Skelestruder V1 for Prusa MK3 and MK3s

by jltx Mar 30, 2018
Download All Files

Thing Apps Enabled

Please Login to Comment

I started getting ready to put everything together. I bought new Bondtech gears to pre-assemble all I can before touching my MK3.
Then I started reading through the comments and got scared. I have no idea what changes are needed to fix the firmware and how to make them. I've never done any firmware compiling in my life.

For what I've figured reading, there is still a problem with the filament unload, or have this been fixed?

Anybody can help me? At this point, I have all the parts but I'm not sure I will keep going without guidance.

This is what I have:

Stock MK3 with indirect filament sensor mod. I'd like to use the indirect sensor in this new extruder.
I've printed the parts for the 3.5 gear ratio.
I have bought a new stepper motor with 0.9 degrees steps.

I'm hoping somebody already has this configuration and can help me with the firmware changes.

It’s not bad. Keep in mind comments are where people post problems. Most people go smooth. Just follow instructions on wiki for firmware. Hop on discord to get some help. I’m in mountains right now so losing communication for a week.

Comments deleted.

Not done with my build yet, but with other community firmware (i can recommend the FW by Josh Jones aka vertigo235) and a 1:3.5 gear ratio with 0.9 motor, you would use 1/8 microsteps and 415 esteps.
If put into a gcode file: (If not, run these from octoprint/pronterface/whatever console without ";")
M350 E8 ; set Microsteps to 8
M92 E415 ; set esteps to 415
M500 ; save values to EEPROM

The M350 command has only recently been added in FW 3.7.2 - either original or forked.

I have almost finished upgrading all to your amazing work! thank you very much for this great effort and sharing this with the community!!!!!
All works great for me excet i cant get the extruder to work at all..
Ordered the motor you suggeted from the aliexpress, all installed and good.
Snapped the cable from the original (i know.. i know..) Judging by your pictures the cable colors are not in the same order as yours. anyway this is how i connected them:
blue from motor - to orig. prusa blue cable
green from motor - to orig. prusa gren cable
red from motor - to orig. prusa red cable
yellow from motor - to black prusa cable

what is wrong with my wiring :)

You need to look at the data sheet for your motor. But looking at your motor connector, I'm guessing the two coils are: Red A+, Green A-, and Yellow B+, Blue B-. The stock LDO motor is black, green and red, blue. You crossed coils so no current. Try:
Red - red
Green - blue
Yellow - black
Blue - green

If it turns backwards, just swap two wires on one coil.

That works great now. Great improvement in quality. Cheers mate.

Increasing the printing speed by turning the knob creates an extreme underextrusion like problem. It looks like it is still extruding at the normal speed.. for that to work I assume I will need to patch the firmware and change the values I assume, am I right?

I am also considering updating to the moons motor linked here: https://prusacommunity.com/to-the-moons-and-back/
Would you recommend using that with the skelestruder. I have paid attention that my new motor is not smooth enough.

Has the filaments sensor for mk3s been updated yet?

Added link to a good looking option on the front page.

I am testing one for the MMU2. Then I will get the single version going. This thing is 40mm long with connector. It's stupid big and awkward to fit anywhere. Taking longer than I planned.

Been using it for a few weeks. Can recommend.
EU dudes can buy the shaft now at lecktor: http://lecktor.com/en/extruders/229-skelestruder-shaft.html

No I did not see it, I wen't straight to this skelestruder page.
Now I have no excuse to not buy a mosquito.
Just a quick question, how does the Omega nozzle look?, I've heard that due to the off-centre heat block, it can foul existing solutions.

It should just barely fit. But need testers to confirm.

Do you have any plans to use a Mosquito all metal hot end in a Skelestruder design?: https://www.sliceengineering.com/
Or is is simply too different from the e3d v6 to be able to work.

I really like the fact that the PINDA height is so adjustable and your SP_Inlet_STD keeps the PTFE tube in place and the SP_Bcart_beta works so well, cold pulls are so easy and I do them every time I change filaments and the fact that you can see the hot end through the Omega nozzle which also works very well.

Did you follow the mosquito link on the front page?

I’m glad you are appreciating those features. Tell your friends!

@jltx im not sure what went wrong but My mk3s started stalling during print. It wasnt doing this before. Any idea of a fix for this? It seems like it’s taking it’s time processing data before moving to another position.



E steps: E2010.25
Slic3r 2.0
Custom gcode: M907 E400
Retaraction length: 0.6 mm

Your esteps is very high. I’ve seen others who got some stalling when they went that high. Seems the firmware might have trouble keeping up. Try cutting micro steps in half and then lower esteps to see if that clears it up.

Can you teach me how to cut my micro steps in half? In slic3r or pronterface?

see posts by mikegi below.

@jltx - I received my MK3S 2 weeks ago and will hopefully be starting on a Skelestruder soon, I was wondering if you had received your MK3S upgrade kit yet since I only have the new filament sensor and am waiting on whatever changes are going to be made to support it. Or do you think I should just order one of the old filament sensors and switch to that?

@jltx - Thank you for answering all my questions below, I will be sure to post a make once I get everything completed.

I personally like the old sensor better because it can detect jams. The new sensor is a step back because it only detects runouts. That said, the old sensor wasn’t perfectly reliable for everyone and it seems the new one is more reliable. So it’s up to you. For now they support both in the firmware, but could stop at some point, or only add new features to the mk3s version, though I’d be surprised.

@jltx - I have one other question as I did not see it addressed in your build document. With the MK3S kit, we have a nylon rod that runs through the E-axis cable bundle to help it move correctly, this is not mentioned or shown in your build document, do we remove this or is there a place to attach it? This is shown in Chapter 5: Step 55 and 56 of Prusa build guide for MK3S. I am leaning towards the fin option myself.

I'd leave it in the bundle to provide some support.

Yes. Note that if you go that route, I'd recommend the indirect mod which requires specific bearing(s). See docs. But you can go direct which is easiest and see if that works for the filaments you use in printing.

My 17hs08-1004s is getting very hot at 400mA while printing ABS on enclosure with ambient temperature of +/-43ºc, i can't stand my finger there more than 1s it burns! Motor temp: 60 to 70ºc. You think hanpose will run cooler since it have larger body?

How did you pick 400mA? You try lower?

Didn't try lower because i already run at limits, with volcano i can print at 23mm^3/s volume and if i increase lets say to 25 or 27mm^3/s it start to skip, but with 600mA gives same volume rate. Stock parts and motor gave-me 27mm^3/s max before. Sure i can lower the current but that may end in lower volume capability or performance, on low layers heights i have max speed of 150mm/s. I don't know how motor run so hot if is 1A, it should run at 50% without that excessive heat. How much current do you run yours?
Note: I print with 0.45mm layers, at 1mm extrusion width, so motor have a hard life pushing filament. 0.6mm nozzle

Noticed BLTouch was mentioned and not added; any chance you could add .step files for the parts involved do make my own adaptation easier?

I already opted for BLTouch for my i3 Rework clone, and will want to occationally use glass so i don't think the inductive probe is a good.

Oh, and impressive work on the design!

Bltouch is huge and no good place to put it that doesn’t block airflow or run into ends. I gave up on it.

I understood that from the previous request, thats fine, 10-20mm travel lost is not a problem for me, I think it might fit on the carriage as I have a little extra travel on Y anyway. Its just so slow to work with .stl files in fusion, .step would be much easier to change. But I'll figure it out.

Any advice for printing the spider? I've tried more than a dozen times and have yet to get it to stick. I even flipped it over and tried printing with supports, but it didn't like that either!

I haven’t had specific trouble with that part. Other parts printing ok for you? What filament? You try cleaning with acetone first to remove any residue?

Each of the little lines refused to actually stick onto the heatbed. This despite acetoning it, adding glue, adding tape! No matter what, I couldn't get it to stick to actually finish the print. Then I flipped my steel plate over...and it came out perfect.
I dunno. Printers can be weird sometimes, I guess.
Additional question: which pocket on the extruder should I put the belts in with my full taurus upgrade? Stock or bear?

This is a really good design. I've been printing with V1 for the last month, with some mods.

  1. 40/16 gear ratio, with 122mm belt (couldn't easily get 130-140mm belts).
  2. Cables exit upwards. The filament cover is modded to support the cables and prevent it from interfering with the wheel (smaller wheel helps here). This gives me 229mm Z height (with a modified Z-axis).
  3. Switching between v6 and Volcano is easy if you forego using socks and trim off where the sock screw goes on the EcageS. I unplug my cartridge + thermistor (I have connectors for this), remove the fan shroud, unscrew the heatblock.

Cool. I would have been happy to make a wheel for you but seems you are self sufficient.

Very loud noise from Skelestruder. Any tips to get rid of the noise?


Decrease your max feedrate:
M203 E120 (default) - divide by your gear ratio (or greater).

If not, some commands will exceed the extruder motor maximum speed, and you may see the wheel go the wrong direction on retract/deretract and during jerk movements. These are the loud clicks.

I still get quieter clicks on fast retractions, but without abnormal movements or artifact.

will you be able to teach me on how to decrease the max feedrate? is it thru pronterface?

You can use pronterface, but can also just add a line to your start gcode in slic3r.

I would try lowering your motor current since it may be too high for that slim motor. A couple others had a similar sound and it was fixed by lowering current.

Could you please teach me the way how to lower my motor current? I really have no idea what is the command or how should I input it in my slicer settings.

Try adding this to your start gcode in slic3r:
M907 E400

That’s 400mA in theory. Increase if you get skipping or other feed issues. I don’t know what your motor needs.

Thank you very much! It works at E400. Now my skelestruder is much more quiet than before. My extruder motor does not over heat anymore.

I actually have the stock motor installed. (the 0.9 slim motor has errors with unloading filament)
Should i lower the current for the stock motor too? How should I do it?

yes, because you have multiplied the torque by 3.5x so you don't need so much oomph. That is likely your noise culprit.

I had a similar noise, turned out to be the Bondtech gear screw. It was this tiny bit loose, and it was hitting on the shaft when switching directions. It eventually got mega loose, the gear stopped rotating alltogether, and i got a hold of it. Use a good tool to tighten it properly, cause if you strip it, you are pretty much done

Where are the M6 covers? I can't find either of them.

follow the MMU link in the details section,

For some reason the Filament sensor that came in with my MK3S is different and will not fit in the printed part. Is there a different printed version for this filament sensor to fit in? I'm almost done. It's just this sensor won't fit in.

Yes, I need to add support for new IR sensor. Currently only laser sensor supported. You can disable sensor for now if you don’t have the laser sensor. I am waiting for Prusa to ship my sensor upgrade that I ordered in February so I can make and test the modification.

Can you give us some guidance on what firmware to use in the meantime? I am getting ready to do this build and I want to be prepared in case you haven't gotten your IR sensor revision done by the time I'm ready to start.

I am using 3.5, but I will be switching to 3.7 soon. I will get a patch out for that. You can see my old patches and probably just make the same modifications.

I've modded 3.7 firmware based on your patches, Everything works fine. My only issue is that the filament unload sequence is way to slow. Is it supposed to be like that?
If not, could you give some guidance on how to tweek the parameters to make it faster/slower/whatever?

did you find a better config for faster unload?

I just found time to try stuff. I used 500 / 60 instead of 100 / 60, and PLA unloads just fine from 205C degrees. Full unload sequence takes around 15 seconds, which is fine for me.

because the tube is shorter, I was giving it more time for the tip to cool to guarantee a clean unload. It could probably be made quicker and still work. The main slowness is the speed of the cooling move:
current_position[E_AXIS] -= 10; plan_buffer_line(current_position[X_AXIS], current_position[Y_AXIS], current_position[Z_AXIS], current_position[E_AXIS], 100 / 60, active_extruder);

Note the original line speed was 1000/60, so this is 10x slower. Try adjusting that. If you go too quick, the tip will still be soft when it tries to pull through the gears and it may jam. Not too hard to clear, just pull the cartridge.

I need your help big time with this...

Filament will not load because of the sensor being disabled.

Now if i try to choose the "unload filament" option. The stepper motor (nema 0.9) loads the filament instead.
(I'm not sure if there is a problem with the wiring of the motor because it was sent to me with no cable and i just bought a cable online.)

My stock stepper motor works properly...

This is so frustrating...

What Mikegi said. Need to swap a coil. There is not really a std as far as I can tell so cable you bought may be reverse of what your stepper expects. Easy to fix though.

Did you turn off filament sensor in settings menu?

Reverse/ swap to two wires to one of the phases.

Not sure what connector your cable has, but most likely, use something pointy to press the metal tab while putting some tension on the wire and it should slide out. Be gentle.

Thank you guys for your response. However, i’m not sure how to change the coild (which coil should go on which slot).
I took a snap shot of my stepper motor and it’s currently cable set up. Hope you guys can help me figure out the right cable connections for this type of cable.

Just reverse the connector into Einsy so blue wire is on bottom.

Thank you for the tip to fix the cables. The Extruder is now up and running.
Problem i have now is when i try to unload filament. It will beep loud and display “Error static memory has been overwritten” and then the printer will reset or do crazy things. Habe you encountered this problem before? Anyway to fix it?

PS: i did factory reset + install latest firmware. Still same result after unloading filament or changing filament during a change color print.

I don't think I have seen this before.

A quick look through the firmware source shows this is printed on the display by a routing called stack_error()
This seems to only used by SdFatUtil.cpp, the Library for handling the Filesystems on the SD card.

If I remember correctly, the Arduino FAT library has some limits to the amount of files or directories it can read from a given directory. FIlenames are likely stored in RAM memory when navigating through the list, and there is probably not enough RAM allocated or free for larger directories.

Try removing the SD card. Or use a freshly formatted one. If you want to keep old gcode on the SD, make new directories for older files and move them there.

Thank you for your response! I did format the SD card but still the same result. I tried removing the SD card and tried to unload filament on standby mode and same result. Either the printer will restart or will display the same message.

hoping it's not a motherboard issue.

Here is a video: https://youtu.be/QbYwcjKh0WU

During the unloading filament process, the extruder rotates so fast and it doesn’t even pull out the filament (filament stays in the same place) and then will blow out an error.

would it be because of the change in E-steps? I have to change my E-steps to E1979 after doing the E-steps calibration by measuring and extruding filament and calculating for the E-steps.

Ideally, the motor driver chips, have thermal and current limits to prevent damage to the mother board and themselves.

Are you using stock firmware? Which printer?

I think it would be good to see what is working and what doesn't, and find the way to the cause of the problem.

Can you browse files on a SD card via the menu?

You could go to the settings menu and check that you can move the axes ( start with X, Y, Z and then E )?

Can you adjust the fan speeds?

My next suggestion would be to see that cables are plugged properly and check the new cable and new motor for shorts.

Also, you could check again that everything works properly with the stock motor. Then you can focus on figuring out what the problem with the motor is.

One possibility could be that there is a problem with the stepper motor or cable ( maybe a short) and when the E stepper is switched on a voltage drop or glitch in the power for the Microcontroller which also powers the Display and SD card reader

I had a quick look at the diagram for the motor that you posted... Notice that you have a 6 pin stepper and this goes to a 4 pin cable.

At least from memory, the cable in the drawing looks ok... So if your cable equivalent, the only thing that might need to be changed is the polarity for either the A or B phase.

Another thing you could test is the new motor with the stock ESTEPS setting ( 400?) The extrusion will of course be off, but at least you can see that the direction is correct.

I remember some discussion in the discord group having to do with problems with some E values and certain motors... Maybe 0.9 deg/step
that may have needed a workaround. There also some discussion in this comments section below....

Are you using stock firmware? Which printer?

  • Yes! currently using 3.7.0-2201 (i3 MK3S)

Can you browse files on a SD card via the menu?

  • Yes! I can print files via SD card menu.

You could go to the settings menu and check that you can move the axes.

  • Yes! all axes are moving with no problem.

Can you adjust the fan speeds?

  • Yes! I can do that via settings and no errors.

I will try to change back to the stock motor and see if I get the same error..

Did some tests and this is what i got...

  • reinstalled the stock motor and factory reset + re flash firmware. (stock E-steps 280) no errors.

  • installed the new slim motor and factory reset + re flash firmware. (stock E-steps 280) no errors!

  • changed the e-steps to M92 E980 with the same slim motor. runs ok and no errors! However the extrusion is so slow.
    (needed to unload filament option twice to completely remove the filament)

  • finally changed the e-steps to M92 E1960 with the same slim motor. Faster extrusion (Loading filament ok!) unloading filament... BOOM! Error occurs!

BTW im using 3.5 gear ratio


M92 E980 and decrease the microsteps?

It may be possible to use GCODE to modify Microstep settings

  • M350, M351 ... but I think these operate on all motors
  • M42 to manipulate microcontroller pins directly

Someone please correct me if I am wrong, I think the best options are...

  • Use firmware with decreased microstepping on E (with 0.9 deg/step)
  • Change to a 1.8 deg/step motor

I could not resist digging into the firmware source...

Here is how unload filament is defined in the Prusa firmware ( so what is executed when you click that item in the menu)

// Unload filament commands
#define UNLOAD_FILAMENT_0 "M83"
#define UNLOAD_FILAMENT_1 "G1 E-80 F7000"

I dug a but more, but figured it's time to stop since I could dig all day...

F7000... is fast...

At some point, the firmware make some calculations to do with timing the steps and calculating acceleration, checking that a move doesn't exceed a limit, etc.

Most likely, 7000 and 80 are multiplied by ESTEPS somewhere. With ESTEPS at 1960, perhaps the limit of a register is exceeded perhaps this error is not handled by the firmware and leading to the unpredictable behavior in your video...

Without changing the firmware, or to a 1.8 deg/step motor, maybe you could run with default firmware and current settings
and avoid using menu option to unload filament... And instead maybe a GCODE script as follows for example...

G1 E-40 F3500
G1 E-40 F3500

Assuming there is nothing else in the GCODE that will cause a problem.

Thanks to mikegi for all the help.

Wow, I've never heard of that error. Honestly I have no clue. Since you are using a 0.9 stepper you need to double Esteps, so 1960 is correct. Others are using that. May be a bug in 3.7 FW which I have not yet used because I am still waiting for my stupid MK3s upgrade (Grrr!). Anyway, worst case you can use the stock motor until it is solved.

The unload sequence needs to be adjusted since the distance is shorter. But that just leads to occasional filament sticking in gears, no errors. I will take a look at 3.7 and try to get a patch out.

No problem... looking at Marlin source is more interesting than what I should be doing instead. :-)

Managed to find a local source for the little ball bearing for the indirect sensor... so I should be ready go get my skelestruder going next weekish... ( found them at a RC hobby store which will ship small orders by post ).

Is there a direct way to adjust the microsteps? I'm not familiar with it (never done it before)

Unfortunately it seems that the M350 GCODE is not enabled on when compiling for Einsy controller boards. (as least in Prusa's fork of the marlin firmware)...

The MK3S microstep settings are configured in the firmware..

#define TMC2130_USTEPS_XY   16        // microstep resolution for XY axes
#define TMC2130_USTEPS_Z    16        // microstep resolution for Z axis
#define TMC2130_USTEPS_E    32        // microstep resolution for E axis

So if it was possible to run GCODE to set all motors to 16 microsteps. The others should be the same as default and the E-steps would be half.

Maybe someone else has an idea.

I read that the 3.7.2 FW Beta has the M350 GCODE.

What I did was I've just modified the default steps and µSteps in firmware and then used M502 to reload the default steps from firmware. (Don't forget to save them to EEPROM with M500.)

R4G Extruder for Prusa, a DIY Bondtech Upgrade


Just wanted to let you know that in the MK3S extruder the length of the PTFE tube was reduced to 44 mm (from 50 mm).


now that I've installed the Skelestruder, I realized you need to cut the PTFE tube. I measured 39 mm. is it correct? I'd add a remix for that to perfectly cut it

  • Posted with 3D Geeks Thingiverse Browser App

I’m actually stuck here... after installing the bondtech gear, i found out that the PTFE tube from MK3S is sticking out. Should i cut it? How much should i cut it?

Correct, you need to trim it to match the inlet. I just make mine a triangular tip and it works fine.

I guess you have used SP_Inlet_ptfe ?

you are supposed to cut it so that the height does not exceed the inlet. I read 42 mm somewhere but I cannot vouch for it. I guess it's easier to stick it through and then cut it to length. then you need to cut the sides that tube so that it follows the shape of the inlet.

Hey @jltx, would a 71T belt be ok to use with the 56/20 gearing?

I think so but will have to check. You found that size?

Yeah. The 70T is gonna take a loooong time to arrive, and i found a 71T that will arrive sooner. Will test out as well when it's here.
I think it is gonna be ok, cause the 56/20 took reasonable force to tension properly.
I'm currently using a 70T flex printed version i designed (works just fine, but a few layers disintegrated, i think because of rubbing on aluminum pulley)

Reprinted wheel and 56 tread using polycarbonate and 0.25mm nozzle as I was never quite 100% satisfied with the PETG version, but it was most likely due to my calibration etc, the PETG wheel never quite sat central on the shaft and the tread never quite sat tightly enough on the wheel, this resulted in movement as the whole thing turned about the shaft.

With the polycarbonate, the tolerances are a lot better, I also used 100% fill (I know, no need) but I wanted the strongest parts possible.

Here is what the prints looked like. The first is the PETG in black and the second is in PC in clear, it took quite some force to assemble the two parts, but it runs nice and central to the shaft now.

Errmm, maybe thats the other way round. :) Actually the PETG part is for the Y axis wheel, Oh well, you can still see the detail achieved using PC, I've added a PLA print close up as well.
The PC and PLA are the same perimiters, layer heights, line widths etc, just different speeds and temperatures.

Is your extruder compatible with the bear extruder x-ends?

Pardon my ignorance, but what is the cleaning brush supposed to do? Or where is it placed?

It is for soft filaments like tpu. If they are soft enough the drive gears will shave off bits that stick in the teeth and eventually reduce the grip. The brush helps remove those as they form.

Is there a link to 5x45mm shaft with two flats? The existing links doesnt have the flats pre made. I have no grinder to make flats, if there is a seller that sells it pre made please do let me know. Thank you!

The US and UK links on ifixit should get you that. I just double checked the US one and it worked.

The UK link is useless to most people unfortunately. It links to Misumi EU who won't sell to normal people, they will only sell to businesses (you have to provide business registration details to them).

Do you have another source of "read to use" shafts? As that seemed to be the only source.

Others have fabricated a business name and been able to purchase. Unfortunately they are the only company I know that offers custom shafts. But it really is pretty easy to make your own and feel a sense of accomplishment.

I actually got the heli shaft, It didn't have the flats. I will get the US one next

I'd keep the shaft. that's what I am using. just get a file from a hw store and spend a few minutes. maybe like this: https://www.homedepot.com/p/Corona-6-in-Bastard-Cut-Mill-File-FE405506BC0CD/205980201

Any chance you could put up the xclamp version for hitting endstop? :) Thanks so much!

Comments deleted.

By assembly i noticed that holes on carriage have a diameter of 3.2 mm. There is no way that a M3 screw will hold in that. I used some epoxy glue to fix that. Is it a design mistake? Or am i doing it wrong?

Definitely wrong. You are supposed to slide a square nut into the slot behind.

I'm currently printing the parts to 100% infill. Would that make any issues regarding the performance later on?

That’s generally considered a bad idea. Do a search on infill testing. Strength does not increase beyond a point.

Thank you for letting me know. How much infill percentage would you advise me to go for?

With gyroid maybe 40% max. I’d have to look. I usually look to see if I get one wavelength for the smaller (but not smallest) sections of infill.

Regarding one of the motors, in your parts link you specify the Stepperonline 17HS08-1004S which I have and works, but after reading this thread on Prusa forums:
and also noticing the slight under extrusion issue on startup, this motor Stepperonline 17HS10-0704S might be a better choice.
I have ordered one and will test it and report back.
Seeing how my motor current is set at 300mA on the motor fitted now, this replacement motor is still well within specs as it is rated at 700mA max current.

Reviews are kinda scary on that one. But what attributes does it have that lead you to think it will perform better? Lower current?

But let me know how the results turn out.

I'm hoping that due to it being a lower current motor, it will have more windings and may produce more torque sooner (Slew rates of driver) than the higher current motor. Granted the 1A motor will produce more overall torque and at higher speeds, but we are not using this stepper anywhere near speeds where any of this matters. in both cases maximum torque occurs at around 75 rpm, the 100 motor produces around 1200 gcm torque at 1000 ma whereas the 070 motor produces around 1060 gcm at 700 ma, so if we apply a very rough rule of linearity to both motors, at 700 mA the 100 motor would only be producing 840 gcm.
I know steppers arent quite as simple as this, but if it responds quicker and produces more torque at slower speeds and currents, then surely this can only be better.
Well, I'm going to find out :)

Here we go, finally got the information together on why I think this motor 'may' be a better option. Since it is a lower maximum current motor, I'm hoping it gets closer to its 63 % therefore producing more torque than the 1A motor will, at about 441mA to be precise which happens to be right in the ball park for what Prusa are driving this extruder motor at, ideally we want to set the current to this for maximum torque.
Granted we could drive the 1A motor at 630mA, but asuming the rate that the power increases from 0 to maximum remains constant, the lower current motor will achieve maximum torque sooner than the higher current motor also more current = more heat which we don't really want anyway. ( I may have this all wrong, but only one way to find out! - Everyday's a school day!)

Taken from here: https://www.motioncontroltips.com/what-stepper-motor-type-is-best-for-high-torque/

Hybrid stepper motors have the ability to produce higher dynamic torque, but in order to make this happen, the electrical time constant of the motor must be taken into account. The electrical time constant, τe, is the amount of time that it takes the current in the windings to reach 63 percent of its maximum value. Because torque is proportional to current, the current needs to be at its maximum in order for the motor to produce maximum torque. And, maximum current (and hence, torque) must be reached before the current is switched to the next phase.

In order to achieve this, the voltage supplied to the motor must be higher than the motor’s rated voltage. The higher the ratio of supplied voltage to rated voltage, the faster the current will enter the windings and reach its maximum value. Hence, maximum torque can be produced before the current is switched and the cycle repeats.

The Stepperonline 17HS10-0704S works perfectly, but then again so did the 17HS08-1004S.
I thought that after reading about the slight underextrusion upon restart that it might help with bridging, or rather just the first strand which most of the time apears to start late so drops, but alas, it made no difference to this.
I'll leave it on for now and run some more tests when I get a bit more time so see if there is a difference in print quality.
Just for the record, motor went straight on, no wiring issues or anything negative. I did order it direct from https://www.omc-stepperonline.com/
at £8.05 and a couple of £ shipping in the UK, I thought it was worth a try, the 17HS08-1004S is £7.82 from here if anyone in the UK is after one, they also have supplies in the US (cheaper than UK) and worldwide.

Had to reprint the Omega nozzle in Polycarbonate since PETG softens and droops whilst printing T-Glass PETG at 270 Deg C.
Printed using Rigid.Ink Natural Clear PC filament, just had to play around with supports a little to minimise them and make them possible to remove and tweaked some Slic3r settings to improve my print.
Came out really well without any holes or issues or layer de-laminating or warping. Printed at 295 Deg with 100 Deg bed temperature and no enclosure onto a Buildtak surface, taking it slow so it took 2.5 Hrs.

Now to get bridging sorted with PC and going to reprint more of the skelestruder in PC starting with the SP Spider as this has a little too much flex in PETG for my liking then going to try the cage, but only if I can get the bridging sorted out first. Maybe the wheel as well as this is ever so slightly off centre when tightening the cap head screw onto the shaft, again I believe due to me using PETG.
There is no actual problem with my skelestruder as is, but I like to get things just right. :)

Nice. Haven’t tried pc for omega yet. Glad to hear you got it working. Be careful with eCage though. Others have had problems where it cracks eventually since pc so brittle. Spider might work.

This Rigid.Ink PC doesnt seem to be brittle at all, if I were to describe it, it's like PETG on steroids!. I am still testing, but where a small part in PETG will break, I can't break this without some serious abuse and even then it's not from cracking it's layer adhesion faults.
This stuff is seriously strong, slightly flexible but not as much as PETG and seems to print a lot more accurately than PETG.

Still playing with bridging settings though, but day job gets in the way, so have to leave it for another day.
When I've got reliable settings and sucessful prints, I'll post my details.

If 39mm rods would work / could be adapted to work instead of the 45mm, Ikea may be an easy alternative source for some: their Ivar shelf pins (#101324) are 5x39, although not precision ground. It's one of their oldest product lines, so spares aren't too hard to find online. I've also seen compatible pins in my local hardware store, but YMMV.

I would be hesitant to use something like that. It’s probably just simple rod stock chopped to length. May not be particularly round or straight, though fine to hold a shelf.

Any chance you could post a version of the indirect filament sensor for 3x6x2 bearings? I cannot find 3x6x2.5 locally to me unfortunately. Would thickening the cartidge by 0.5mm suffice?
Could i maybe just place a 0.5mm spacer in between?

I didn’t know they made smaller ones. Anyway, I see no reason this wouldn’t work as is for the indirect sensor. For the hybrid it might also work as is. Which are you planning?

Hmmm.... Turns out my local shop lists MR63-ZZ as 3x6x2mm, although they actually appear to be 3x6x2.5, based on the part code, the one you use. I'll have to check in person then. I'll post back with results

Hi, do you have any plans on adding support for the new MK3S filament sensor for both single and mmu configuration?

I have one printer with a V6+mmu and soon another with volcano and I like your design so much that I want to try switching both printers to your extruder.

Also I want to use cable chains in the future, which I need to adapt to your extruder then. I'm almost afraid to ask but do you also plan to release step or source files of your design? Or maybe only the part I have to adapt the cable chain to.

Yes. I am just waiting for mine to test. still backordered.

Mine is backordered too. But hearing that you will integrate it, gives me confidence in switching to your extruder.

Any thoughts on my question about releasing step files, since I want to try to adapt a cable chain to the extruder?

Another question I havn't found an answer yet...
do you recommend using a 0.9° or 1.8° pancake stepper motor for your extruder?

I am having an issue with filament slippage. It looks like there are a few issues. The first one is that it appears like the filament path is not aligned and that the filament is further away from the bondtech fixed gear then it should be. The other issue is that the bentech tensioned gear is hitting the piece on top of the hotend and not allowing it put pressure on the filament. Both issues lead me to believe that the piece on top of the hotend did not get printed correctly and is offset (away from the fixed gear). Not sure how this would happen though. Has anyone else seen this issue?

Something is definitely wrong with your assembly, possibly caused by a bad part print as you surmise. I have not seen that fail before. I would remove the cartridge and watch the feed to see if you can see why the idler is not meshing fully. If that looks ok, then i would suspect the cartridge. Which config are you using?

Hey jltx! Can I add you thru Discord? I'm gonna need your support on building this project soon. Just prepping my PETG and all the HW needed.

I am on the skelestruder channel in the wasteland on discord.

Thanks for the reply. I used the latest stl's from February. Not sure what you mean by "config".


which cartridge, etc? direct?

I figured it out. I reprinted the part above the hotend and it is better. The other one was slightly off center. Not sure why.

Also, I was calibrating the extruder steps by extruding 100mm of filament using the OctoPrint GUI. It was feeding the material way to fast and it was backing up in the hotend. I ended up using the gcode commands and it didnt grind the filament anymore.

As a side note, I used this on an Ender and I had to set the steps to half of the value that is in your table.

Any chance of a version of this that takes a Mosquito hotend (preferably the Magnum)?

Yes. I am working on a mod for that. It takes a bit more modification than I was expecting since it is wider, plus some other integration issues according to slice. It I expect I can get that done eventually. I’ll have to look into the magnum.

It looks like the magnum is exactly the same of the regular mosquito but with an insulator on the block end of the heatbreak.

Has anyone weighed a mosquito yet? Slice doesn't seem to be publishing a mass spec.

For the record, I would consider the mosquito a worth while upgrade, even if it ads a few grams over a v6, if it meant not having to deal with loose heat-breaks anymore.

Anyone tested the max exutrsion volume i can do? (mm^3/s)

Any ETA on the hotend maintenance guide? I'm really interested in making this, but as of now, i'm scared that it is gonna be a pain to remove the hotend in case of jams, just as the original MK3 assembly is

Soon I hope. But curious what sort of jams you are having that require full hotend removal? I’ve never had that. At any rate, nothing is as hard as the original mk3 assembly. Their new one is supposed to have improved that. Mine does require you to remove a few more parts but they are all screwed on and the hotend just slips out. But I’m guessing you won’t need to really do that unless you are upgrading something. This has better hotend cooling which may help with your jams.

Well, my jams happen right below the Bondtech gears right now, just with PLA, from motor heat, and in prints with loads of frequent retraction. That will definately wont happen, since there is no metal connection on your extruder. But since i've busted my ass of with current MK3 assembly, i'd like to have some peace of mind on the subject :D

Only now i notice that my PINDA wheel have no clicks... That due a missing detail with 0.6mm nozzle.
For people with 0.6mm i have attach a fix
See PNG, my fix: Left part, Original: Right Part

Today i got a false runout after 50m of printing, using beta cage.
ABS black filament

For those that were wondering about the beta filament sensor, here are some pics of mine printed with PETG, this is after quite some use and plenty of changed filaments and a few cold pulls as I do one every time I change nozzles.
There is no issue using PETG, but I did print this with a 0.25mm nozzle.

The filament sensor is quite dusty as well so I used a camera lens blower to clean the dust.

I will reprint this and see if the bearing posts come out slightly bigger as you can see, the bearings are quite a loose fit.

I printed that with ABS and 0.6mm nozzle. Come out perfect. It seens your piece is a bit deformed. Use less speed, PETG need to cooldown a bit. Or plate with other pieces

The part SP_EcageS_r1 is susceptible to heat damage, so when assembling make sure your heater block is not ever so lightly rotated so the back is close to the part else it will soften, droop and end up touching the block at which point it will melt.

What did you print your cageS with? I made that a separate piece so you can print with high temp tolerant filament, e.g. ABS, PC.

I used PETG, I must have missed something in the instructions about using ABS or PC.
Since I don't have either of these materials and will have to buy some filament, which is the best for temperature and print-ability?



Sorry, you can use PETG, just that if you print at high temps frequently or use enclosure, you may want to use something different. I used PETG for a while. It did sag a bit but worked fine for most of my prints which are PLA or PETG. I'm now using PC version but not critical for my needs.

I've been printing with ColorFabb XT which prints at 260 Deg C so this may be why my PETG part has softened and sagged.
I'm going to buy some Colorfabb HT in white as this will have the most titanium dioxide to give it the white colour. This will also make it stiffer than clear filament and ColorFabb HT is good up to 100+ Dec C, whereas normal PETG is only good up to about 70 to 85 Deg C.

Then again I've just read up on PC and maybe I would be better off with this but what is the smell like?, is it as bad as ABS?.

I would try HTPLA+ first. You have to anneal and it will shrink about 1-2%, so print a couple versions, one regular, one at 101 X,Y scale. That can go over 100C and is easy stuff to print and work with.

Reprinted it using Rigid.INK Poly-carbonate and 100% infill.
You have to get the printing temperature up to 295 deg C for this stuff.
When I took the head apart to fit the new PC part, I found the original SP_EcageS_r1 back had completely gone and melted all over the heater block!, so definitely a good idea to print this part with a high temperature filament.

I found these cleaning brush with 19mm diameter.

1) https://www.aliexpress.com/item/100PCS-Steel-Wire-Wheel-Brushes-Dia-19mm-22mm-25mm-Dremel-Accessorie-for-Rotary-Tools-Polishing-Brush/32864563404.html


1) have the same material (Brass), does stainless steel sill work or can damage hobb?
2) This is also apropriate because is two units instead of 100 which i don't need. What the best hardness to use without problem?

Tested new beta sensor with clear PETG fialament, it work's very well. Still a short test but stock would fail just after load when your keep going :) Great work!

Part cooling is touch volcano heat block on one side, the block is vertical and parallel (No side rotation), is it normal to hit the block and do i need to rotate it? I guess it will melt due direct contact. What's wrong?
Note: part clip is hook to shroud

Off this question but: Another sock holder should be added at pinda side for Volcano

It should not touch but is close with sock on. You can rotate block slightly though should not be necessary. You can see gap in photo above.

I find the volcano sock holds itself much better so screw not needed like v6.

And can you tell me whats wrong? Everything else fit in place. If i place part clip under the shroud (Touching each other) it looks like the correct position still i don't know...

updated. should be able to print with 0.6 nozzle now. let me know how it goes.

I’m impressed you could print that in ABS. It is closer on one side because the volcano block is asymmetric and cooling is centered on nozzle. But that looks right now. You can rotate block a little for more clearance.

The yellow spring sheet was detaching print near end or 80%. But my new black sheet as you can see on picture have very good adhesion, much superior to the yellow one (Earlier ones) it stick ABS like PLA and i think my brim was not usefull at all :)
Also i will replace X and Y idler with that GT2-40T 6mm: https://www.aliexpress.com/item/1pc-2GT-40T-with-teeth-or-no-theeth-Idler-Timing-Pulley-3mm-4mm-5mm-6mm-Bore/32894094735.html The printed ones don't close all the sides and not perfect, it twist when moving

Yes, the b5 version for volcano touches the corner of the shroud so this could cause it to not seat fully. I will update to R1 this afternoon. In the meantime, how did you print with 0.6 nozzle? I would have to thicken more which would restrict airflow.

I've "Make Solid" and add 0.1mm thickness in MeshMixer to able to print, use ABS.
Your new model slice well, will print now and report later

I think the b5 version for volcano is a bit off for r1 cage. I will update to an r1 version (I meant to anyway).

i use it without problem with volcano, mine is rotate. printed un petg it handles 240° without problem.

  • Posted with 3D Geeks Thingiverse Browser App

The printing and assembly instructions on IFIXIT appear to have vanished as of 12th March 2019.

OK, I got it fixed. Link should work now.

It appears one of the ifixit admins moved a section from in progress (which should be good) but has messed up the link. I’m looking into it. In the meantime you can get to the intro link here and chain them from there.

I thought the link I posted above I got from the thing details New Instructions link was supposed to take you there.
Oh well. at least anyone else reading this can now find them.
Thanks again



But you are correct, the link should redirect people for the right page

Does the first link still not work? Works for me and I haven’t received other feedback. Not sure why you can’t get there if you can see the second link. If you back up one level, that should be the same URL.

Works fine for me now as of 14/03/19. Thanks

@jltx : It worked for me, but now i'm just seeing category 3 on the main page...

Can anyone upload volcano omega with thicker walls?
Edit: I have thicken the walls by 0.1mm, it now print with 0.6mm nozzle with 1 to 2 walls. I will print and test out
Edit2: It works perfectly, printed in ABS

I finally got around to printing and installing one of these today. I have a Titan Aero installed on my frankenprinter (wanhao I3 plus with all the guts replaced) and this reminds me a lot of it.

Here are a few notes:

  • I didn't see an obvious way to use the mk3s filament sensor. Am I missing something?
  • Some of the screws sizes seemed off. In a few cases longer screws than what was listed was required.
  • Printing the spider was hard. The lines on the front were the most difficult because they didn't want to stick to the bed.
  • Using the stock prusa motor. Not sure I want to lose torque switching to a smaller one and I like that prusa fine tuned the TMC drivers for this motor.
  • Its loud for retraction/extrusion switches. Not any louder than my titan but if you are used to a quiet printer, expect some more noise
  • Printing at higher speeds seems more stable. Could be in my head, but I am happy with it.
  • Printed all parts except the knobs and inlet using the black prusa PET that I was sent with my mk3s upgrade kit. It was very stringy and I needed to manually clean up each part.
  • Used filamentone gint PLA+ filament for the knobs/ptfe inlet. Mostly because its my favorite, strong, and easy to print.

My rod was ordered from misumi using the link provided however the grub screw in my bondtech gear wouldn't screw in flush. I used a dremel to file the notch deeper so I could screw it in more and I don't seem to have any problems after.

I can't remember where I ordered the belts from, but I believe it was amazon. No issues the seem to fit and work perfectly.

The retraction extrusion switches are loud because the motor is being run at its normal full power.
I used a slim motor and had it running at the stock PRUSA power and it clicked loudly, so I turned the amperage down and it extrudes and retracts perfectly and is now nice and quiet and runs a lot cooler.

There's currently no way to use the MK3S filament sensor with this extruder. But I guess jtx will come up with a solution, soon.

PSA for anyone putting this together... A few useful links below for harder to find parts (maybe update the instructions?). These can be potential gotchas, and you should know about it before you do the upgrade.

I got the 1.8 degree stepper motor for the extruder from Amazon, as listed in the Skelestruder intro instructions. This motor has 4 wires coming from it with bare ends. If you are using an MK3 you will need a 4 pin connector to go into the Rambo board. One option is to cut the Prusa motor wire and solder to it, but I am planning to keep that motor / wire intact. The connector for the Rambo board that the Prusa stepper motor uses is 4 pin, 2.54mm pitch with a latch (same as regular dupont, but also has the latch). You also need some 26 awg wire because the wire that comes with is not long enough- I also prefer to use silicone rubber wire which you can find on amazon. You will also need a crimp tool to make the cable. I use the Engineer PA-09.

If you are in the US, ultimachine (TN) sells the 4 pin connectors for a 5 pack for $5 + shipping.

Here are some other links for hard to find parts that may be helpful, if you need them and are building the Skelestruder without borrowing parts from an mk3. For example, if you are building for a Creality Ender 3. I could not find a good link for the springs in the US, but the ones for Germany should, work they just take a while to ship. Other springs may also work that you can scavenge or find at a hardware store, but I'm not positive.

https://zaribo.com/other/47-extruder-springs.html (Germany)

Square Nuts - make sure you get the Din 562 which are thin.
https://www.printedsolid.com/products/m3ns-nut-din-562-thin-m3-square-net (US)

Crimp Tool

Bondtech Gears (clones)

I designed a printable jig support to help with cutting and filing filing flats on the 5mm rod.

You can find the thing Here

Grinding and cutting jig for Skelestruder 5mm shaft
by mikegi

Awesome! I wish I'd used one, my flats aren't as flat as they could be, caused a day or two of head scratching when I started getting strings and poor retractions.

Picked up non printed parts from the post office today... I'll see how it goes. Thought I might get away with a hacksaw and file, but we'll see.
Just in case, bought Dremel cutting discs ... should fit on the proxxon no problem.

Hybrid filament sensor working perfectly.
I printed this in PETG (Rigid.INK black) using a 0.25mm nozzle and 0.1mm layer height due to it's small size.
I have previously had trouble with Technology Outlet PETG. Not any more. It prints this without any filament detection issues now.

Technology Outlet's T-glass, short for Tough Glass, this new and unique filament has some very interesting characteristics with regards to clearness and strength. It is made from the highest optical quality PET-G material. Unlike other polymers, some PET-G, including T-glase are not considered "transparent". Instead, it belongs to a group of polymers that are considered "colourless" and or "colourless with a percentage of reflectivity". T-glase is also considered "water-clear" and "optically correct".

Pull your pulley carrier out at some point (you don't need to remove it from its block, pull out the whole indirect sensor module) and see if the spring holding the pulley broke. I broke two, rounded some edges and reprinted, and broke that. Perhaps I should try PET or something.

The spring part of the carrier is fine in PETG, although I did print it with a 0.25mm nozzle.

I have my first test print in 0.4 and to break it I had to lift the bearing to a 30 degree angle from it's normal plane and pushing forwards and backwards did not break it, so it will not break under normal operation as far as I can see.

Just to check, I just printed and tried a 0.4mm nozzle print 0.1mm layer height using Technology Outlets T-glass which I know is more flexible than what I printed the first part with, and it works perfectly. I cant make it fail pushing backward and forwards within the confines of the slot around the movable part, and had to bend it up to 45 deg before it failed.

For anyone not familiar with this part, it does not bend up or down during normal use, I am just showing how much abuse PETG will take before failing.

Great! working for some others and myself as well. No trouble printing it from PETG? I thought that might be too flexible at that scale.

Alas, the Hybrid filament sensor fails using a completely clear filament. Rigid Ink Poly-carbonate Natural Clear will cause minor blips, not the usual filament change error issues, but the heating pause issue, maybe if the sensor was angled slightly more this might solve this minor issue.
More than willing to test, but I don't have the original source files to modify the part myself, and am using Design Spark Mechanical for designing parts so can't import the SCL for alteration.
I do think the sensor angled a bit more to look at the fixed bearing edge and outer bearing face, may be enough to work 100% with completely clear filaments, but there's not a lot of room to play with, I would have to find a low profile screw to hold the sensor in place as altering the angle as I have stated, would cause the existing cap head screw to interfere with the moving bearing.

I can look at angling more. But you should definitely use a button head screw instead of cap. I think I put that in instructions. Not sure how you got it to clear.

If I were to make any comments about improvement I would put three raised ribs along the bearing posts so that the bearings sit a little tighter and the same for the filament sensor screw mounting hole, just so the screw has a bit more to bite into or the posts ever so slightly bigger, but the ribs are a better idea as with PETG they will squash a little, we are really only talking maybe 0.05 mm and hole slightly smaller or ribs 0.1 mm on the inside, this is for PETG.
Also possibly a pin hole notch where you can use a pin to pop the plate out of the carrier.

I printed the whole skelestruder in PETG using Rigid.Ink Black PETG, this PETG tends to be a bit more rigid than other PETG's.

I am somewhat committed to building one of these now. Seeing that the Prusa R3 to R4 upgrade is print it yourself, I figure this would more worthwhile and interesting.

Have printed a few parts, have ordered some non-printable parts and a couple more colours of PETG to make the PETG parts are more easy to distinguish. After ordering parts I realized I already some 5mm rod scavanged from a 5.25" floppy drive.

Have gone through the guides on I fix it, and read through this discussion a few times. One needs to read through a few times and look at the parts and files a few times before it starts everything starts to make sense. Part of this is correlating the part names in the manual and the thingverse files.

The guides on ifixit are good for assembly and I appreciate the time and work that has gone into the.

I think the following would have helped me get started faster, and could help others.

  • a CAD exploded view with labels
  • something about installing the brush ( or maybe this becomes clear when one has all the parts in hand)
  • one page checklist or table for the printed parts ... CSV to markdown is actually pretty trivial. I put what was in my spreadsheet at the end of this post

Happy to contribute as I go through printing and assebling. New part photos for I fixit?
Ifixit.com is a bit limited compared to a wiki and not the best place for discussion but does offer a better possibility to edit in a collaborative manner than thingiverse.

Printed parts

[x] Part Material options Filename Notes
[ ] X-carriage PETG or PLA
[ ] X-Clamp PETG or PLA
[ ]  Umbilical connector PETG TAIL or FIN
[ ]  E-Cage B PETG
[ ]  E-Cage F PETG
[ ]  E-Cage S PETG , HTPLA , ABS , PC High Temperature filament preferred
[ ]  Fan Shroud PETG or PLA
[ ]  Key PETG
[ ]  Inlet PLA Std or PTFE
[ ]  Spider PETG
[ ]  Wheel PETG
[ ]  Tread PLA or PETG 56 or 54 teeth
[ ]  Pinion PETG , HTPLA , PC optional if you already have a GT2 gear for the for the stepper
[ ]  P-Rack PETG maybe HTPLA
[ ]  Idler PETG
[ ]  Tensioner Knob any
[ ]  Pinda Knob any
[ ]  Spacer PETG or PLA
[ ]  Cartridge PLA Direct or Indirect Dark Material
[ ]  Direct Cover PETG or PLA Direct: Std , Bowden , M6
[ ]  Indirect Cover PETG or PLA Indirect: Std, Bowden, M6
[ ]  Spring PETG or PLA Indirect SP_spring… Indirect filament sensor shaft
[ ]  Omega PETG Standard or Volcano SP_omega… See printing instructions

Well... a pleasant surprise to find more info in the PrusaOwners wiki. While it is not as up to date as the ifixit manual. I think it is good supplementary information. Of course this has been mentioned in comments earlier but it is valuable and informative enough to be worth mentioning in the "thing details".


Every time I think I am getting OK with CAD, some hot design like this pops up and I feel inadequate! Ha! Looking forward to the MK3s version! For fun, I am trying to match all the colors in your render, because it just plain looks cool ;)

Hi, I made a Prusa Mk2 conversion for this. I don't have all the parts to build the extruder yet so I have not tried printing with this, but the conversion should work as it is right now (Consider it a beta version). If anyone wants to try their luck, it's at https://www.thingiverse.com/thing:3443561

Skelestruder for Prusa i3 mk2

There are three changes needed that I know. I created an mk2 mod but no testers to try it. Need trigger for endstop (I used Xclamp), move PINDA closer to X for mk42 bed probe points, and new tail that fits in narrower Z frame. Someone did a Fin mod that works, similar to MK2 thin tail. I think there are a few ways to attack it. Let me know how yours goes.

One issue I have noticed so far is that the omega cooler shroud is actually too efficient. If you print small objects at the center of the heatbed (the bed thermistor is right in the conter, on the other side of the PCB, the cooler cools the bed so much it triggers a bed temp runaway error. This is with the old omega version, not the R1.

My parts arrived this morning, so now I finally have the thing built in front of me. The x endstop trigger and the off center cable holder work fine. The pinda holder works too (the probe is at the right place for the heatbed calibration points) but has clearance issues - the probe is so close to the trapezoid slide that the adjustment knob is stuck. Would you mind sharing the files for the Skelestruder in an editable format? To fix the Pinda issue I need to move some bits in the E Cage B file and Fusion is not good with STLs.

I did some test fits (can't print with the skelestruder yet, Aliexpress bearings take very long to arrive) and the only part that still might be making trouble is the tail fin. I think I can make the mk3 or mk2.5 fin fit. The other parts should be ready to go, the x ends top lever is in the right place and the pinda holder appears to be ok too. I copied the sensor position from firmware and confirmed it with test fits and as far as I can measure it is in the right place.

Comments deleted.

Is it possible to do a manual toolchange version?
Ideia: Using e3d beta heatsink so is possible to unscrew without unmount parts. It will require manual solder and connectors to allow detach thermistor and heater from the connector.
Detailed steps:

  1. Unload fillament
  2. Detach thermistor and heater connectors eg: XT30
  3. Unscrew heatsink (At this point we have hole extruder on hand)
  4. Swap with another and do reverse steps.
  5. Ajust PINDA for new setup using the amazing knob

That would allow easy toolchange and combinations, Change nozzle frequently is a bit of pain and fell something bad can happen

My beta is on the way, with just such an idea. Unscrewing it seems a good idea.

Prusa just did an Extruder upgrade MK3s. The manual came out today. You might want to check it out. I think it maybe in response to issues with the MMU 2.0.

There are parts and small details which don't print with 0.6mm. Missing details don't seens critical (i think) But two parts are not printable, see image. Would be great that Omega can be printed with 0.6mm nozzle, it need thicker walls. The shroud also incomplete, don't know if will affect anything.
The small details are like tops of inlet and others, think is ok...

I design all parts to be printed with 0.4 nozzle (most common). You should review very closely that the slice is clean. There are many thin sections that may not resolve, though may not be critical as you say. I have a thicker omega someone asked for but may not be able to post until I get back.

I wasn't able to find it on comments, but no problem i will wait, i also have parts on order.
One thing that wasn't clear for me: I can use a metal GT2 16T pulley or i have to print one from your design? At one section it say to print one, but in another says i can use a metal one if already have one, i'm not sure.

Either will work. I try to provide options to minimize required hw.

I am going to start printing out parts this weekend. I like several other of the upgrades you have done also. I am going to try to print a 5mm shaft for the wheel. I am going to print it in either PETG or PolyCarb. Will print out a prototype to check the sizing to your wheel. Thanks for all the hard work you have done.

Here in Germany a solid Carbon fiber shaft costs nearly the same as a steel one, there are also hollow ones. I think I will try the carbon one as a light weight alternative. But don't know yet when I have time for the build.

Misumi is the only pre-fab shaft source I’ve found. Otherwise need to make your own. I thought about a printed shaft but it will limit your print speeds (if it works at all) and I’m guessing retractions are out. Curious to hear how it goes. I’d order a steel shaft for backup.

I also couldn't find ready to use shafs at aliexpress, so i buy 5x150mm to cut them at size

Nice find plus cut to lenght! Thank you.
What is better bearing steel vs stainless steel?
Also do you have a link for aliexpress Cleaning brush wheel?

Note that I bought some 5mm rod that didn’t work because it was not milled to enough precision to fit bearings. I’ve also seen ones not round. Let me know how these work out. Because it is dowel it might.

Ok i can confirm that shafts are OK, precise 5x45mm. I order the 50mm option but leave a message for 45mm if possible, seller sent 45mm for my surprise. Thank you

Comments deleted.

Which is the better choise to print at limits with volcano (+/- 30mm/s^3 max):
1) 17HS08-1004S
2) 17HS4223
3) Stock

I think any of those should work. I would avoid a 0.9 stepper if you are going for speed. I use the hanpose and it worked fine with my volcano. I hit > 40mm^3/s reliably (30 is not the limit).

Ok good to know. Current with stock i was never able to go more than 30mm/s^3 without have artifacts or underextrusion (clicking) and got same values with Ender3 with Bondtech BMG geared direct extruder. 0.6mm nozzle user

That's odd. You should have gotten higher with the stock extruder. I used a 0.8 nozzle and 40W heater than came from Prusa. But I don't think that is needed.

Yes, but some users are also reporting same values as me under Prusa Forum. Maybe you use the copper block and nozzle plus other better parts? With 0.6mm nozzle i can print 1.2mm perimeters @ 0.45mm layer height which is good to break the limits :)

It works great.
It's the best extruder I've ever had.
Is there a chance to fix it to the ender?
2020 mount.

Glad you like it. Post a make.

TPU, speed 50 mm/s external walls. Internal 60 mm/s


Yeeeeeesssss!!! :-)
I have an ender 5 but it will fit.

This looks great! I'm ordering parts now to build one of these. The only thing better would be a design that would accept the dual Extruder adapter such as this. Any reason this wouldn't work on the MK2S asa well?


Hey Jitx,

I finished assembling the extruder over the weekend and the new instructions are great. I made a few comments on some steps that you might want to take a look at. One thing I think you should add is about the change in build height and how that affects the initial calibration routine. I spend 2 hours trying to figure out why I could not get the calibration to find the points before I found a comment buried here about the build height increase. If you could add a section on how to deal with that it would be great!

Good feedback. I will add something for that. Which solution did you use to calibrate?

I haven't done anything yet. Didn't find the comments until I was at work this morning so I am going to try to get things going this evening. What do you recommend, I have never modified the firmware but that doesn't scare me if you think that is the best way.

If they haven’t changed the routine in FW, here is the easiest hack. Run Zcal and after homing at the top, you acknowledge, and about half way back down hit reset. Then just run a mesh bed level under calibration menu. That’s all it was going to do next. But it errors out because it didn’t reach the bed when it expected to. The mesh level is just confirmation you are within tolerance. The actual “z calibration” is the mechanical homing at the top. There is nothing stored in eeprom. The two z motors are synchronized with the x gantry being level. Let me know if that unblocks you.

Do I need to run the xyz calibration? That was where I kept getting stuck or can I skip that and go straight to zcal and do as you said?

You shouldn’t need to run that if all you changed is extruder. Pinda is in same position as stock so old data points are still valid. Just Z cal should be enough. I have only run full XYZ when I first built printer, even with all my Taurus upgrades, because the basic frame and bed have not changed.

So while I was messing around trying to get the calibration work I ended up factory resetting the printer. In this case I think I do need to run the xyz calibration to get the skew values saved. But I can't get that to work even lowering the pinda almost to the bed still fails on the second of 4 points. Does your patch file to the firmware adjust the build height or can you point me to where that value is in the firmware so I can try to change it and recompile. Or if you have another suggestion I am all ears.

Build height should have nothing to do with it. I misunderstood what calibration you were doing so was giving you the wrong advice. If you find the first point but not the second, what error are you getting? Can you get picture of where it is when probing second point? Any changes to printer other than extruder?

No other changes. here is a video showing the problem. You can skip past the first point to see the second fail and then the error on the screen. Let me know what you think.


  • Posted with 3D Geeks Thingiverse Browser App

Ok, I see that it is starting the hunt way too high. Surprised it is picking up any signal. Anyway, should be simple hack to get you unstuck without flashing new firmware (but you will want to at some point). Just hold a spacer on top of each x end so that it hits the z tops early. Then it should work as normal. You want the spacers equal size and about 10mm high. Needs to fit behind lead screw. Maybe AA batteries?

I found two nuts which were about 9mm and used those. That has gotten me past the xyz calibration. I was going to try to apply your patch file over the 3.5.1 firmware but before I did I wanted to check that your patch includes the 10mm build height change. If it doesn't could you let me know which line I need to change to include that.

Running through the first layer calibration even though I ran M92 E980 Followed by M500 through octoprint. It seems like the printer is really under extruding so I am going to try to do a e-steps calibration to see if it is way off for some reason.

Thank for the quick help!

patch file should adjust height, yes. Note that you may need to switch to 16 micro steps with that e3d stepper. Others have had trouble with that one and I don't recommend it anymore. You can always put on the original LDO stepper and see if things are working as expected.

To be more clear, you either need to double the steps to 1960 because that stepper is 0.9. that is why you are under extruding. Or halve the micro steps to 16 and use 980. It seems the latter works better for that stepper, but a few people never got it to be reliable.

I can confirm that 980 e-steps with 16 microsteps works better with the e3d slim stepper. But yes, it still isn't strong enough, and I had skipping issues with that motor.

are there any other steppers you would recommend from Amazon. the one listed won't ship for several weeks and while there are plenty of others I am not sure what makes one good or not. trying to find one that has the removable cable so I don't have to redo all my wiring again.

I haven’t heard issues with others. I would lean toward 1.8 since you already get smoothing from gearing. You want around 1/3 torque of stock. Lower internal resistance is nice because it should run cooler. But really most will work ok. I understand about wiring. You could throw on stock and just zip wire outside umbilical while you wait for shipping. It will hold up a couple weeks that way. Be careful that those plugs don’t necessarily use a standard pin out so you may need to swap pins in connector.

This might not be helpful since it's not from Amazon, but PrintedSolid has a 1.8 degree stepper that should work really well (from LDO motors): https://www.printedsolid.com/products/ldo-nema-17-motor-overthemoon-ldo-42sth25-1004ac

Another non-Amazon option is Aliexpress. Only €10 and (for now) the thing runs smooth and cold to the touch. Actually, I bought all the needed hardware on Aliexpress for less than €20. Great deal! Happy to share the links If anyone cares.

Hi jltx, I started to print the B4/5 parts but am now printing the R1 parts, but I have a question / issue with one of the parts, it may be nothing.
SP_EcageS_r1.stl the hole for the e3d v6 measures 22.3 mm dia whereas the actual heatsink measures 22.2 mm dia.
This will allow cooling air to leak onto the hot end from above affecting temperature and possibly the top of the print, although I doubt there is enough velocity of air, but you never know.
Is this correct or should this part also be 22.2 mm not 22.3 mm.


22.3 is correct. It will fit nicely. Don’t worry about small gaps. The fan doesn’t have enough static pressure.

OK Thank you, got to finish printing the second half of a cyclone filter so that I can finish plastering and sanding before I can carry on with the important stuff like the skelestruder :) , but she who must be obeyed has been having a go at me for not doing my house duties!.
Oh well, no more parts for the next 44 hours. :)

The newer e3d v6's at least the special edition ones come with a new collet clip at the top to lock the PTFE tubing in place. With this clip the inlet piece won't quite sit flush. Is it possible to add a little more room to the inner top of that piece to accommodate the clip. Looks like another 1mm raised should provide the clearance needed.

Sure. I’ll take a look. Can you measure how tall the clip is at the highest point?

The clip is 2.1mm high measured from the top of the hotend. If you could make the change at least the the ptfe version that would be great! That is the last part I need to start assembly.

updated. try and let me know. I don't have a collet clip on mine.

Fits perfect now! Thank you for getting that done so quick!

I will do that when I get home this evening. Judging by the gap last night I think 1mm more head room would do it but I will get a more precise measurement tonight.

I asked several months ago, but I am finally getting this put together for real this time. Do you know if anyone has made a remix of this part

that will work with the skelestruder. I tried to adapt it using the step file but could not get it correct. If anyone could whip one up with the above model that would be awesome!

Cable chain upgrade for Prusa MK3

I'm trying to modify this for another build because it's a great design. I'm wondering if you would share the step file or similar of it assembled. Need to see how things interact so I can make the needed modifications.

Hi there,

I have assembled the skelestruder, but I have some problems with the filament-runout sensor parts. It looks like the mechanical_top doesn't 100% fit perfectly. When I screw both screws in, the hole isn't straight over the hole in the mechanical_bottom part. Therefor the filament gets curved and doesn't fits smoothly. This both makes it hard to change filament and let the extruder grab the filament.

This gives me layer shifts.. :-(

Any good solutions for this problem?

Thanks in advance.


Do you mean the F_cover and cartridge? I can see such a problem causing trouble with loading, but not clear how it causes layer shifts. I am not following exactly what you mean. Are you on version R1 or earlier?

Could you explain what you mean by "mechanical_top" and "mechanical_bottom" parts? Do you mean "SP_Dcartridge_r1" and "SP_Fcover_r1"? Do you have the approprate PTFE tube in place above the filament sensor?

Hi there,

I am facing problems with: "Mechanical Filament Detector Cartridge" and "Mechanical Filament Detector Top". When I mount the top with the two screws, then the two holes isn't matchning perfectly. This generates a resistance for the filament to be pulled by the mechanical gearing.

I think I need to somehow move the hole on the top part a few mm to make it fit perfectly.

If the extruder cant pull the filament through perfectly and smoothly then this can generate problems with z-layer shift as I read on this forum:

I have borrowed the picture from the reddit forum, since it was easier than taking my own - but it's the same problem I am facing. However the filament has a good flow from the spoil and it's not easy to change filament, since the two holes in the filament-sensor part doesn't match.

Should I try to draw it or is it understandable?

I didn’t see you post there. Did you get your issue resolved?

Oh, this is for Ender? Please post your comment in https://www.thingiverse.com/thing:3364504. I don't know about Ender but I will try to help. The parts you mention don't exist here so it is confusing.

Ender 3 Skelestruder Remix

2 questions. Noticed you had models for the pulleys. I am concerned about the printed ones. Are there metal ones that i can get. If so can you provide info?

Lastly will this work with a stock MK3 v6 and a Moons Stepper?

Yes, 16T just like on the x and y motors. Very common.

Yes and yes. What made you think it wouldn’t work with v6? I want that to be very clear about that so want to clear up any language in the description.

I'm really interested in building one of these for my MK2.5 if possible.
From my reading of the other comments on here, there are two issues with this.

Firstly the extruder needs to be modified so that it triggers the endstop on the x-axis. There seems to be a mention of a version that triggers the endstop but I can't seem to find it in the thingfiles. Has it been removed?

The second problem, as I understand it, is an issue with the opening in the mk2 frame being too tight for the cable bundle at the back of the extruder.
Is the issue that the frame clearance is too tight from side to side or that the cable bundle hits the top of the frame?
I've seen mention of a possible workaround by tapping the rest of an extra long M3 screw but I'm not exactly clear on how this solves the clearance problem.

I really just want to get a better idea of the problems and see if there is any way to workaround them for the MK2.5 or if I'm just doomed to fail with this upgrade attempt :-)

Sorry for delay. I was getting new assembly instructions up. check them out!

Anyway, 2.5 support has 3 issues.

  1. endstop trigger - solved with X clamp variant (available on demand)
  2. Frame clearance, top and bottom - requires MK2 style tail which is where I ran into a snag. But ultimately solvable I believe.
  3. pinda postion - the MK2 and MK2.5 heat beds have different probe points than MK3 so I need a modified pinda clamp. Theoretically possible, need to research it (or have someone tell me the answer). Offsets are in the FW I believe

If you are bold, I can give you the rope. ;-)

Question: For #3, why do you say the MK2.5 heatbed has a different probe point than the MK3? I thought the MK2.5 used the same exact bed and that the pinda now probes the steel sheet as opposed to predefined points on the heatbed like the MK2/2S? I'm almost sure that the MK2.5 and MK3 use the same exact bed with the same exact probe points, but I could be wrong.

I am pretty close to pulling the trigger on printing this just to give it a try on the MK2.5... Just curious about the heatbed comment.

No, MK2.5 is 12V I believe and for some reason the bed is set up slightly different. But it may be that they are close enough to work since you don't have to be right over the probe point to get a reading. Look at diagrams on GitHub. Also I think the offsets are listed in the FW.

Ah, you are right, there is a voltage difference. I'll post the question on the Prusa FB community group and see what they have to say about the probe points.

Nice job. New instructions look great. They give a much better view of the assembly process.

I can see now from the pictures of the cable routing what the clearance issues would be with the MK2.5.
I guess that as someone else mentioned the best workaround for that issue would probably be to mod the firmware to slightly reduce the build volume of the printer on the x and z axis.

Maybe on Z, but X should not be necessary. Let me know if you are willing to test. Good to have a second printer though in case you need to update a part while building.

Yeah, I do have a second printer, its actually a stock MK3. I'm building up a MK2.5 clone at the moment as a second printer that might be better for printing flexibles which is why I was thinking that the Skelestruder might be a good fit. I'm still waiting on delivery of the MK52 bed though so I cant get measurements from the probe points for the pinda just yet.

Based on the changes required, I'm sort of thinking that I may be better fitting a stock extruder to the MK2.5 and then upgrading the MK3 to the Skelestruder once the second printer is up and running.

Has anyone successfully routed heater wire and thermistor wires as per the assembly guide's suggeston Optionally, route heater & thermistor wires up channel in back of cage B. Slip through hole and route out under shaft, away from idler? Could you share a photo if so? With just the thermistor wires and one heater wire through, it looks dangerously close to the bondtech, and would have to be somehow held back very securely if all 4 wires will fit at all - photos attached. I'm not sure it's such a good idea to have the heater wires (or any wires) so close to rotating metal teeth without protection.

There are two zip tie channels there to help secure the wires. You cannot fit both heater wires?

I have the stock heater wires that came with the printer, each has diameter 2.5mm. The thermistor wires are about 1.1mm diameter each. As best as I can measure in the confined space, there's 4mm space along the shaft axis between the bondtech and ecage, and about 4.5 mm space beneath the shaft. And then there's a chamfer that occupies some of that space. It really seems too tight.

I've assembled my Skelestruder, wiring still needs to be done (I'm upgrading Einsy case to Prusa R3/B7 at the same time). As suggested by @jltx below, I probably did have a slightly over pressed first layer, so I will probably get this build going and then do a full reprint. Assembly will be much easier second time around, and notably I'll have an informed opinion about cable routing. Some further observations:

  1. I could not get SP_Ecage_S_b5 to print successfully in ABS despite several attempts, different extrusion temperatures, support/no support. The shroud section would always be very rough and have holes in it, and the roofs of the screw/clip holes would be very messy. I don't normally print in ABS and don't have an enclosure. Has anyone else managed to print this part in ABS? I resorted to using PETG, which prints perfectly without support.

  2. A few screw lengths were different from the docs, even combining the knowledge of the assembly guide and the updated notes in the Thing Details here. I will make full notes as and when I disassemble. Notably, the screw required to hold down the indirect cartridge had to be 14mm - 12 was too short, 16 too long - I had to cut a 16 down as I don't have 14.

  3. Regarding the filament inlet for indirect cartridge: I'm using Capricorn tubing, and it was lose enough that it could pop out on a retraction. I had to wrap a bit of sellotape (scotch tape) around the PTFE so it stays put. Could you add some kind of retention mechanism?

  4. As per Update 11/13, you mention adding a second hole to SP_I_cartridge_b5. There is a new horizontal hole which aligns with a hole in ecage_b (just below the sensor connector), but both holes are M3-clearance sized. Should the hole in SP_I_cartridge be smaller to grip M3?

  5. After attaching spider and motor, I had to take spider off and press in 2 x hex nut for the blower and omega, then reassemble. Suggest others watch out for this as the assembly guide does not (at the time of writing) cover this part of assembly due to significant design changes.
  1. I am going for PC next. I will also try ABS. I have a new version that should be slightly easier to print.

  2. See new Dozuki docs and let me know

  3. Will look at it

  4. the force is lateral to the screw so should not need strong tapping. Mine works. Too loose?

  5. see new Dozuki docs

Re 4: Meshmixer says that hole in i_cartridge has max d=~3.1mm (see measure1.png). Calipers measure 3.03mm from printed part. Measuring a few actual screws, I get measurements from 2.89 to 2.96. Measuring another thread tap hole (in ecage b for noctua, see measure2.png), max d is ~ 3.0mm. Surely the hole in I_cartridge should also be 3.0mm, not 3.1mm?

I'm mid-assembly.

  1. Indirect cartridge: There seems to be some magic in the exact angle of the spring part. Is it supposed to fit in one particular orientation (it has one slightly flat side)? I tested pushing both hard (PETG) and soft (TPU) filament through and fiddled with the spring orientation until there is only gentle resistance, and the bearing still turns nicely.

  2. The P-rack cam seems to be too big in any orientation, and it deflects the p-rack screw enough that it's very hard to get it through the lower brace and screwed into the nut in the p-rack knob. Also, the instructions seem to suggest that the cam is optional. Should I use the cam? How is it intended to work? Should I need to file something down so the p-rack can be assembled without excessive force?

  3. The little tooth that engages with the p-rack knob: I noticed that this appears to be designed with the intention of having a bit of spring. I used a fine blade to ensure that the spring action was free to move. Have others done this?

I would orient the flat toward the filament. But shouldn't matter too much.

Cam is optional. Let me check if it needs a tweak after I updated the rack.

Yes, should spring. I'm guessing you might be over pressing your first layer and getting bad elephant foot. That could explain cam too.

Re: Cam

It looks like there is interference between the P Cam and the side wall of the ECage_B parts, at least when the current .stl files are assembled in 3D land. See attached.

This is a view of SP_cam_b4 and SP_Ecage_B_b5.

Cam is totally optional and I am just going to remove it. No worries.

I am in the process of printing the parts for your Skelestruder, and I have been referencing the wiki on printing and assembling the unit. I have one question (so far). In the breakdown of the parts and instructions for slicing/printing etc, I see on some of them it says to "plate" the part, and on some others it says to "do not plate". I am unfamiliar with these terms in reference to 3d printing, as I am very new to the process. Can you enlighten me to what exactly the term "plate" means in this. Thank you

plate means to print multiple parts at the same time on the build plate. This is convenient and saves time. The print head has to travel between parts on each layer. For small parts, this can be advantageous because it allows more adequate cooling of each layer so you get a better print. For others, this can introduce some surface irregularities and stringing that may be undesirable, for example the gear teeth.

Just finished putting together my new skelestruder over the weekend, I will post pics soon. I was wondering if the SPV1 3.4.0 patch on the instruction page will work with the latest official prusa patch (3.5.1 is the latest)? Right now all I have done is manually changed the e steps from the terminal function in octoprint. I changed the extruder current to a lower value because my motor was very hot to the touch but that only holds until a power cycle.

PS. To anyone else building this extruder (or really anyone with a printer in general) double check the calibration on your e-steps. I ultimately ended with a value of 1014 after using the calibration fixture found here https://www.thingiverse.com/thing:2598993 there is also a useful video explaining the process.

Extruder e-step Calibration Fixture

If you are using the latest prusa firmware (3.7) the commands for changing the current have unfortunately been disabled. If you go the software section of JLTX's instruction page you can find a link that will show what changes you need to make to the firmware in the ardiuno IDE, these changes will adjust your motor current along with a few other tweeks that are specific to the skelestruder.

If you're not comfortable making those changes let me know and I can send you the hex file of the latest prusa firmware that I've applied the skelestruder changes to. I flashed my mk3 with it over the weekend and haven't had any issues.

can you teach me the command to lower down the current? My motor also gets super hot during printing.

Sorry, I replied to my own post above yours so it might not have sent you a notification. Here is the same info again.

If you are using the latest prusa firmware (3.7) the commands for changing the current have unfortunately been disabled. If you go the software section of JLTX's instruction page you can find a link that will show what changes you need to make to the firmware in the ardiuno IDE, these changes will adjust your motor current along with a few other tweeks that are specific to the skelestruder.

If you're not comfortable making those changes let me know and I can send you the hex file of the latest prusa firmware that I've applied the skelestruder changes to. I flashed my mk3 with it over the weekend and haven't had any issues.

I always have to chase something with their changes. The patch will usually fail one or more chunks but it is easy to track down and fix manually. I have a few changes in there that are not strictly necessary. I have a 3.5 patch I will get posted.

Looks like you were underextruding a bit. Did you go with 100 flow or default 95 flow in slic3r?

100% flow. I simply commanded 100mm of filament from octoprint. I had done the same thing with the stock extruder when I got the printer and found that needed to be adjusted as well. With 3 test runs I was only getting about 96.67mm so i adjusted the e-step by about 103%. I think this is a fairly common problem given the tolerances in the bondtech gears and how the extruder motor itself is assembled.

After I posted my first comment I did go through and track done all the changes in the 3.5 firmware and I will probably flash tomorrow since my already running an 8 hour print (quite successfully in an enclosure).

Kudos on this design, I think its a great improvement over the stock. I was having a lot of issues with PLA softening in my enclosure and cause jams in the hotend. Not of hint of that now.

A great design.
I'm finishing him.
I praise the photo 1 :-)
I do not understand two things.
What are they for?
Photos in the attachment.

I like your color scheme. :-)

The first is cage S which goes around the heat sink and completes the whole cage. See 'Update 11/1' notes.
The second is a the cable connector that is more like the stock Prusa version (instead of my fin style) for folks that prefer it old school.

Thank you very much.
Is there a chance for an Ender 3 printer mounting system? ;-)

The extr2 is the tail to attach the cables. Is shorter than the prusa :)


do you have made a Fcover to use with a palette+ ?

What is the connection? Different than mmu?

Your Fcover is thinner than the original of the mk3 stock.
My adapter doesn't work ^^

Prusa MK3 Filament Sensor Cover Adapter for Mosaic Palette
by briankb

Is this a bowden connection? Why not use my P cover?

Yes it's a bowden connction.

I only see the file name Fcover wicj is not compatible because the short tube is less larger than the stock one

That's an alternative solution to the tail. You print in PETG and it flexes with the wire bundle for strain relief without providing to much resistance to X carriage motion like the stock solution.

Also...I think it makes sense to reference the discord group in the “thing details”. Appears there’s a lot of information there which I wasn’t aware of until a few minutes ago.

Can I get a link to that discord?

It’s called “The 602 Wasteland”, under channel mk3-skelestruder. Seems like a lot of pictures and development notes etc are available

I didn't list it for a reason. Please stop posting on this project.

Eh...what's the reason? I'd probably have finished in half the time with that resource (lot's of user photos, troubleshooting, discussion). I'll stop posting, but, I hope my trail helps anyone else who tries this build. You said yourself that my comments would help for when you eventually update the instructions. I think you could warn people that at this point the existing instructions are half a year old and many of the photos etc were removed...

Anyways, thanks for your thing. I think I will have it finished tomorrow.

If you're mad at me for being honest, or having trouble due to lack of documentation and hidden resources, then I don't know what to say. But...again thanks for making it, thanks for helping me. It was an interesting journey, and in the process this thing has become my baby. So I'll stop posting here and on the discord per your request. I'm sorry if I upset you, as it seems I have.


Because the discord channel is a chat stream without organization and is populated by people doing wild experiments that won't help mainstream users, in fact probably frighten them off. So I didn't want to encourage people to go there by advertising it. I'm fine it people know what they are getting into. There are some very helpful people there, along with some very disgruntled ones.

I'm not mad I just want you to stop treating this comment section as your personal chat room. There are over 300 comments now, most are yours. Nobody will bother to read through them all. TLDR. So it not really useful anymore.

Well that ended catastrophically. My copper nozzle cracked in the heat block, and the heat break is stuck to the nozzle. Guess I need to go buy some m6 nuts now

Is it normal for the heatsink to spin? Having trouble with hot tightening. Are my cages too loose or should I remove the heat block and hot tighten outside of the extruder?

I tested generally the load/unload works, and it does extrude. Still can't figure out the wiring


And I have interference on the top too, lost about a half inch and the wire bundle hits the frame. No idea what's wrong...Is the main wire bundle supposed to attach on the sides vs sitting on the top?

If I do look at the prusa extruder tail, it's way lower, about level with the lower rod. Unless I'm missing something, the wires have to be mounted on the sides of the tail.

Edit: Could be the textile sleeve makes the bundle too thick. I was looking at the Prusa's install procedure for the new textile sleeve wiring. It seems to clear if I remove the sleeve, so I could cut the sleeve length a bit and try attaching it later on the tail. Still not sure about the hot end wires. Maybe I'll try removing them and switching the orientation so the wires come out on the right side, maybe that would help prevent them from snagging on the heated bed input box..

Edit 2: I think a combination of the textile sleeve, the e3d motor which has plastic sleeving, and then I sleeved the blower fan. The textile sleeve adds most of the bulk, but in combination it caused a bigger problem. I think if I partially remove the sleeve from the blower fan, and attach the main textile sleeve later it should clear this issue.

Edit 3: lol. I wonder if the fin option would help in this case, could be another option. Have you thought of also adding another option to have a sideways mounting tail? That could free up a couple mm of space.

Hmm, do you have any photos of how you wired/routed your hotend wires? I'm finding that those wires are hitting the bed connector and screw. Seems it will require a bit of thinking


Any tips on how to apply patches? Or...is it actually possible to have the full source firmware folder with the patch already applied? Applying manually seems pretty daunting...but I think editing an already patched file might be easier for us (me) noobs.

I tried downloading git as a first time user, and I ended up using a "patch -p1 -i SPV1_3.4.0.patch" command that at least partially worked. But it throws a lot of errors such as "can't find file to patch at input line 18". I really think having the source files already patched would be leagues easier than having to learn programming...Could just have a basic one to serve as a base, and users could edit if needed.

Not sure what to do since the patch is only somewhat successful...

With the patched firmware source, I think this would have taken like 2 minutes

FYI, if you use "I" instead of "i", it will throw an error, so that's a mistake in the instructions, at least using gitbash/mingw64

Hrm...could be that it's just throwing that error since it's going to the next section. I should have successfully applied the patch in theory, I'll have to check. This section could use a lot more explanation though, and I still think providing a pre-patched source would have been a lot easier.

Edit: I'll test these, and if they're good I can share. There aren't many variations though, not sure why they aren't made available. Thanks

Wiring finished! Up next firmware!

Also attached some photos from the backend showing my choice for wiring routing, tail etc. Non of the makes show the backside where there’s a lot of wires and routes etc so it might help some future makers.

Edit: Deleted photos due to routing/blocking issues, don't want to confuse anyone

Also, a possible option to use stock v6 without height firmware changes:

There are a bunch of recommended springs, but I’m thinking the 12 or 15mm springs (or even 20mm) might offer the possibility of skipping height firmware changes. I’m currently using some much shorter springs, so I can’t test this theory yet, but I’ll order some of the taller springs and test. Might be an option to reduce amount of required firmware changes.

What a coincidence, I ordered a bear upgrade last weekend haha. What’s the new spider for, I guess the way it interacts with the updated jammer required changes? Do you think it makes any sense to try/switch to the beta heatsink?

On an unrelated note, how come your idler pulley is only available in 22t and 24t? Is it too difficult to print at a more standard 20t size? I’ve been eyeing that thing forever haha.

And on another unrelated note, did you consider adding a version of your feet for the bear upgrade? I’m kind of sad I’m going to need to give those up when I switch to bear later. Originally I tried sorbothane on the mk3 but that was horrible due to the way it sticked. I bought 1” round cork pads and love those feet!


No reason to change to beta. New spider is just tweak.

22T actually has same effective diameter as stock smooth pulley.

Oh cool, I didn’t know that. Do you have a bear upgrade? You think it’s possible to use your 22t idler bearing in the bear x axis ends?

Do you know what size screw is needed to screw in the tail/fin? It’s definitely not 25mm. A 40mm doesn’t reach the key. Omg...now I see you put the screw into the tail at the midway point. Or use a 50mm screw. Omg lol, I was so confused and sad I thought I needed to go to Home Depot again hahahaha

25 is correct

You are one sneaky dude lol

On the filament sensor...it can actually attach in both orientations. Which way should the lock clip be facing? I’ll check the Prusa manual...and it seems the clip should face inwards. Could be useful to include in future instruction updates.

Clip is down. That is, on same side as sensor faces

A possible wiring suggestion could be...you could actually route the cooling fan wires through the oval opening in the carriage, if there was a little wire channel in the connecting cage/key. Added a photo for consideration, think it could be a pretty elegant solution.

Also maybe an idea/suggestion, don’t know if it’s possible. I tried sleeving my blower fan cables to make them look a bit prettier, but the increased bulk might impede some airflow in the omega position. Maybe with an omega inlet change, and a blower fan bracket, maybe it’s possible to rotate the blower fan out 45 degrees? It should probably get more cool air that way also I’d imagine.

Or I could just run the wire over the blower fan lol. But I still wonder if you could get colder air with that blower fan rotated

I’m not sure if you plan to support the 2 piece b4 cage in the future, but if you do, could you add a larger hole on the b4 like you did for for the b5 cage, to secure the indirect style cover/lid? There is a tiny hole there, but I’m not sure if I can use it. I tried fitting a small grub screw but it doesn’t fit in the b4 cage design. On the b5’s that hole is much larger.

I had a center drill bit I was using for shaping the pfte tubes, I was able to enlarge that hole twisting the drill bit by hand to accept a M3x10mm. So lid is secure now.

Right. That hole on b4 was just to enable removing the nut. So you would need to drill it out as you did.

How tight is the idler supposed to be? It seems pretty tight on there, even at loose setting. Like really tight. You think it would make any sense to get a 42mm screw? Or is it supposed to be fairly tight?

I did use a nylon though, I think a normal hex nut might give a slightest benefit, but my hex nut was harder to get in. I’ll try switching it out

It is snug but not too tight. If you thread the knob onto screw to the tip of screw is just flush with the top of the nut down inside the knob, that is where I set it.

So weird, if I do that it’s super tight. Are you supposed to be able to slip filament in just using pressure? If I spin the wheel it goes in ok, but just using pressure there’s no way it would go in. Is that as intended or something incorrect?

Definitely using the 40mm screw, hmm...

can you message me? You are doing something wrong here.

Sent a pm, but can’t attach photos. Looks ok I think, not sure though. Does the b5 idler work on the b4 cage?

I fit the teeth a bit better, loosened the long idler screw, seems I can push filament with pressure but it does appear to require a decent amount of pressure. On the Prusa extruder the idler I found was still way looser

This is expected due to the increased mechanical leverage your are fighting (extra torque) which works in your favor when the motor is doing the work. Filament is not pushed into the gears, it is pulled.

Hex nut doesn’t seem to help. It’s still really tight, is that normal?

What do you mean hex nut not helping? There is one in the knob, yes.

I had a nylon hex nut in there, so I switched to a non-nylon as in the instructions, since they are a bit thinner. Didn’t help though. I think either I need a longer screw maybe 42mm, or cut the spring. Not sure, based on how the Prusa one worked it seems too tight at loosest setting.

Also, I noticed my indirect pfte cover has some wobble, not 100% secured. Is that normal? Pretty sure I printed the latest b5 version.

I see a screw hole under the filament sensor, but there’s no way to use that since the screw head would hit the filament sensor. Am I missing something?

There is a hole in the back left side of the Ecage that you screw in an M3x10mm to help secure it. In addition to the one in the front right top.

Oh I see, I was thinking that was for a square nut so a bit confusing. Makes sense now

It will accept a square nut when using the normal cover and you screw down from the top. With the indirect setup, that screw is covered so have to come in from the side

Btw, mine cracked a bit when I tried inserting a square nut. Luckily I just bought some super glue last week so no foul but maybe tolerances could be increased/decreased or general reinforcement. I had a vertical crack from top to that little nut window. I removed the nit to use a screw but just wanted to make a note

For the tail, do you reccomend attaching that with the opening up or down? Looks like it can be attached either way..

Welp...I’ve successfully assembled something lol.

Wiring, belts, and first time firmware modification, all that’s left to join the club!

looks good so far. I admire your fortitude. Don't forget to double the steps (1960) for the 0.9 motor. Looks like you are planning a reverse bowden or mmu? For initial testing, I'd just feed the filament straight into that coupler like it was stock so you don't have too many variables to check.

Also...I see I left the collar unused. That’s supposed to go between the motor and the plastic pinion to give correct belt spacing I’m guessing?

no, that's for the cleaning brush which is optional

One more lol.

Attached a photo, is the heater block in the correct orientation? Looks like the wires kind of touch the nozzle a bit. Seems lee did it with the block the other way, not sure which way is the better way? Is my orientation ok?

Also, it looks like with the omega on the standard v6, you need to use a fan bracket? I’m guessing the thicker part goes down and attaches to the fan since it has a nut space, and the thinner piece attaches to spider?

Not sure, it seems to fit better with the thin part on the wheel, but then there’s no nut for the fan. The other way it seems to hit the wheel though. Maybe I printed the wrong version or something? Or does the bracket attach from the top of the fan?

Or I’m seeing now, in Lee's make, the motor is oriented 180. Maybe the connector is causing some obstruction, did I mount the motor incorrectly? Hm..

heater block looks fine. Yes turn your motor 90 or 180 deg so the connector points up and not interfere with fan. Then it will fit fine and no bracket is needed.

Actually the motor doesn’t interfere, I totally missed the clip-in part. If there’s no obstruction, can I keep the motor position as is? What’s the best orientation for wiring?

Added a photo of the omega nozzle clipping into b5 fan should. Did not see that at all. It fits perfect now. Do you think it’s ok to keep the motor position or is there a better orientation for wiring?

OK, I see your connector is far enough forward to clear. Really personal preference and how you want to wire things up. I have mine pointed up to the left and send wires over top left shoulder of carriage. There are other ways that work too.

Oooh, ok I see. I didn’t notice the omega had a little piece that fits into the b5 fan shroud. If I do that, it appears it will fit! Is my heatblock and motor in the correct orientation? I need to take it apart to fit the collar, I’m guessing that goes between the motor and the plastic pinion. Can you confirm?

Think I might be done for the day, can’t seem to find a 20mm screw, might need to go to the hardware store.

Couple more questions, originally you had a metal pinion on the motor that you replaced with a plastic piece. I noticed while it is on their fairly tight, I can shift it as I did to position the belt alignment a little better. Is it possible to post a link to purchase the original metal one? I’m not sure, but I think if it did have some microscopic shift during printing, couldn’t that cause some issues or problems? I think if the original metal one has a grub screw, it could get rid of this possible variable. I suppose tightening the belt could make it shift less, but since mine is all pla at the moment, I don’t want to risk cracking anything this far along.

Also, I noticed a trend of people printing structural parts in nylon/cf nylon. Where does nylon fit into the chart of reccomended plastics, do you think nylon would be better than pla or petg? I was thinking I might buy a roll for some other upgrades so just curious.

Also for the pfte cover, I needed to use a 14mm screw instead of the 10mm listed for the original cover, in case anyone runs into the cover not securing properly.

Looking good so far, thanks!

I don’t think you need 20mm screw anymore with new spider / fan setup. I might be forgetting one.

Pinion can float. Belt will keep it aligned. No worries. But you can buy metal. It’s a std 16T pulley.

Nylon has some nice properties but extremely hard to print and would not be good for some parts. I don’t recommend.

Oh cool, I’ll try a 16t pulley. In instructions says for the other gear ratio only. What’s the disadvantage of nylon assuming you could get it to print okay?

Oh weird, the place I thought needed a 20mm screw actually only needed a 10mm. Wonder why lol...

On a side note I forgot to put the nut in the spider for the blower...but I managed to get it in without removing anything. Rotated the wheel and stuck the nut in, feel so proud of myself lol. So happy I didn’t need go remove all that hahaha

Another question, on the Misumi shaft, there’s a long part then flat, and from the other end, a shorter part to flat. It’s not symmetrical. Do I insert, I’m thinking, the shorter part leading into the cage? Instructions don’t say anything about the shaft orientation..

longer part first

Nice thanks! I might be able to finish today!

Another question, on the idler/bondtech idler, is the bondtech idler supposed to be hollow inside, or should those little black pieces be inside the bondtech idler? I attached a photo of the part in question. Thanks!

Based on the mk3 assembly, I’m guessing yes. Didn’t notice them fall out

Comments deleted.

Also for this part:


Insert inlet and heatsink, aligning key, into E-Cage-B. Press E-Cage-F on top, first aligning top pin, then making sure inlet fits slot. Rotate heatsink so that heat block is straight. Also make sure Bowden aligned with inlet if tapered

Insert screws into middle

2x M3x25mm

I think at least one of those needs to be 30mm, cause 25mm doesn’t appear to reach the nut...

Both are 25mm. Did you remove support ring on left side?

Er...I don’t see anything that looks removable. I attached a photo, is it supposed to look like that with the screw head not being flush? The b4 cages have a removable part?

btw your prints look very clean. good job.

If you look real close at the stl, that part where your "not so flush" screw head is, actually comes off. it's only there as a support for printing. You'll see when you check out the STL up close. Freaked me out when I snapped it off.

Thanks guys. Interesting, it seems on my prints this part kind of fused together in my petg/pla prints. I had some failures that I tried to yank a piece of this off, but the whole ring/column comes off rather than just the top...

I photoshopped the stl and highlighted what I think you guys are talking about, can either of you confirm? If that's the part, I think I will need to cut it out with an exacto knife. I need to look at it again, but it definitely seems fused together pretty well.

Thanks, feel free to use the image if it helps anyone else in the future.

The red part stays put. The conic cylinder below it should snap off. If it fused tightly, that could need coaxing with knife or wire cutters.

Oh I see, was looking at the model from the wrong angle. I can barely make out the line, but I think I could cut it out with an exacto. I tried on some warped petg parts, but it won’t snap cleanly, the whole part wants to come off. So yea think I need to cut it out. Thanks

Ok I got it! I learned cutting pla with an exacto isn’t exactly fun, after a while, it suddenly popped off and while I can’t say I was left completely unscathed, it came off clean and I thank God for protecting my precious, sole remaining cage lol.

For the tensioner knob nut, I found threading the nut with a screw and inserting from the top, and hammering it in while holding the sides of the knob worked to get it seated properly.

I attached a photos of the tensioner knob nut fully inserted, and a photo of the cage with support ring removed and screwed in.

Correct. If you just insert the tip of your pliers and tilt, it should snap right off cleanly.

Oh, didn’t see that mentioned in instructions. I’ll have to take a look. Any photo of what the support ring looks like or what that finished assembly is supposed to look like? I used a 30mm screw that sticks out atm, need to check.

Btw, any tips for getting the nut into the tensioner knob? Is it supposed to go in really deep, or just slightly? Hard to tell, looks like the nut could be seated really deep, or is it supposed to sit just at the surface?

Nut goes in deep, all the way to the bottom. I'm not sure what you mean by right at the surface. I don't see how that's possible.

Hmm...I warped one of the columns on the b5 cages trying to screw the two cages together. Took a moment to figure that out. I think in the instructions it should be highlighted as important to assemble the cage off the ciarriage, like lees1183 did. It’s much easier to see what’s going on. I really wonder if the 3 piece cage is an improvement or not, it was not very easy to put together lol. If you reccomend petg anyway, the two piece cage might be better imo.

I guess I’ll use my pla b4 cage to finish, and print petg spares first thing when I get the extruder online.

What’s the function of the cam? Tab down, but what about that little pointy indent, I’m guessing that needs to be placed some particular way to function?

Minor amendment to instructions. Don’t screw the key to carriage, p-rack knob blocks it. Should attach key to cage, then key to carriage.

You can do it that way, but knob should not hit. I’ll check that.

Also, if you use the cam, I think maybe it’s better to use a longer than 25mm screw. With the cam, I’m not sure the 25mm screw would hit the top post, it might not extend past the knob. Not 100% sure, but certainly looks this way

Suggestion: the bevel jig for the pfte tube, could be same height as the piece that’s goes over it. Have the pfte tube inserted, a little scared to pull it out since I might damage the tube. But the bevel jig is way taller to use over the heatsink. I’ll try a spare piece I have and use that as the bevel jig. Maybe could have 2 versions of that jig piece

Also, do you have any photo for this step?
“Screw key onto front of carriage using hole below top bearings”

Seems the orientation is kind of important, not sure where it goes exactly..

The tapered end goes up. Someone offered to take new assembly photos but I haven’t seen them yet. Make sure you test fit all before mounting so you know how it goes together.

Figured it out, I was looking on the wrong side. Attached some photos if you want to use..

Happy Thanksgiving by the way!

When you say “If using new ptfe tubing, make sure bottom is both tapered and chamfered to match heatbreak”, do you mean the bottom should be both tapered and chamfered or that the bottom is chamfered and the top should be tapered? Isn’t it normally one end that’s chamfered and the other tapered?

I did both on the bottom. I think the Prusa one is that way too. But minimally you need to chamfer the bottom or it won’t seat in the heat break properly. The taper I did was very light, just to take the edge off. Look at your stock one and let me know. I’m away from home so can’t check.

I got through assembling and hooking up the extruder, however, my calibration is failing. For whatever reason, when it tries to find the first calibration point during the XYZ calibration, it decides that Z=0 is a good 1-1.5cm above the bed. Since the pinda can't detect the calibration point from that height, it fails calibration. Any clue what the issue might be?

Oh. Apparently the prusa firmware assumes the height of the extruder when it searches for the calibration points. This part of the test requires dropping the pinda well below the nozzle, and then needs to be set back up to its normal position for tests after.

Sorry for delay. If you are using v6 (stock) heater, you gain an extra 10mm of height. Bonus! But you need to adjust the firmware (Z height) or trick it during Z calibration. Set pinda lower is one option, or better, use a 10mm spacer at the top so it hits early (only during calibration, remove after).

Thanks. Yeah, I'm seeing that I will probably need to make a couple firmware edits to make things easier for me. At the moment I'm on first layer calibration. I'm using a thingiverse test print since I'm sure the e steps will be way off for the built in calibration.

Minimally you need M92 E980 to account for the gear ratio
I have not had to adjust any from there.

Thx EmilyLyn1 for the firmware mod! Works like a charm :-). You don't accidentally have the modded version of the latest release?

Yeah, sorry. My printer has been down for a bit (I suspect the issue is that the bed is way out of level. It's not just printing the first layer bad enough that prints are failing, but it actually failed on mesh bed leveling once). Since I've been too busy with birthdays to fix my printer, I haven't gotten around to updating its firmware yet. I'll get the next firmware version up soon.

I just recently did the nyloc nut leveling mod and it went really well for me. FYI.

I did that mod, but I think my nylocs might be slightly too lose, giving the screws too much play to move around, as my bed was pretty well leveled, but it's fallen out of being leveled to the point where it's got something like a 0.5mm range from the lowest to highest point now. It's really bad. I should just be able to tighten them up slightly and get the bed releveled, but it just takes time I haven't had the last few weekends.

The never ending struggle ;-). I'm pretty sure we ow it ourselves, but hey what would be the fun of something that always works and can't be improved... right?! I'll keep an eye on this thread. Keep up the good work and the best of luck with your bed issue!!!

I added a link to my 3.5.1 patch on dozuki instructions.

Great, thanks for that jltx. Now I'm really sorry to ask, but what exactly do I need to do with this .patch file? My software skills aren't all that great you see ;-). EmilyLyn1 linked to a .hex file, and that I knew how to use :-). Providing .hex files might be a great addition to your otherwise very clear and usable instructions for people like me.

Well, you don’t strictly need to do anything with it. You can make gcode changes instead . See dozuki instructions. But firmware allows more customization. There are instructions on github for compiling the firmware. Look up Prusa Firmware. Once you do it you should see it is easy.

Ok, I'll give it a try!

Or I could adjust the e steps in the firmware itself. And the load/unload filament. And the 220mm vs 210mm thing.

And yeah, I currently have the M92 E980 in the gcode I've tested with, but loading filament is basically impossible if it's not basically already loaded by the end of the fast load phase as it pulls it in stupidly slow..

Certainly. I thought you were trying to avoid fw edits. But yes, change e steps and load / unload. My patch addresses those. I need to publish update for 3.5 but you can review the changes which are minimal.

I wanted to avoid it, but then I decided I had to, and it's really not as bad as it seemed. So much is just configured.

Glad you took the leap. It is very easy. I have trouble convincing people of that. And it’s not strictly needed unless you do the indirect mod.

Editing firmware itself is very easy. I found this tutorial that explains it better than anything else:

Hard to describe this question but...when you use a reverse bowden fitting with the heat set nut, obviously you put a pfte input from the top, but the bottom of the heat set nut that goes into the 3D printed part, it has exposed threads when looking at it from below. Is it recommended to cut and fit a mini pfte tube to cover up that little threaded area of the heat set nut, or do you just leave it bare?

Just leave it. This diameter is much larger than filament and won’t catch in that short span. I am using that way without issue

Took a photo if that helps, showing the bottom of the heat set nut. Should I leave it bare/as is, or cut and fit a little pfte insert here?

For the fin option, do you have any photos of how it’s installed? Nothing in instructions or user photos. As far as printing goes, the fin might be the hardest part to print. The top layer over the circles is so thin it’s very difficult to get the top/bridge layer completely solid. Could it be designed a touch thicker on the bridge part?


That part is not supposed to bridge or connect. It creates openings to insert the wires.

I’m printing some of the new b5 parts, is the fan bracket no longer required?

correct. attaches directly to new spider

regarding the I-lid_b4, it seems to have only one screw hole that can be used to attach it to the rest of the body. The other screw hole is within the filament sensor enclosure and that surface is also at an angle. It seems that the cover isn't securely attachable because I can only attach one side. Am I missing something?

That is correct and I have it fixed in b5 which is imminent. The mmu Bowden puts some torque on the top so needs two anchor points.

Cool. I'll be on the lookout for b5. The pre_b5 stuff seem to be fine. Once I got my petg dialed in, it all fits well and is sturdy.

The right vent of the omegav when looking at the printer from the front inevitably warps, but not enough to cause issues I think? I printed it in petg but I'm guessing it is a tad close to the heatblock. I have a sock on mine as well.

uploaded b5.

are you using b5e omegaV? I raised that off the bed more and thickened the ribs. Are you in enclosure?

I was using b5d omegaV. Yeah I'll try out b5e and let you know. Are you using the volcano? The volcano heat block is not symmetric along the nozzle so that is why one side of the shroud gets warped eventually. The picture is of my second extruder. I've rotated the heatblock so that it aligns with the shroud. I would hope that doesn't cause issues.

I am not using an enclosure.

Correct, it’s asymmetric and I have the airflow axis angled to clear the pinda. I use a 20mm circle centered on the actual nozzle point to set spacing. So one tip gets close. Been holding up for me on my volcano. I plan to reprint in htpla and let the tips self-anneal before going to high temp prints.

A few more questions...

  1. What's the difference between the Spider_b4 and the Spider_b5? I guess one of them is for the stock v6 and one is for the volcano, but which do I need for the stock v6? I'll print both, I'm guessing the b4 is the one I need, but I think the files names should be more descriptive?

  2. There are two different shrouds, "SP_shroud_b3" and "SP_shroud_b45b". Do you print one of each, or is one for volcano? Kind confused since in the instructions it only shows 1 mounted. I'm guessing one of each and they connect together. Can you confirm?

  3. The files designated "pre", those are for future versions/pre-release revisions? If so could be nice to have them in a separate sub folder to be a little less confusing...


Oh I see, I guess one spider and one shroud are for stock, and another one of those is for the newer Omega style cooler? It might make sense to label the omega versions more descriptive, like "SP_spider_XX_omega" and "SP_shroud_XX_omega". Assuming I'm correct on this...

Originally I thought the Omega style cooler was only for the Volcano for some reason. I'm gonna try printing it!

Got my stock extruder back together lol. Since I got the recommended .25 nozzle I might try printing some spares at 10 microns for fun, also trying the pretty pla/petg profiles which gives a tighter surface texture. The .25 nozzle is a lot of fun!

I do have another question in the meantime. In regards to the Skelestruder gearing, what are the advantages/disadvantages of running the 4:1 setup? And if I was running a 0.9 degree motor, which gearing setup would work better or give more torque?


Noooooooo! I took apart my extruder only to find I missed printing one piece!

What’s the difference between the SP_PI_Lid and SP_I_Lid?

If I could make a user/noob friendly suggestion, as you add more parts to the thing files, it would be great to have some sub-folders, for example “V6 files”, “Volcano” etc. A lot of the files aren’t well documented, some aren’t listed in the instructions. Anyway I feel dumb lol.

Would be awesome if instructions were updated too. Looks like some assembly photos have been removed without any replacement photos.

Another suggestion, for wiring/tidiness, it would be awesome if there were some snap/push in cable clips built in, wouldn’t add much weight. Not sure if this suggestion is applicable since I haven’t gotten there, but could help with wiring.

Wish me luck lol. Thanks!

The P stands for pneumatic which are those PTFE connectors as used on MMU2. You can use either one. I have PI lid installed but feed it directly when I want to bypass my MMU.

Sorry about the instructions. I need to do a pass to correct those. They got moved from a wiki to GitHub with markdown which is less flexible.

Also, for the small pfte tube that goes in the indirect/direct inlet, I guess cut to stock Prusa pfte inlet size?

You can either just trim the top of the PTFE that Prusa supplied (it sticks out too far since the filament path is shorter now). Or install new tubing but you need to be careful to chamfer it properly in the heat break. If you are not sure how, don't change it.

Got my belts today!

I have a couple questions before I start disassembling/assembling...

  1. You say to mark the bearings and “Align bearings with bottom facing out, top two facing out but skew slightly (see picture)”. There’s no picture though. Should the bearing alignment be “Y” shaped from a cross section view? The Prusa y axis assembly requests to align bearings with this “Y” orientation, I guess you mean the same?

  2. The pfte jig file you provide, are you supposed to screw the two pieces together? It looks like there’s a hole to screw in both pieces. The top piece I guess you take an exacto and shape the angled edge, and the bottom piece you put in a blade and twist to get I imagine a beveled edge? If you join both pieces, do you get the correct height I suppose?


  1. You can do it however you want. I suggest have the bearings skewed for smoother travel (maybe)

  2. It's hard to explain the jig, I need pictures. You can skip it as it is just a guide.

The new S cage is difficult to print because of the overhang that starts from layer 2 when printing at 0.15 or 0.2 layer height. It doesn't adhere anywhere.

Are you talking about the Noctua inlet? There are two bridge supports to address that. Or a different place?

Yes. At 0.15mm, slicer didn't put any anchors on the first layer for the two bridges. But using 0.2mm height they were there. I was able to print at 0.2 just fine. Thanks.

Odd. I willl fix that. It should also print at 0.15. Thanks for the feedback

Apologies for this but the issue was because I had elephant foot compensation set to 0.2mm which caused this behavior where some parts of the second layer was floating.

Ah, makes sense. I generally design parts to already compensate for elephant foot or make them agnostic, or even exploit it in certain cases. So I'd recommend not enabling that in the slicer.

So...I probably should have printed that tolerance test first. Today I printed a bunch from .9 extrusion to .99 in .1 increments. What am I looking for? The differences between them are very minor, not sure if it’s a placebo effect even. I chose .95 to test fit two obvious pieces, the clamp and the x-carriage piece, however they still don’t fit. At .95 the tolerance test seemed pretty good, not too tight and not too loose. Should I just keep trying printing the parts at lower extrusion ratios, or are there any other settings that might affect tolerances in this case?

How loose should the extrusion test be? Slide in/out super easily?


Trying to print at .9 to see if the parts fit. I’m not sure they will fit though... In slic3r it says not to go down below .9, is it okay to try say .88? Or does it make sense to try and adjust xy compensation? I guess I could try printing some calibration cubes to see what else is going on in terms of dimensional accuracy in relation to the extrusion multiplier. Hope I’ll have some filament left over to print the extruder parts lol. At some point, does it make sense to just use sandpaper?

I’m using colorfabb semi-matte black pla, it’s a bit more difficult than standard pla in that it seems to behave more like PETG with stringing. Hmm

Think maybe I’ll try this calibration print instead: https://www.thingiverse.com/thing:1622868

It has a calculator that you can measure the dimensions and output corrections, so maybe it would be more helpful. I’ll update this after I give it a shot.

Fast edge over/under extrusion test

Well that’s interesting, based on the calculations it says set it back to 1.0. Ugh...I’ll try printing/measuring again. I’ll try retraction tests, maybe some of the wisps are the problem, even after clean-up.

In what way is the clamp not fitting? It seems you are working way too hard at this. It should be a quick calibration and off to good prints.

Yea you can delete these posts or I can delete them let me know. I was test fitting with the pieces in the wrong orientation, 180 degrees off

I cannot delete comments. So they remain for posterity. :-)

So upon further examination...I was trying to force fit the parts in the wrong orientation LOL. I feel so stupid

Some people are using GT2 16 idler pulleys (for instance for tightening mechanisms). In this case the belt is on the same level in the carriage. Would you be so kind and make a variant of it?

I also wonder if it would make sense to offer the possibility to install 40x40x20 fans. They have the same flow rate but higher pressure. It seems to improve heating capability quite significantly.

It's only 1.1mm. I don't think this will cause any discernible issue with 16T.

40x40x20 won't fit on the side. Would have to be front mounted which is a major redesign.

Would this be suitable? For those of us that want the symmetry.

see new b5 carriage and jammer.

Due to all the requests, I have a solution I’ll be releasing soon.

Comments deleted.

I'm starting to print parts for this. I suppose I print one of each item? I don't mind to print extras of things I don't need and waste a little filament to be sure, but it would be horrible if I took apart my extruder, printed 1 item, but ended up needing 2 of them.

For example, "SP_pinion_b1.stl, SP_pinion_deepflat_b2.stl". I'm not sure if I need only 1 (either or), 1 of each, or 2 of each, etc. Just want to be sure! I suppose some things will make more sense during assembly of course. I'm trying to be careful and printing the different versions just in case. Thanks!

Thank you

For your first build, if this is your only printer, I’d recommend printing 2 of each, or minimally a few specific parts I list in the instructions. You only need 1 pinion but depends on your stepper shaft.

Do you recommend lubricating any of the metal parts pre-assembly with something like Superlube for less friction/smoother motion?

I skimmed the instructions but didn’t see anything about duplicates/quantities for specific parts, but maybe I read too quickly. I’ll print 2 of each to be safe, and I’ll also try pre-assembling everything before printing the duplicates to make sure the fit is good as reccomended in the instructions.


do not lubricate anything.

Comments deleted.

Could you post a picture of the fin (SP_fin_b4.stl) in use, and showing the cable routing?

One last feedback I have is the cam's arm is weak enough where you can easily break it if you aren't careful. Otherwise this extruder design has so far been very nice. Thanks for putting this work out here for us to benefit from!

When using the 4:1 gear for the extruder motor I noticed I cannot tighten the belt as much. Is there a different spider that can compensate for the smaller 4:1 gear?

14T? I will add some more range on the spider. Thanks for the feedback.

It looks like the ecage f17 needs changes too as I'm sure you will find out.

of course. I am about to release update cage anyway

Would you advice waiting for the new cage update, I have everything printed at this point but If the changes to the cage are worthy I would rather wait for you to release them before I attempt to assemble everything.

uploaded for you to try and decide for yourself.

I have full hotends (heatsink,heatbreak,heatblock,nozzle) that I tend to swap out as a whole each with a different sized nozzle. This makes it so I don't have to deal with unscrewing nozzles when I want to change. Will this new cage design make it more difficult to swap out full hotends?

Are you removing the heater cartridge and thermistor? or do you have wire plugs? The idea of the key is you can replace the entire extruder if you make the wires pluggable. Of course that means duplicate feed (bearings, bondtech, axle) too.

But to just swap the hotend, it should be only marginally more because you have to remove the cage S and slip it on the new one, or print multiple copies. Since you have a second hotend, you can print it up and do a test fit on the side and see for yourself. Let me know.

Can cage S be printed in PETG?

No. The heater cartridge and thermistor are using wiring plugs.

Regarding the key, I thought that was the idea but I was not able to remove the extruder entirely off of it because (perhaps mistakenly) I could not find the screws to loosen without disassembling some parts. Could you let me know which screws to loosen so can I simply slip off the extruder?

That is the direction I will go. I hate having to disassemble the thing and would much rather take advantage of the key.

If you route the wires up through the key then you should only need to remove the two M3x10 (one above Noctua, one above pinda) and loosen the M3x40 that holds the idler so the head clears the top slot on the carriage.

Upon actually trying to detach from the key, the volcano heatblock actually gets in the way. I have to remove the heatblock to remove the assembly via the key.

I must not have mine tightened as far on the heatbreak. I can add a relief on the key to make that clear. You can grind off a bit on the bottom of your key in the meantime.

I would still think your heatblock would hit the key when trying to slide it out. The heatblock protrudes enough so that it blocks the key.

I am thinking I could turn the heatblock 180 degrees instead, so that the parts don't need a redesign. Otherwise it seems the keyhole actually needs the adjustment but that might not secure it enough.

Are you on latest key version? I reduced the throw to about 6mm to remove (for this reason). I am in MMU hell right now but will take a look.

I deleted one of my comments because I failed to realize that the heater wires come out through the gap above the key instead of into the body where the axle is. Ugh I'm stupid.

It seems to be a tight fit with the wires going up through the key but should be fine. Also, it may not even be necessary for me to route the wires through it actually.

Comments deleted.

No I am not on the latest key. I will check it out. Good luck on the MMU stuff. I'm noticing a bunch of people having issues with it.

Yes the GT2-2M Timing Pulley - 14 Tooth

Is there a way to be notified when you make an update? Sorry I'm new here.

not sure. maybe if you follow me, but I think that only says if I put out a new thing.