HEPA Filter for 3D Printer

by jmillerfo Jun 21, 2017
Download All Files

Thing Apps Enabled

Please Login to Comment

I was thinking about a low noise setup, so noctua
to get a decent airflow would you lean more towards changing proportions and fit a 120x120 fan or having 2 twin blocks running at the same time?
would this kind of solution be enough?

thanks in advance

So much depends on the efficiency of the fan. Most consumer fans are designed to move air past objects (i.e. heat sinks) where there is little resistance to flow. In our case there is a great deal of resistance to the airflow and therefore you need a fan capable of over coming that resistance (High Back Pressure). Unfortunately a byproduct of moving so much air in a high resistance to flow environment is cavitation resulting in noise. The Noctura fan is very efficient but it does not provide a lot of pressure to move the airflow through the carbon pile.

Whether increasing the size of the fan or placing several smaller fans in a series would provide the best solution is subject to experimentation. If you are up to doing so, let me know your results so that we can share with others.

Comments deleted.

Maybe I'll give it a go then
but about the series variant, I was thinking more about doubling everything, I mean: having 2 completely duplicated and disconnected packages (filters, fans, etc), two instances of the whole thing

actually come to think of it it makes way more sense to have 2 fans in series

I would try two fans in series first and see if you get enough airflow.

Impressed that this all fit on my Up Mini 2 build platform (120x120)!

I did have some problems with printing the mount as the support material ended up stronger than whatever we should call the structure between the to arms. See others have made mentioned or made alternatives for the same reason.

I made a change where the thickness increased a little but more importantly it all has the same cross section viewed from above so less support material. Feel free to take inspiration from it or add it to your files or leave it here.

Thanks for sharing!

Thank you for your input and contribution to the Project. I have added your version of the mount to the list of printable files with attribution.

Can you confirm its just a Roomba 800 series filter and that there isn't anything special about the one linked in the BOM?

The link to the filter @ amazon currently says unavailable and they don't know if/when there will be more. Just searching for Roomba 800 compatible filters it doesn't appear there is any risk of them becoming unavailable, but maybe not that particular manufacturer.

There is nothing special about the Roomba 800/900 filter. As long as it fits one of those sweepers it will fit the HEPA Filter.

I did check the URL for the filter and it comes up with the ones I previously specified but shows as unavailable. I will update the URL to one that shows availability.

The HEPA filter chosen here is only effective to 10 microns. Any thoughts on using a .3 micron filter like this: https://smile.amazon.com/Electrolux-EL012B-Electro-Filter-Package/dp/B000MISO8A/ ?

When I get a chance I will do a little research on that filter. Given the very low cost of the filter I doubt that it has 0.3 micron pores. Micron filters get their effectiveness by randomly overlapping layers of material that has a relative small porosity. 10 microns is typical of a good household filter. That being said, HEPA filters generally filter high 90% of of the particulates one order of magnitude smaller than the porosity. Marketing folks often spin their specifications using this factor and I suspect Electrolux is doing just that.

While marketing hype does abound around the false use of HEPA terminology in the consumer-grade filtration industry, Edsai is making the right call here. The iRobot filter is only rated AT ALL for particles down to 10.0 microns, which inherently mandates that it is not a HEPA grade filter. For a filter to be HEPA graded, it MUST be rated to retain or otherwise filter a minimum of 99.97% of 0.3-micron particle by US legal standard and 85% of such particles by EU legal standard, with the EU specifying eight such classifications ranging from 85% total retention (E10) to seven nines five total retention and six nines local retention (U17). The companies that list 0.3 microns aren't just "spinning" their specifications; they're following the law and international standards.

Again, by legal standards in the US and the EU: filtration-rating specified at 10 microns? Not HEPA.
---------------------------------------------------------------: filtration-rating specified at 0.3 microns with a listed retention percentage? HEPA.

That having been said, with the preponderance of unlitigated[sic] false-advertising abundant in the industry, it wouldn't surprise me if the Electrolux filter is being sold under just that; particularly since the product page claims "Class H12" and "99.97%" when the closest legal classifications are Class E12 at 99.5% total, Class H13 at 99.95% total and 99.75% local, and Class H14 at 99.995% total and 99.975% local. There is no Class H12 according to the sources I've read, and if there was it wouldn't be 99.97%.

I am trying to understand the fan and pressure design for a HEPA Filter. Can you share your understanding about filter designs (things such as flow through filter must be 0.5m/s, carbon has 0.2" mm H2O static and the fan static must be HEPA+carbon+CFM at that point etc.)?? Thanks!

PS: Currently I got a filter made from someone for my MK3S enclosure. But I would really like to learn these basics.

Its been some time since I designed this filter and don't have access to my notes at the time as I am traveling out of the country for the next several weeks. The first step in designing any filter is to determine what the objective(s) are. Do you want modest airflow and quiet operation, maximum filtering in a short period of time (Greater Airflow required), etc.

Beyond that there are several fan designs, some move a lot of air at low static pressure (PC computers are good examples) and some filters present a high resistance to airflow and therefore a fan with high static pressure is required.

This design combines a carbon pile and a HEPA filter and requires a high static pressure to get any reasonable airflow.

There are a lot of good fluid dynamics primers on the web that can guide you toward a design that fulfills your needs. In the end you will likely resort to some level of trial and error varying the fan design since the pile and HEPA filter are a given.

Sorry, my mistake, I was talking about support not raft... Anyway, thanks for the feedback I will try first with Slic3rPE then if I'm not satisfied I will try with Simplify3D.

Hi, I'm a newbie.... how do you achieve to print this Exhaust tube. I'm using SLic3r Prusa Edition but with raft it seems there will be all over the place and I wondering how I will get rid of them.... maybe I should try without, or raft limited only to some part of the piece but how ? Any help would be great....

I use Simplify 3D for my slicer. It is very good at building support on its own and it also allows the user to add or subtract from the automatically generated support. I don't use rafts for this part. There is enough material in contact with the build plate that a raft should not be required. For supports I use 1 dense bottom layer and 2 dense top layers. In between I use 15% fill. This yields a good finished part and easy support removal.

great desgin, i see in the thing to buy list, thad you are showing 3 different types of fans.
is i just one of those fans i need to install?
i cant see properly on the pictures or cant find the anwsers in the discription.

Only one fan is needed. I list a variety of fans so that you can choose the best one for your application. The major difference is Airflow vs Noise. I chose the highest airflow for my application because the printer lives in my workshop. You might want a very quiet one if your printer is located in a quiet space. Below is a copy of the fans in the description:

Fan      Part Number      RPM   CFM  Noise Length
Noctura  NF-A6x25 FLX     3000   17  19Dba 25mm
Sunon    PMD1206PMB3-A    6000   42  47Dba 38mm
San Ace  9G0612G101       11,800 65  58Dba 38mm

Thank you for the quick response, a question about the fans listed,
Is the noctura 3000 strong enough to get the air flowing and filtering? Because the 3000 rpm vs 11.800 rpm is a big difference.
And how is the best way to power the fan? My tornado is a 24v so best I power the fan to a wall power supply in som way. Not my best expertise btw. Installing this filter is a must for me because I have a small kid running around in my house

Its on the low end of acceptable but it is virtually silent. I would be good for PLA or PTEG but not ABS. All of the referenced fans are available in 12Vdc or 24Vdc. Last time I looked Amazon carried all the fans.

Any suggestions for a High Static Pressure, 24v fan that would be appropriate? (I have the Prusa MK3). Would this Sunon fit the bill?

I haven't tested any 24VDC fan yet but the one you selected should work. Let me know what you think of the fan. If it works for you, I will add it to the list.

Thanks much.This is really a superb looking build.

My concern is that the fan is two wire, so I'm not certain we can control the CFM and that's really essential to get the most out of your components, as we'll be able to control temperature in addition to air quality. I noted that someone had PLA in the case fail due to temp, so I'd expect we'd really like to be able to control the temperature from an external mounted rheostat, rather than having to open the case and adjust the baffle.

Am I on the right track here? I'm trying to visualize the unit build and installed.

rluci, did you end up trying the 24vdc fan you linked to and if so did you have any success?

No, I opted for a 12v fan: https://tinyurl.com/y47oktve. I stepped the voltage down from 24v to 12V with this: https://tinyurl.com/yy325azq Then connected a variable speed motor before the fan: https://tinyurl.com/y28gvwv9

In the end, I decided that it's near impossible to get enough pressure through the filter and charcoal to guarantee a purge of the cabinet; that there is no way to tell when the charcoal is deactivated; and that the science of air scrubbing requires a lot more math to determine the amount of charcoal needed for a given CFM than I was willing to do.

So, while I'm quite impressed with jmillerfo's design and idea, I replaced the fan with a larger, 140mm, fan and printed the parts for a 5" window passthru to the exterior. (My printer sits in my workshop.) So, now I just vent directly outdoors. The fan speed controller allows me to balance cabinet temp with exhaust.

Venting the enclosure environment to the outside is certainly the best alternative unless you consider the outside environment. I tested the enclosure printing ABS and found the filter removed all odor produced by extruding ABS. I also used a "Smoke Bomb" we use to detect aircraft pressurization leaks. The aerosol can doesn't actually produce smoke but a fine particulate that is highly visible for inspection. This makes it ideal to test the effectiveness of the filter. I found that a 65 CFM fan that has high back pressure capability is more than enough to completely filter both the ABS odor and the smoke particulates.

In my case I also have an area exhaust fan that removes about 5% of the room air per minute with fresh air as a makeup.

Great news. I'm familiar with the aircraft leak-down tests and hadn't thought to beg a smoke bomb from my friendly FBO. My problem was that I was getting such a hack and headache from PETG printing, I finally decided to just vent it all. Will likely revisit your excellent build in time. I still have the rear window and all the prints.

Thanks for making this. I'm collecting materials now and am wondering if this is correct, as the description does not match the item:

1 set Power Polarized Jack/Plug

That is the correct item although you can use any connector that insures proper polarity is maintained when unplugging/plugging the filter.

Thanks. Any idea if the fan can be powered by the Prusa i3 mk3 power supply?

It can. That is how mine is powered. Just be sure to match voltages. I have a MK2.5 which is 12Vdc. I believe the MK3 is 24Vdc.

Thanks for the design, printing right now!

So about the Acurel EXTREME activated carbon pellets, they are hard to get here, would have to import.. What are the measurements of those pellets? I could get my hands on ⌀3mm to ⌀4mm pellets - just big enough to not make it through the grate?

Thanks in Advance!

Any Activated Carbon pellets should work as long as they are large enough not to pass through the grates.

safety WARNING:
mount failed in high temperature, filter crashed down.

this could be applicable to more:
today in the dutch heatwave temps are 30+ degrees celcius and inside my enclosure 50 degrees while printing ASA. The mount gave way resulting in the filter crashing down and thereby ruining a 4 hour print. Luckily I noticed the failure quickly and beside the failed print, no more damage was done.
The mount I used was printed in PLA. On monday (EDIT: tomorrow) I will do a reprint of the mount in ABS and reinforce it with a full circle velcro supporting the legs.

I have been using the filter with satisfaction for half a year, doing occasional prints. However a redesign of the mount might be in place.

Sorry that you experienced a failure in the mount. I incorrectly specified a requirement to use PTEG or ABS if ambient temperatures exceeded 100°C. The temperature should have been stated as 50°C. This has been corrected.

The Tg (Glass Transition Temperature) of PLA is only 60°C at which point the material will become very malleable and is easily deformed by the force of gravity on the filter as can be seen in your photo. The mount in my printer is made of PTEG. Depending on the type of material I print, the enclosure temperature varies from 35°C to 60°C. The mount has been in place for more that 1 year and is showing no signs of deformation.

kudo's for the prompt action

this increased the airflow quite a bit

This likely because the majority of the airflow is now bypassing the carbon pile.

My original design had airflow holes full diameter and lacked the upper baffle. Your change effectively duplicated that design. The problem was the airflow largely bypassed the carbon pile since it is difficult, if not impossible, to get the space below the HEPA filter completely filled with carbon pellets. Hence the addition of a solid baffle.

One way to test the effectiveness of the carbon pile is to inject inspection smoke into the air-stream and observe how much of that makes its way through the filter. For this test remove the HEPA filter so that it is not removing the test smoke. Amazon has a number of products that generate detection smoke. Search for "leak detection smoke".

As it turns out the carbon pile provides a significant portion of the drag on the airflow. That is why it is necessary to use a fan that can overcome this drag (High Static Pressure Fan). They tend to be a little noisier because of the airflow cavitation around the fan blades.

can you make a solid expanded grate so I can print it with different infills to test airflow?
I am trying what I can to increase the airflow so I can run a more powerful fan at a lower speed and still get alot of flow across the charcol
I am not adept at fusion 360 to do that one, sorry

I don't currently have the spare time to do what you are asking.

Varying the infill will not accomplish what you want. The infill layers depend on upper and lower solid layers to hold the infill pattern in place. If you use a sparse upper and lower layer infill, the grates will not have sufficient strength to contain the carbon pile. The only effective way to affect airflow with a given air pressure is to vary the size and/or number of holes in the grates. The Filter to Motor adapter also has a grate component that would have to be modified.

Sorry, should have been more specific.
If it were solid, absent the holes, I could print it with an infill, and no top and bottom layers, extra perimeters and it would be completely a mesh then
I’ll figure it out, thanks

ANyway to doctor this to take a 80mm or 120mm fan? Maybe go to a double filter?
The 60mm fan noise is annoying and looking for options

I am trying it without exhausting, just recirculating the air. I suppose I could just put a grill over the fan instead of the cone/divertor.

If noise is the main concern and you are only recirculating the filtered air, you might try the Noctura NF-A6x25 FLX fan. It is virtually silent.

Doubling or tripling the fan size would be very difficult with the present design. It would be easier to start over from scratch. Keep in mind that this application requires a fan that is efficient with high back pressure. Ordinary case fans will not work and I have not researched options for suitable larger fan.

In my case I have the fan in a significant enclosure that muffles most of the noise. In fact the motors and fans on m Prusa i3 MK2 printer completely drown out the HEPA Filter noise.

Did you even think to roll up an activated carbon mat and place it in the container? - is not there a bigger surface with less resistance?

I have not tried this but I don't think it would reduce the amount of resistance to airflow assuming the mat was rolled tight enough to keep airflow from avoiding the carbon by passing between the layers. That having been said give it a try and let me know your findings.

If we're filtering the air, wouldn't we want to make the mix higher for the recirculating and lower the escaping air? Am I missing something?

Depends on what you are trying to accomplish. Here are some thoughts:

  • Enough filtered air needs to be exhausted to provide a slightly negative enclosure internal pressure. This keeps the nano particles in the enclosure.
  • For materials such as ABS where bed temperatures are high (100 deg C) more air needs to be exhausted so that cooler outside air can be used to cool the interior of the enclosure.
  • For PLA the temperature is relative low and less cooling air is needed.

I have supplied a restrictor that will divide the air about equally between exhaust and recirculate. Since this is a simple shape you can make your own restrictor to better suit your specific needs.

But for ABS it might be good to have a warm enclosure. Like 40°C to 50°C. Sure - still cooling is needed to protect the hardware. But...

Exactly. That is the reason for the restrictor which can be drilled out as needed to get just the correct flow in order to control the internal temperature.

For the record, the San Ace fan is ribbed so it requires screws the full length of the fan, nozzle and filter to motor, about 60mm. Just ordered some so I will see how well that works.

Also, where does the restrictor go or is it still needed?

also the fan control box wont take a 16mm screw, or a 20, so you may as well make it a bit taller for stuffing the wires into and use a 25mm screw

You are correct. I had cut down M3x40 Stock Crews as required for the enclosure. The resulting length was 22.5mm. Per your suggestion I increased the Control Bottom height by 3mm in order to allow a standard M3x25 screw to secure the cover. The updated file is online.

You are correct on the screw size. I will add a note.

The Restrictor goes on the end of the Exhaust Tube if you need to force more aircraft through the bypass hole in the Nozzle.

Have you thought about a bigger fan diameter like with the noctua nfa14? Specification says about 140m³/h (about 82.4cfm if my calculation is right)

I looked at it but I did not have the physical space that such a large fan would require. Secondly most large fans do not have high static pressure capabilities; i.e. they can move a lot of air if there is little resistant to that movement such as in a case fan. Where there is a lot of resistance to airflow such as in our case, they do not perform well. I have spec'ed 3 different fans that have a range of airflow under pressure vs noise. If the fan you selected is too noisy, move down one option; if there is too little exhaust airflow and you can tolerate the noise, move up one option.

Tank you for the fast reply. Static pressure might be a problem for the Noctua. Very valid point.

Have you tried it without the expanded grate and baffle and just fill the whole chamber with carbon?

Yes and it works to some extent but a lot of air bypasses the carbon pile and is recirculated without the benefit of carbon scrubbing. The HEPA filter does remove the nano particles however.

OK, first I have to say this is very thorough. All sources are there and some alternates
I have printed the whole set and a few comments.
The grommet and restrictor are so thin on the flat that I had a hard time removing them from the sheet without destroying them.
The hanger I think is too weak to hold the whole device up when screwed to the back. The base of it is coming out at .94mm thick and the curved pieces, even at 100% infill, were to weak. Might be the filament or my abilities, but on both Hatchbox PLA and ngen, they snapped with ease.
I like the design and now I am invested with all the parts bought, can you assist in any way?

I printed the Mount in PTEG and do not have any problems with strength. I have increased the Mount base by 1mm. Hopefully that will be sufficient for PLA. You can re-download the Mount file and print a new part.

As for the Grommet, I apparently created this on the fly and so not have the design file any longer. The Grommet is for aesthetics only and you can install the filter without the Grommet. If you would like to create your own, the object is very simple as it is only a tube with a flange. Tinkercad.com is a free and easy tool that you can use for that purpose. All that being said, I looked at the Grommet I printed and the flange is more than a mm thick. That should be sufficient and you should not have any issues separating it from the print bed. I would check your printer settings, in particular the settings for the 1st layer.

I hope this helps.

I have Fusion360 so I am good there.
The new mount looks bulletproof. I will try that out - thanks

So, super stupid question for you. I printed and assembled the new design. However, once powered on, it seems like it is not sucking any air through the chamber. The chamber is filled, per your instruction with activated charcoal under the new plastic piece. Then the air filter on top. When I do a smoke test, no smoke is drawn into the chamber with the air filter on. Remove the filter and some smoke is drawn into the chamber. On the old design, with just the air filter, it sucks like a Dyson... I also checked to make sure I had the fan orientation correct and it's sucking from the chamber and blowing out the back - hell that thing can really blow w/o restrictions, and will eat a paper towel for lunch. So what am I doing wrong?

No Idea. Try without the carbon pellets. The problem can only be:

  1. Blocked Filter,
  2. Too fine of Carbon pellets,
  3. Other obstruction in the flow path. Check both grates to make sure they are not somehow blocking the airflow. I assume that you do not have any restrictor on the output. I just checked my installation and there is plenty of airflow.

Thanks - I checked the grates and even cleaned the holes just in case. I did remove some carbon pellets and that seems to help significantly. Do you happen to know if packing them too tightly will cause a restriction? (I ordered the same as stated in your BOM). Basically, how do you know you have too many in there? :)

I am curious how well these would do as replacements for the carbon pellets - thoughts? https://www.amazon.com/Carbon-Filter-Activated-Weller-Absorber/dp/B009AE8DWO/ref=sr_1_5?ie=UTF8&qid=1517011950&sr=8-5&keywords=activated+carbon+filter

There are mat filters. They may work. You will need to cut them to fit.

Nice design. Can you post a step file of the assembly I can insert into my printer design?

See File Section.

Great, Thank-you.

Jmillerfo - Super stupid question - Once you install the new baffle, seal it up, putting you carbon pellets into the new chamber created by the baffle, then installing the grate - How in the heck do you remove the grate when you need to replace the pellets? See photo, or am I not understanding the instructions correctly.

If you look closely you will see a horizontal slot near the top of the Grate centered across the Grate. This slot is set at an angle that will permit a small screw driver to be inserted. Once inserted, simply apply a small amount of aft lateral pressure. If the Grate was properly printed it should just pop out and can be moved aft to expose the pellet chamber. You should test fit the grate to insure that during the printing the grate did not become oversize. If it did, you can lightly sand the excess material from the grate.

Oh ok. So insert screw driver and pull back away from the fan? Seems to have worked, I will need a smaller screw driver :) Thank you so much for your work on this project and taking the time to answer questions.

What is the best way to print the exhaust tube? On its side? Supports?

Nozzle end down with supports. See attached photo:

Finally got around to printing this up. This is a much better implementation over my original design; much easier to print. That said, there is a small design flaw in my original work (which I added an easy fix), but it doesn't appear to be present in this remix. This design requires a small extension piece to be added to the carbon chamber. This extension keep air from by-passing the carbon at the front most lip. (See Picture - Black Piece)

I tried fitting a baffle of varying lengths to see if the there was a meaningful improvement in carbon filtering and determined for my usage there was not. It seems the size of the pellets allowed some air to bypass the carbon pile no matter the length of the baffle. I thought about omitting a portion of the holes in the upper portion of the plate separating the carbon chamber from the fan but that would have lessened the amount of air that would ultimately pass through the filter assembly. The HEPA filter is taking the majority of the harmful nano-particles out of the the air flow in any event and I cannot detect any noxious odor in the processed air even when printing ABS.

While not remotely scientific, my testing (using an TMA40-4 Anemometer) (with and without the HEPA) showed a reduction of output CFM with the baffle as compared without. My summation of the results at the time was more air was being forced through the carbon pellets.

While I've not tested your revised design, I suspect an even greater amount of bypass. Your design doesn't appear to incorporates, (to help facilitate ease of printing and part reduction) any chambered area between the top grate to guarantee carbon passage. My original design with the extension incorporates a 20.22mm carbon chamber where the air must pass through the carbon before exiting. My thinking at the time was primarily to help control VOC's

Perhaps the pellets I'm using are of a smaller diameter and length and thus cause greater bypass.

I made some similar tests with an airflow meter we use to test aircraft gasper flow. My findings are similar to yours.

I have made some changes to the design to insure all possible airflow is routed through the activated carbon bed in the Chamber. In summary I have added a second flange in the chamber below the flange that the HEPA filter sits on. I have included a Baffle that can be inserted or removed but held securely in place in the slot between the two flanges. The Extended Grate now fits into the Chamber at the aft edge of the Baffle thus capturing all the carbon pellets forward and insuring the area of the Chamber having pellets is completely filled. Finally I removed the top several rows of holes in both the Extended Grate and the Chamber to Fan parts. This will insure that all airflow is forced through the carbon pellet bed even if the area is not completely filled.

Part of this update (not related to the Baffle issue above) is the addition of two new fans, one that is virtually silent and another that has considerable more airflow and higher static pressure. These have been added to address other comments relating to noise and/or airflow.

Do you no longer recommend the original fan?

The original fan is the preferred on IMHO. I had a comment asking for a "Silent" fan. The Noctura fan is one such fan but the airflow is 50% of the original. Another comment expressed concerns that the enclosure internal temperature was rising too much for the Prusa i3 MK2S MMU. The third fan was introduced to double the airflow at the expense of noise levels. I have elected to stay with the original fan for my printers.

All fans support the speed controller so even with the highest flow fan, speed can be adjusted as desired with a commensurate reduction in sound as speed is reduced.

The MK3 power supply should NOT be in the cabinet, but remote to a mount out side the cabinet, like on the cabinet leg. Continued higher temps will only shorten the life of the power supply.

it never hurts to run the power supplies in as cool an environment as possible. That having been said, moving the power supply complicates the installation and future maintenance of the printer by separating the printer from a major component. To determine the environment that the integral power supply lives in, I instrumented the enclosure in the vicinity of the power supply and observed the following:

Enclosure temperature never got hotter that 35°C with a bed temp on 70°C. 70°C is what I use for PLA and PTEG. On ABS I run bed temperature of 100°C. The internal Enclosure temp is <45°C for ABS. This is using the restrictor orifice. If you overboard all the filtered air, the temperature should lower another 5°C. The HEPA filter fan used was San Ace P/N 9G0612G101 moving 65 CFM air out of the enclosure. I concluded the trade-off between running a cooler power supply vs ease of maintenance favored easy of maintenance since 45°C is within the typical operating range (<50°C) an ATX power supply.

I'll not argue with your engineering experience or the amount of thought you've put in on this, jmillerfo. And I agree- my cabinet has not seen more than 45 degrees C even with ABS. Still, the power supply would prefer a cool environment over a warm one, I suspect.

For the MK3, if you look at the Lack Table Enclosure and Power Supply external bracket, it's a pretty easy split to accomplish. I used (and swear by) Anderson Power Pole connectors for a clean buildout.

As a result, you can quick disconnect the printer from power supply for servicing. Or simply remove the cabinet, disconnect the fan cord and the printer and power supply come off the lower table, still connected, one in each hand.

A lot of what we do with these builds is not an issue of right or wrong as opposed to "because I can" ;-)

Thanks! It is a great design. I've been reprinting all the internal bits all night. If folks want temperature control, they could also do what I am doing, have dual fan setups with a temp controller for a second fan.

Fantastic Work!

What kind of power supply did you use? I'm using 12v laptop PSU....doesn't seem to give enough airflow.

I used a spare 12Vdc tap off the MK2S power supply but any 12Vdc supply should work as long as you use a 12Vdc fan. I also used the same tap to power the LED lights at the top of the Enclosure I made for the printer. See https://www.thingiverse.com/thing:2396990.

Prusa i3 MK2S Enclosure

This post is in answer to Tinchus2009 who asked whether or not my comments regrading 60/40 airflow was backed by any tests. They were using an airflow meter. while the openings in the two ports are substantially different leading one to question how this could be true. The answer is in how the airflow is picked up down stream of the fan. Note the scupper inside the Nozzle that picks up a disproportional amount of the air in relationship to the small port. That along with the friction with the overboard pipe causes the airflow to divide 40/60 percent. I played around with the scupper surface area until I got the desired airflow from each port.

Hi. This is one if the best designs I found, really nice, thanks for shareing.
I haev printed all pieces, in my case I wanted to use what I have here right now, and I have a fan of 80 mms, I really dont know if it is a high pressure one (but I added this to improved its pressure: https://www.thingiverse.com/thing:432630, for those asking, YES it works). I used 2 adapters (https://www.thingiverse.com/thing:21112) to fit this fan into your design and the assembly looks great.
But I noticed something while testing: you say that the noozle gets th 60% of the air out and recirculates 40%. At the entrance of the noozle I can perfectly see and tell that the air deviation you design keeps that proportion, BUT the exit of the 40$ part, the one wich recirculates the air, it is too small and, I have to say I didnt run a simulation in solidworks because I dont have the original files, Im sure that small hol will create some back pressure, I think makeing that hole bigger would be more effcicient, what d you think?

Laminar Flow Fan plate
by qboney
Various fan size conversion adapters.

Would it be possible to post the original .step or .sldprt files for the chamber and filter_to_motor parts? Solidworks feature recognition is having trouble translating them.

The files were generated with Fusion 360. Unfortunately the Model is part of a larger project which contains proprietary data and cannot be shared in Model form. The chamber is a relatively simple shape and should be easy to recreate in Solid Works.

The original work is CC BY-NC 3.0.

wher buy the filter, do you have a part number


Its also in the BOM at the end of the narrative along with the other part that are purchased and not printed.

Why do you use activated carbon pellets rather than an activated carbon filter?
I have extra Carbon Pad for Air Purifiers from amazon and I want to make sure it's ok to just use that.

The purpose of the HEPA filter is to remove as much of the ultra fine particles as possible. If there is an activated carbon filter that has the filtration properties of a HEPA filter that you could use that instead and eliminate the Activated Carbon Pellets. I am not aware of such a filter but they could exist.

I guess I wasn't clear that I was saying use a carbon filter AND the HEPA filter rather than carbon pellets and the HEPA filter.

I don't see any reason why that wouldn't work but it would require a different housing.

Hello, I really like this design, especially being diy.
I print using a Rostock Max v2, which I have already enclosed. Wasn't aware till recently about the particles and odor problems, must be old age sneaking up.
I started building a dlp resin printer, but stopped when I became aware of the fume and odor problems. Which seem to be more prolific than
abs or pla, and nylon. Do think this would work for resin fumes? Any suggestions would be greatly appreciated.

I don't know much about the particles that resin printers give off. It may be less or more than FFF printers. A little research would shed some light on the emission issues. HEPA filters are fairly good at removing particles down to around 10 microns according to the iRobot literature. If the particles are significantly smaller that 10 microns then this filter may not be effective.

Thanks for the quick reply.
I have noticed that some commercial scrubbers for printers recycle the air back into the enclosure, like probably back in at the base of enclosure. Do you see a problem doing this, will it affect flow or back pressure. Or am I barking up the wrong tree?

I'm bad, went back and reread your description and it answered my questions.

I liked this filter design and have print all the parts. I'll use it in a Lack table enclosure. I have not been able to figure out where the part call the "restrictor" would go if I decide I need it. Can you tell me?

The current design has a small opening that is intended to recirculate filtered air back into the enclosure with the majority of the filtered air going overboard (Outside the Enclosure). In some situations that require high ambient temperatures (e.g. ABS) too much air may be exhausted out of the enclosure and cause the ambient internal enclosure temperature to drop too much. In these cases the Restrictor is designed to fit on the Exhaust Tube and reduce the amount of filtered air going overboard.

Thanks for the quick reply.

Great design!
Is the "Chamber Top" part of your final design, or a remnant of a past iteration? It doesn't seem to be necessary nor is it shown in your assembly renderings and physical prototype photos.

Oops! The Chamber top was from a previous design and it not required any longer. I have removed it from the "Thing".

I had to modify the Chamber Top piece for my printer. There appears to be a hair line gap where the top cylinder joins to the base of the print which caused the top piece to always fail while trying to print.

which now that I look closer at all the assembly picture, I probably didn't need to even try to print this part (let alone 6 times).. sigh..

Glad you were able to resolve the problem you were having. I not sure which part you where having problems with. I looked over the .stl's and could not see a line that would prevent printing.

Looks like chuteroo beat me to it. Don't get me wrong, I'm a complete noob and thought this would be a good practical learning project, just didn't realize how much I would learn :) Once I get my printer back up and operational, I'll definitely continue with this build. thanks for sharing.

Hi James, great design. I would like to use 80x80mm high pressure fan which I have laying around. Please is it possible that you share parts which are connecting to the fan in any 3D modelling software format? thanks

Great solution. Thanks for responding.

aeroms: I would be interested in the airflow you get from the 80mm fan with the adapter in place. I havn't measured my setup yet but I plan to borrow a meter from work when I get a little more time.

Thanks James & Olav. will go with the reduction piece and see if it does any harm to the air flow. I do prefer to use what I have on "stock"...
trying to stay on the "green side" :)

I'm not sure that the current design could easily accommodate 80mm fans and still work efficiently. This would mean increasing the width by 10mm on each side while maintaining the chamber as is. The fan I used is only $8.49 on Amazon.

To incorporate the larger fan, you would have to stretch the neck from the chamber to the fan and the fan to the nozzle. It would be better to just scale up the whole thing but that would require a different filter, etc. In general I don't share models because they generally contain proprietary functions but you can take the .STL, scale it to the size you want and reprint all the parts or simply trace the design, creating a new model of your own.

2 changes that would make this way better 4 tabs on the outisde around the filter so you can attach a fascia to hold the filter in for vertical placement and also an option of different fan sizes like 120mm fans

The iRobot HEPA filter is an interference fit in the current design. This fact and the force of air trying to pull the filter in will keep the filter in place in all orientations except for inverted ones. In this case just add rubber bands around the chamber and filter.

With regard to the 120mm fan, this would result in a part that would be too large for the enclosure I designed. Do you have an application that would require filtering a large amount of air? If so I can design a filter for you but it will not be as simple as increasing the size of the fan adapters. Rather I will have to scale most of the parts up to accommodate the roughly 4 times airflow and physical size.

Did you have any proof that the particule are filtered ? I read a lot of write who say that the diameter of this particule are to small to be captured by hepa filter

Nothing that would stand scientific scrutiny. The odor is certainly removed but that is due to the activated carbon pelts. I did place a white piece of paper that I covered with double sided sticky tape while I printed black filament. I did not collect anything that the eye can see after several hours. That having been said, I have not noticed any particles sticking to the Plexiglass enclosure covers either after 100's of hours operation either. Plexiglass normally has a natural charge on it that typically attracts any dust or dirt. So the questions remain:

  • Is the type of material I am printing with (ABS, PLA, HIPS, and PTEG) generating Ultra Fine Particles?
  • If so is the HEPA filter capturing the UFP?
  • If the UFP are getting through the filter, what type of filter would remove them?

I did review some articles on HEPA filters that claim 99.9 efficiency in removing particles of 300 nano meter and larger particles but did not give data on the smaller 10-300 nano-meter particles recorded in some studies.

I my particular case, The filter exhaust port is within 3 feet of an exhaust fan that is moving about 200 CFPM of air constantly. Hopefully this would carry any UFP that is airport outside the structure.

Probably but you will need to secure the actual filter with something like rubber bands otherwise the weight of the carbon pellets along with the CNC vibration will cause the filter to fall out releasing thousands of carbon pellets.

I recently backed a combination 3D printer, laser engraver, and CNC machine. I am definitely going to use this!

Will it work as well if the filter is mounted facing down? Otherwise I worry it will clog with CNC dust.

Looks Nice, Great work!

Thanks RJ. After researching lot of other solutions, your idea was the best. I just improved on a great idea and adapted it for my application.

When printing PLA at bed temperatures of 60°C the ambient temperature remains between 30 and 35°C regardless of Fan Speed. I have not measured the internal temperatures with bed temps of 100°C that is common with material such as ABS. I plan t do so in the next day or two and will update the "Thing" with the results. My expectation is that it will drop the temperatures 10-15°C at full fan speed and without the restriction in the exhaust outlet.

Great design! I'm curious to know if you had a chance to measure temperatures while using ABS, and the temp differences with the different restrictor plates?

I'm waiting for my Prusa Mk2s and it's already fun making a list of things to print. This will definitely be one of them, well, right after making the enclosure. Thanks again!

does this lower the temp in the box when printing?

Not in its current configuration. Maybe when you close the hole that releases 40% back into the enclosure.

Did you try it with the hole closed? For me it's basically useless since my PETG and PLA prints fail with the enclosured closed as it gets too hot inside.

I did not. I did monitor the internal temperature and it never got hotter that 35°C with a bed temp on 70°C. 70°C is what I use for PLA and PTEG. On ABS I run bed temperature of 100°C. The internal Enclosure temp is <45°C for ABS. This is using the restrictor orifice.

Those temperatures should not affect PLA or PTEG. What kinds of failures are you seeing?

Well they do. Once I get up to 30 degrees ambient (in a corner of my enclosure, so I suppose it's hotter near the printer) I get jams because of heat creap. I can consistently reproduce this. If I leave the door open 100% of prints succeed. If I close it they fail after about one hour of running because of the heat.

Running a Prusa i3 mk2s MMU

It may be the MMU is more sensitive to temperature. I just ran about a 30 Min PLA run with the Thermocouple hanging from the middle of the upper horizontal bar. Bed temperature was 60°C. Max temperature was 30.5°C Printer is a Prusa i3 MK2S. I don't have a MMU feature to test with unfortunately.

See attached Pix of the test setup and readings.

Closing off the small internal port and removing any restrictor will maximize the airflow out of the enclosure but you may need to open a few holes in the enclosure to allow more make-up air to enter. ABS wants to have a high ambient temperature to keep from warping during build. It is also very sensitive to drafts. Increasing airflow into the enclosure would not be the best thing to that material.

Is the jam occurring in the hot end or in the MMU adapter?

Yeah but that's quite a short print. Usually when it happens on mine it's after more than one hour. Even had it go up to three hours before it happened (it also depends on the print and how many times it needs to retract the filament it seems). Then, as you said, it's quite possible that the bowden MMU setup with its different retraction settings has something to do with it.
The clog always happens just below the ptfe tube in the hotend.

Setup picture attached. The blue thing on the left hand side is the temperature sensor.
Lot's of cables :)

My enclosure is much different than yours. The front and two sides of my enclosure are Plexiglas set loosely into aluminium extrusions. This allows a more distributed source for makeup air. The picture I sent shows 30.5°C. That temp was stable for about 10 minutes. Earlier in the week I printed a Millennium Falcon (13.5 hrs) https://www.thingiverse.com/thing:1559425. The print settings were identical. The Max temp recorded over the period was 34°C. I regularly print 10+ hour objects for work and have never had a clog. I use 2mm for the retraction length. Bowden takes longer lengths but there is a limit. I have a XYZPrinting 1.0 Pro with the same E3D hot end. It uses a Bowden tube of about 12" in length. I use 3mm in this case. If I increase the retraction length to 4mm or greater I will get clogs at some point in the print, often many hours into the run.

A few questions:

  • What is the retraction length?
  • What Slicer are you using?
  • What is the printing speed in use at the time of the clog.
  • What was the enclosure temperature when the clog occurs?
    • Was the enclosure temperature stable for some period of time or was it ramping up as the job progressed?
Millenium Falcon cufflinks
by MrMilan

I am using Slic3r Prusa edition with the default settings recommended by Prusa for their MMU, this means 4mm retraction length. I don't want to mess around with those settings, because they should be ok for the filament they tested on their printer.

Printing speed is 40mm/s for perimeters and 60mm/s for infill.

Enclosure temperature (the time I monitored it to see if it would clog again, which it did) was 29 degrees. I don't know how much hotter it would get if it continued as I don't continue printing once this happens. I don't think it's the highest temperature it would reach.

I presume I need to work on the air flow. The issue with the filter I printed is that the fan makes a huge amount of noise when going full force, which defeats one of the purposes of the enclosure, being to reduce noise levels when printing during the night :D

I'm going to install a dual 120mm extraction setup with some high pressure Noctua fans behind a carbon filter/HEPA filter setup. Hopefully this helps whilst keeping the noise down. Will drill some extra holes at the bottom to get more air in the enclosure too.

The settings seem reasonable. 29°C Enclosure temperature is well withing reason and cooler than I generally see after 3 hours. If the ambient temperature is the only factor causing the clogs, then the MMU feature is running on the edge of workable. Given Pursa's fanatical attention to detail, I would be surprised if this is the case. Have you checked the E3Dv6 cooling fan for proper operation and also the air path around the hot end? Tg (glass transition Temperature or the temperature PLA begins to soften) is 55-60°C... a far cry form the ambient temperature you are seeing.

The fan I selected is noisy but compact. I run my fan at about 80% max RPM and this quiets it substantially. I played around a little with larger diameter fans but them became an issue given the physical space constraints I have. The Noctura Fans have a good reputation of acoustics but I have not tried one in this application yet.

I disassembled everything multiple times because I first thought there was a problem somewhere or a clog inside the hotend/nozzle/... but nothing to be found.
I've got room to slap 2 120mm fans on the ceiling of my enclosure. I will use a carbon filter and not carbon pellets to save space.

Thanks for the update.

New update. I just slapped an 120mm noctua fan to the inside of my enclosure. Running without any filters on top as a test. Drilled a big hole and screwed the fan to the enclosure. Right now I am printing PLA, in its 2nd hour and temperature is a stable 23 Celcius. With the setup I printed before I would be at 30 degrees by now. It's a lot quieter too.
Will have to see how much airflow I lose by slapping on a carbon and hepa filter, but I have a second fan too, so worst case I will install 2 fans to extract air. Do you know if having the filter in front or behind the fan makes a difference?

Location of the filters with respect to the fan shouldn't matter as long as there is no air leakage before the filters. I would let physical constraints and ease of design/use dictate where you put the filters. I assuming you will be installing both a HEPA and Activated Carbon Filters inline with the fans.

Did you have enough enclosure leakage for the makeup air or did you end up cutting a few holes?

Finally take a look at airflow over the heated bed. Some materials like ABS require high bed temperatures as well are high ambient temperatures to minimize thermal contraction. This is probably not a issue for PLA or PTEG but will be for most other materials.

I didn't need to cut extra holes for now. It pulls in air from the gap around the door (not air tight) and it seems to be enough.

Concerning airflow over the heated bed, I don't print materials that require a heated build chamber. Then again, if I would I would just shut down the fan so the heat can build up. Mainly printing PLA, PETG and got some XT, XT-CF and NGEN from Colorfabb. I tried to avoid ABS (for obvious reasons) and no need for nylons.