Loading
Hey! This thing is still a Work in Progress. Files, instructions, and other stuff might change!

DLP-Based Resin Printer for < $200

by ScribbleJ, published

DLP-Based Resin Printer for < $200 by ScribbleJ Mar 15, 2012

Description

This is a DLP projector-based resin printer that can make very fine 3D prints in record time.

This is a work in progress. Today I plan to begin actual experiments with the resin.

I believe the basic design I am producing here is very suitable, however, the details of the design need changes in almost every part to be ideal. Most importantly, the vat should probably be much shorter, but the whole thing could be made significantly smaller.

I am currently using resin from Bucktown Polymers:

http://bucktownpolymers.com/

(You might notice their price for resin is cheaper than some current PLA prices, by volume)

A2Sheds has provided a page with a few helpful resources, as well as the beginnings of his own DLP project:

http://code.google.com/p/lemoncurry/wiki/main

This is very much a work in progress. Expect more photos and videos to come.

Here's a link to my flickr site documenting this project:

http://www.flickr.com/photos/13723140@N04/sets/72157629564335469/

Video of resin cure here: http://youtu.be/UVDN2xq575A

Next steps: Experiment with resin. Create working software stack.

Known Issues: Vat is too tall/big, glass may be too weak/thin. Vat probbaly needs a nonstick coating, like silicone or PTFE. Z Axis range is 5", vat is 5", so Z stage cannot quite clear the vat. A2Sheds recommends a heated vat to 50c so the resin flows better...

Recent Comments

view all
Hi . can i please have your Marlin file that is configure for this printer
why not use a broker DLP beamer without a functioning lamp and then use a 20W UV LED as light which costs around 35 USD? that way we can drop the cost considerably :)
Id say just search ebay for "DLP projector" and choose buy it now. If you see one with a good res (better than 1024x768 Id recommend) and the price looks good then google the model and type to verify the bulb isnt an LED.

Makes

Liked By

view all

Give a Shout Out

If you print this Thing and display it in public proudly give attribution by printing and displaying this tag. Print Thing Tag

Instructions

1) Get yourself a DLP projector with a Xenon bulb -- not the ones with LEDs. Open it up, and use a pair of pliers to break all the glass edges off of the color wheel. I'll take some photos later, but it's pretty obvious what is the color wheel; it's just like it sounds, a wheel with some various colored glass bits on it that the light goes through to make the colors. If you break off the colored glass and leave the wheel in the projector, you don't have to worry about whether the projector will sense the wheel missing and refuse to run; it still thinks the wheel is there.

2) While you're all up int he projector, loosen the final optics, the focusing part, and move it further away from the innards of the projector to get a closer, narrower focus. You want it to oply be a couple-few inches across. Remember, when you do print, it's the resolution of your projector over the area of your focus that determines how fine your print can ultimately be, so smaller area and greater resolution is ideal.

3) Construct the frame - you should see how it works from the photos. There's two MDF clamps in the bottom to hold the projector in the proper position.

  • Using a rubber mallet to pound the smooth rod into the Z axis ends is a good plan.

  • I've included a copy of nophead's Z coupler here, slightly modified. Instead of using vinyl tubing, I attach the coupler to the threaded rod, then shoot the open end full of hot glue, and then quickly attach the motor end before it cools. I find this works spectacularly well.

4) Adjust the position of the projector so it focuses on the bottom of the vat, by dropping a bit of paper on the bottom of the vat to see how you're doing.

5) Get to experimenting... if you've come this far you're about as far along as I am. For the record, the "Pronterface" host suite supports (to some tiny extent) prjecting an SVG for printing, and Skeinforge can export them, but I'm working on something a little more tuned to the task as well.

COSTS SO FAR:

1 used but working DLP projector from eBay: $75 1 2'x4' MDF sheet: $10 (some) 8mm threaded rod: A few bucks 1 8mm smooth rod: A few bucks Buncha 8mm Nuts and Washers: A few bucks 1 5"x5"x5" fishtank: $25 (not a good choice) 1 bit of glass for the platform: $5ish 1 Stepper motor: $7

I'm currently using a RAMPS board to drive the stepper because I had some on hand, but any cheap little single motor driver should do.

Hi . can i please have your Marlin file that is configure for this printer
why not use a broker DLP beamer without a functioning lamp and then use a 20W UV LED as light which costs around 35 USD? that way we can drop the cost considerably :)
Has anyone built and successfully printed with one of these? I'd like some info before I start making anything.
Has one been made... There seem to be 2 people who have done it. Downside.. finding a DLP projector for 75 as he did off eBay is VERY rare. most DLP projectors cost @ 200 by themselves if not more.
Even if you pay $200, I'd think a sub $300 Resin printer would be a great deal. :)
Do you have any info on who built them, so I can contact them with a few questions?
Id say just search ebay for "DLP projector" and choose buy it now. If you see one with a good res (better than 1024x768 Id recommend) and the price looks good then google the model and type to verify the bulb isnt an LED.
Scribble, If I were to remove the polarization filters from a LCD display and backlight it with UV, would this allow for the UV to pass thru? Trying to look into using that or a DLP TV base to develop since I have easy access to them?
Hi Nargousias,
It won't work. The lcd display can't take the UV. I'm using a unmodified DLP beamer (infocus x9, 180 watts bulb) and it works like a charm!
Im slightly curious.. what lumens is your projector putting out? Also what res.. HD (720p) or just plain old 800x600. I would think higher lumen and higher res would increase cure speed and resolution of prints. I would also recommend shielding your whole build are so only the projector is feeding for the cure rather than any ambiant light source.
If anyone is looking for a suitable Slicer and Control software, I would encourage you to try Creation Workshop : thingiverse.com/thing:40778
Creation Workshop - SLA / FDM Slicer and Controller
What software do you use to print with? I am confused on what controls the projector and z axis
This is an epic idea. I'm really interested in trying this with the projector from my 62" DLP TV (massive xenon bulb with tons of UV output). I should have enough power for a build surface 1 square foot or smaller. Have you made any new progress since last post? Any recommendations for a noob like myself? I've checked out the lemoncurry site but there doesn't seem to be much in the Wiki/how-to department. You seem to have taken it way past the theoretical department and into the open-source world.
Hi,
I am very interested in building my own dlp printer but have yet start as I do not think I will be able so create a piece of software to run the machine. How is your build going? Do you have or know of an appropriate piece of software?
Josh
I'm a bit confused. I'm trying to make one of these using an old LCD panel from a laptop and some UV LEDs, but I don't know how the model will stick to the stage but not the bottom of the vat. Is it a materials difference, or is the stage somehow primed to hold the model?
davr - in reply to Guest
LCD wont work with UV. The panel will absorb all the UV, not allowing any to pass through to the resin, sorry.
the key is that lcd is supported on glass, which is opaque to uv, whereas plastic is mostly transparent to uv.
Hi! I have TOSHIBA TDP-S8 projector with removed color wheel and 0.5 litre a flexographic UV-cured varnish. On sunlight his cure for 20 sec, but no cure from projector light. What wrong? :(
Something else might block the UV in the light path.

nycresistor.com/2012/04/06/uv-cured-3d-printing/
still looking for specific models of dlp projectors used. cant find anything on lemoncurrys site. scribblej what dlp projector did you use? are you getting good prints out of it yet? anybody else have a model that is working for sure? I have just been given a budjet and need to purchase this asap. any help would be awesome thanks
Also, to get fine resolution, what kind of lead screw is needed?
No one else has answered this question which is the one thing I don't understand about this process. What software is used to send the sliced images to the DLP projector and what software is used to create those sliced images?
what models of dlp projectors have been used. lemoncurry has said there are certain models that work without modification. i see nothing about specific models
Just bought a broken DLP on ebay for $30. However it doesn't use a Xenon bulb or LEDs, but a "UHP" mercury arc lamp.

en.wikipedia.org/wiki/UHP_(lamp)

Should put out some UV anyway though. It doesn't matter unless I get it working, and if I get it working I can always sell it for more than $30 if it doesn't produce enough UV.
bah, ebay seller lied. they said the projector does not power on (which I thought would be an easy fix), but in fact it does power on and the bulb is just burnt out. A new bulb is more than what I can get a working projector off craigslist for. So... returning that.
I use mercury and xe lamps for research and mercury bulbs output way more uv light but it tends to be concentrated at certain wavelengths. There is a strong peak in the 200 nm range. I don't know if polymers work when you are above the wavelength.
Hey thanks for the info! I don't think the polymers respond much to light outside the target wavelength, but hopefully there is still plenty of light in the 385nm range.
Perhaps using this system with a LCD screen and UV led lights would also work. dealextreme.com/p/5mm-uv-led-emitters-20-pack-2398 That would be cheap and nice :D
davr - in reply to lemio
LCD screens do not pass the UV light, which is why this setup doesn't work with LCD projectors.
The issue with LEDs is that if one is not enough we might have to use an array? In that case we'll have issues with uniformity of illumination. Has anyone tried out the LED concept yet? It sounds great if we could do that, since LEDs are more efficient.
leds are designed to put out specific wavelengths (the materials used do not normally emit broad spectra, so they usually put out "spikey" spectra, with pronounced "lines"). and so they might not be putting out very much uv at all, no matter how bright they are. (unless the led was designed for uv, its uv output would be considered "waste".)

other technologies, like mercury vapour and fluorescent, inherently produce lots of uv, which can then be converted into visible light by phosphors, etc. (and they may have "uv shields", either as part of the bulb, or in the fixture, or in some halogen lights, in both. after some negative publicity, the manufacturers must have decided that you can't be too careful with product liability lawsuits.)
have take some test? why dont use simple xeon car kit with lcd? more powerfull light and simple to change.....

i want use small high resolutionion lcd screen on base of resin tank and use alogen or uv lamp as backlight.

but not know it work....and too much difficult find resin in italy to...have possible to try it (have micro lcd screen used for small projector) but no have resin :(

for z asix think use
paralel port with simple unl2003 and 5angle x64 redution...
wildthang - in reply to
lcd is made on a glass substrate, which filters over 90% of the uv.

you'd probably start a fire before you could blast enough through to cure the resin. and waste over 90% of the power that you use (and pay for).

but your actual mileage may vary...
I doubt it. LEDs are nothing compared to real bulbs.
leds produce approximately 10x the lumens per watt of incandescent or halogen. higher wattage arrays, producing 1000s of lumens are definitely a challenge to find, but are out there. and they're easier to make than compact fluorescent arrays, just because of the geometry.

it's not a mature technology, yet, but a 10w led is impressive, plus the mechanical designs are still a bit crude, but i think they're already a realistic replacement for many "real bulbs", and the price does need to come down some more.
Do these polymers react with normal day light? if so wont poring the stuff into a tank that is exposed to the elements cure? if so how do you get the part out?
filtering uv is not difficult. ordinary window glass is over 90% opaque to parts of the uv spectrum. and coatings can make plastic (as in plastic sunglasses) fairly opaque as well.

and it also depends on the resin, which may require either higher intensity or longer duration. (if you think of it in terms of sunburn, most people can tolerate some sun exposure without "burning".)
It all depends on the formulation but, at least with the case of the Bucktown Polymers mixes they should not. That being said I wouldn't do it close to a sunny window. :)
Loving it, thanks! This is way cheaper than this DLP printer:

indiegogo.com/veloso3dprinter?c=gallery
Here is my 2 cents. How about using instead a Black light or UV Lamp with placing a LCD panel as close to the resin as possible to reduce soft shadows. Black lights are rather cheap and hit the spectral range and come in a variety of ways, like bulbs, tubes, flash lights, etc. UV lamps come in high watts which might increase the speed of printing. LCD panel might be a pain, you would need to remove them from their frame/backlight so light passes through them, but they do come in variety of sizes and pixel densities.
I had the same idea when I first saw Junior Veloso's printer a few years back. I made a basic prototype and found that, unfortunately, all LCDs I tired had a lot of UV filtering in them.
That is unfortunate but actually makes a lot of sense, now that I think about it, since you wouldn't want a lot of UV going into someone's face.

Those pico projectors don't seem to be that expensive compared to LCD monitors/panels, and miicraft is using them for their dlp 3d printer.
it's not a "filter" as such, but the glass substrate that the lcd is built on - ordinary glass is opaque to certain parts of the uv spectrum.

lcds that are built on plastic shouldn't have this problem, but they are not common. (and they shouldn't need uv filtering, since there is nothing in an lcd panel that needs to be making any uv at all.)
is it possible to remove the UV filter? Often the filters on things like that are just a plastic film.
I have all these parts and I've removed the films from the LCD panel. Just wanted to know what best combination of VOC-free resin and UV bulb for the backlight.

Also I have Slic3r for my repraps along with using netfabb for the images and was looking for a better setup. is there anything I can download?

I am simultaneously making a DPL based one as instructed here with the acrylic platform. Just removing the color wheel now. Was thinking about jumping f
orward to a larger build area vat with a 2 axis LCD, LED or DPL mount. I'm beating this with a stick in the dark so ALL suggestions and ideas are appreciated before I leap into building the carriage.

Freddy
So maybe make a "reverse scanner". Sweep one of these 1200 dpi print heads back and forth under the bottom glass, using the mechanism from a scanner (or just build your own):

oki.com/en/otr/downloads/otr-161-14.pdf

You should be able to scavenge all the parts you need from a cheap machine like this:
http://www.amazon.com/B431D-Laser-Printer-Monochrome-Print/dp/B003V9YBQG/

That gives a build volume of 8.5" by unlimited by unlimited at:
(1 / 1200) * in = 0.0211666667 millimeters

The trick would be finding a
polymer that is sensitive to the LED wavelength.
I remember 3d printers of the late 80's being a platform lowered into a vat while a UV laser scanned each layer
that's a really neat idea! i like the potential for large build volumes at the expense of mechanical scanning, it seems like a reasonable tradeoff for large stuff!
Any chance you could link to a cheap little single motor driver for a single stepper?
Also I just realized if you have something like an Ultimaker, the extra extruder stepper driver output could be used to drive this.
StepStick is the open-source variant of the pollulu stepper motor drivers, used in RepRaps (see reprapworld.com/?searchresults
&
amp;cPath=1591_1599 ), they are cheaper and open-source so why not :D
Here's another popular one: pololu.com/catalog/product/1182
I have used these on quite a few projects:

sparkfun.com/products/10267

They work very well.
Definitely go with the pololu board. The chip used by that one allows much smoother microstepping and can handle almost 3x the current (the easydriver is for weakling steppers only).
Depends on the resin formulation.
compared to heat softened ABS how is the smell of curing resin?
Bucktown Polymers has a variety of options available. As I was an early adopter, the only choice they had at the time was the 100 formulation, which is pretty heavy in VOCs. Which is to say - you really don't want to smell it in the first place. It smells like cancer.

However, there are someother formulations now that should be VOC-free, and when I run through my supply of the 100, I'll switch to those.
I love the smell of cancer in the morning. :)
Depends on the resin formulation.
Any tips on what projector would be ok? I guess not any DLP projector will work ?

Also saw somebody mentioning a 'chinese'c contact that sells DLP projectors specific for DLP printing.. Anybody have more info on that?
Just found these NOS DLPs: surplusshed.com/pages/item/l13723.html

Maybe they'll work‽
alibaba.com maybe? Lots of inepensive chinese stuff on there, DLPs included.
I don't know what you're referring to specifically, mhensen, but I do know that A2Sheds from the IRC channel is working on sourcing some DLPs that are designed specifically for this purpose. I'll let you know more when there's more to know. :)
here's a group of people that's doing similar stuff with Pico DLP Projector. maybe you might be able to get some infromation or help from them.

miicraft.com
Do the polymers change in volume when cured? If not why not project from the top, avoiding the problem of a sticky vat? The only other argument I know of against top projection is the amount of polymer needed to fill the tank.
Looking at this video, I can see a big (well, tall) benefit of bottom projection:

youtube.com/watch?v=cERllFq2VEg
Arthur has some great photos of his progress on a similar design (used mechanno for the frame instead of MDF) here:

flickr.com/photos/arthurwolf/with/6854626982/

He's had a bit more time to throw at it than I and has just started getting results. For the record, he's coated the bottom of his vat with PTFE tape, and uses a method where he exposes a layer for 30s, then moves 5mm up and then 4.xmm bac
k down for the next layer, because it still sticks to the vat some. His platform is acrylic.

I'll probably be rebuilding a smaller vat and switching to acrylic platform before I carry on.
And for the record, if you want to see how it's all going in realtime, we tend to hang out in #lemoncurry and #reprap on freenode IRC.
Very cool!

I was under the impression that DLP DMDs had UV filters before them since TI's datasheets claim damage at milliwatts/square cm of UV. Do you have any way of measuring how much UV light is coming out of the thing?
I've got a friend with a spectrometer I might go see tongiht, but it's not likely I'll actually have time. Removing the color wheel eliminates the UV filtering in most cases but some projectors do have an additional filter you need to remove.
Why make your vat any taller than the height of the layer you are curing?

You can just add more resin as you go... and it also minimises waste.

I take it you have seen juniors machine? Plus the 2 or 3 others like this?
It was what I had on hand. I thought I made it pretty clear I thought it should be much shorter.
Sorry, the new Thingivese design has turned me off from ever reading the text again...

I read the first line then just scanned the pics!

It will be interesting to see your build times, I know Juniors and resin machines in general are rather slow.
I knwo what you mean, Thingiverse drives me nuts. :)

My initial tests were taking a long time to cure, but it's possible I was "Doin it Wrong" -- I'll let everyone know when I have something interesting to report.
What sort of layer thinkness were you curing?

And iirc junior was reporting build times of maybe less than an inch an hour.
I wonder if a laser could cure this stuff.

If you did the same thing, but with a laser, you could use all the regular 3d printing toolchain. X and Y would step mirrors instead of a stage.

With the right gearing you could make huge resolution, a bright enough laser could probably cure pretty fast.

But I guess you lose the "parallel" spee
d gain that the DLP gives you. Right back to only a single "spot" of action, drawn out in serial.

For that matter, maybe a DVD burner laser on a (lightweight) X-Y platform mounted just under the tank glass? All the drawbacks or a ABS printer and none of the benefits of the DLP . . . . perfect.
Worth looking through... instructables.com/id/Build-a-Laser-3D-Printer-Stereolithography-at-Ho

I believe (but may be wrong) that a DVD writing laser is not powerful enough, nor is at the correct wavelength to cure these resins.
So you project the image up through the bottom of the tank, onto the bottom surface of the movable glass build platform. What determines the layer thickness? Is it exposure time? Is it the distance between the movable glass build platform and the stationary glass bottom of the tank?

Have you considered different material for the stationary bottom of the tank based on spectral absorption? For instance, I think regular glass blocks some UV and plexiglas passes UV (based on urban legend of pilots behind plexiglas getting sunburned, but not in a car :) )
I am pretty sure plexiglass blocks UV, but it could vary based on the type. I actually use a plexiglass box to block a 266 nm laser in my lab and it works great even though the laser is quite strong. Hopefully we could use something besides fused silica, but that would definitely work. (Probably $26 for a sheet on McMaster Carr.)
I believe automotive glass comes with a very light tint that blocks a few percent of visible light, but nearly all UV light. I'd imagine an airplane would have even better coatings though.
If I knew more about materials, I might make that kind of consideration. :)

The layer thickness is determined by a multitude of factors, but primarily it's a function of how much UV your DLP puts out, how long you expose a layer, and how much UV the resin itself blocks. The resins from Bucktown all have UV blockers so that it doesn't cure everything between the projector
and the platform.
Unfortunately UV polymer systems don't work like that. ;) I formulated these for almost 15 years before changing jobs.

A UV formulation consistes of:

1. Resin(s): High molecular weight polymer (typically a acrylic/epoxy/urethane/polyester (meth)acrylate) that provides the primary properties for your formulation
2. Monomer(s): (meth)acrylate functional low molecular weight polymers that are used for viscosity redu
ction and some physical properties
3. Photoinitiator: a low molecular weight polymer that photocleaves when exposed to a certain wavelength(s) and intensity of light. The wavelength of light that they are sensitive to is variable and you can mix and match wavelengths by choosing different photoinit
iators. My guess is that they'd use something like Irgacure 819 or something sensitive to more visible light regions.
4. Additives: these could be surfactants, deaerators, antifoam, flatting agents, colorants, fillers, etc.

Most likely the polymer systems that you'd be working with on this would ha
ve only components 1, 2, and 3.

What happens is that you shine light on your formulation and if there is enough light intensity of the right wavelength, the photoinitiator will cleave generating free radicals. These free radicals then initiate the polymerization of the resins and monomers via a Mic
hael addition reaction.
Anyone else think it's cheating to base your cost on getting a one-time deal for used parts?

Still cool, though.
If the DLP projection system from a big screen TV is usable for this then you can get them even cheaper than he did. I see DLP TVs on Craigslist for free all the time. Gut them for the projection unit, throw away the cabinet and all it costs you is the cost of going and getting it.
Honestly, it's pretty easy to find a used DLP for the price I mentioned on Ebay, and probabyl even easier on Cragislist or at local offices. I feel like it's more cheating to not include the cost of the printed parts at all. :)
I just ordered 6 "small" parts (each about a 1/2" square cross section and about 2" long) from a local prototype company to be printed on their high resolution (0.0006" layer thickness) ObJet, and the bill was $400. LOL--that was supposedly cheaper than their SLA. So even if he put a "2" in from of the $200, if it was used very often at all it would pay off quickly. 8-)
It's called bravado, as long as he doesn't ask for my money I love it! :)
I note from the bucktownpolymers site that they plan to have different polymers that cure at different wavelengths, it would be interesting to have three different wavelength/colour combinations mixed in the same vat and print in full colour all at once.

Still, walk first. Good luck, I'll be eagerly watching your progress. :)
If you talk to A2Sheds, he'll tell you he can produce a resin that can have color ranging from from red to blue as it is printed, based on exposure to an electromagnetic field. That sounds awesomely high-tech. He'll also tell you it costs a prince's fortune to make. :(
Unfortunately UV polymer systems don't work like that. ;) I formulated these for almost 15 years before changing jobs.

A UV formulation consistes of:

1. Resin(s): High molecular weight polymer (typically a acrylic/epoxy/urethane/polyester (meth)acrylate) that provides the primary properties for your formulation
2. Monomer(s): (meth)acrylate functional low molecular weight polymers that are used for viscosity redu
ction and some physical properties
3. Photoinitiator: a low molecular weight polymer that photocleaves when exposed to a certain wavelength(s) and intensity of light. The wavelength of light that they are sensitive to is variable and you can mix and match wavelengths by choosing different photoinit
iators. My guess is that they'd use something like Irgacure 819 or something sensitive to more visible light regions.
4. Additives: these could be surfactants, deaerators, antifoam, flatting agents, colorants, fillers, etc.

Most likely the polymer systems that you'd be working with on this would ha
ve only components 1, 2, and 3.

What happens is that you shine light on your formulation and if there is enough light intensity of the right wavelength, the photoinitiator will cleave generating free radicals. These free radicals then initiate the polymerization of the resins and monomers via a Mic
hael addition reaction.
Shame. Thanks for the enlightenment ;)
Sweet zombiejebus... =-O
Top