And So The Deluge Of Resin-based 3D Printers Begins

It looks like 2012 is shaping up to be the year of the resin-based 3D printer. The latest comes from [Michael Joyce] and is called the B9Creator.  Like other resin printers, [Michael] used a DLP projector to cure the print one layer at a time. The layer height is on the order of 100 microns – crazy for a kit-based printer.

There is a  Kickstarter for the B9Creator where kits are available for $2400 USD. Everything is included in this kit, including the DLP projector and a kilogram of resin. $2400 is much more expensive than even the fanciest melted-plastic 3D printer such as a Makerbot or RepRap, but that’s the price you pay for high-quality prints.

Of course this project comes a month after an earlier, similar, and shadier project called the Veloso 3D printer. The B9Creator promises to be open source once all the Kickstarter machines are shipped out, and [Michael] is very open about his designs and his resin formula – an admirable quality in a maker.

You can check out a load of videos of the B9Creater we found after the break.


71 thoughts on “And So The Deluge Of Resin-based 3D Printers Begins

    1. Traditional SLA requires a vat of photopolymer as deep as your z axis dimension. This upside down build technique only requires a few millimeters of resin in the build tray at any time.

    1. lemon curry? In the spirit of truth and transparency it should be noted that “lemon curry” exists as little more than a conglomeration of information gleaned largely from the work and postings of others. It should further be noted that thus far there is no lemon curry hardware. That Lemoncurry has no unique code. Finally, it should be noted that the lemoncurry forks creator is the owner of bucktown polymers, so it is essentially an advertisement.

      1. It should be noted that bucktown polymers was started *after* the inception of lemon curry, once it was realised there was no cheap source of photopolymer to the hobbiest market.

        1. Hey Graham. I saw an old post of yours on the z402 group. You said you had found its service manual? Did you mean the user’s manual or the actual service manual? If it is the service manual, I would love to get my hands on a copy.

  1. Might be a naive question, but why the steep price tag? I get that the photocuring resing is a bit expensive, it’s not like it’s something most people buy and use.

    But the printer itself? Is the projector insanely high resolution or something? A few fairly basic parts to form the vat, the z-axis and the plate the model is printed on. A stepper-motor and the needed electronics. It would seem, hardware-wize, that this should be around the same price range as the MakerBot and the RepRap, at least if you could get a “bring your own projector”-kit.

      Makerbot replicator $1749

      B9 creator $2375

      I wouldnt call it a steep price…
      especially when compared with $48.9k for a comparable commercial version

      A no projector option would be nice for those wanting to upgrade to 1920X1080 instead of the b9s native 1024X768

      btw: article states a layer height of 100 micron, reading the faq shows that the b9 does 100 micron xy in its wide field and 50 micron xy at its narrow field setting with z axis tunable from 10-100 microns.

      1. 10-100 micron, that’s better. When I read the 100 micron I was like “that’s no improvement, that’s my current layer setting on my Ultimaker”
        But people have also printed 20 micron on Ultimakers.

        We’ll see, maybe my Ultimaker is vastly outdated next year. But for now, the Resin 3D printers have been few.

    1. You need a piece of model-slicing software that can perform a decent infill on the resulting slices then export to any 2d image format. The support software burden for a resin printer is low enough that most of the work can be done by existing software for FDM printers, with the exception of the slicer.

      Right now there exist python programs that will handle very basic model slicing – and there exist programs to cross-convert python code into javascript. Since there also exist JS libraries to handle image synthesis, I can only assume that noone with the relevant skills has had a spare afternoon with the concurrent desire for internet-fame necessary to make an in-browser slicing program. I do not know why.

  2. It looks like the vision of the future in Cory Doctrow’s “Makers” is slowly unfolding before our very eyes. I remember when the hardware hacking scene was a small niche community, then the hipsters arrived to ravage the scene and then moved on to the next trend.

    The ones that have stayed behind have created some marvels and I’m really enjoying watching where the scene is going.

    1. This is very true.

      Tri-colour Reprap = $1088.80
      Dual-colour Makerbot 2X = $2,799.00

      I recently saw a Glenn Beck interview with Cody Wilson. Glenn proudly announces his newly pruchased Replicator, and points to it displayed on a spotlit pedistol. He asks Cody, how much does the material cost for me to print an entire gun on this bad boy? Cody basically tells him be bought a lemon and a steriolithography process is needed to accurately reproduce the parts. LOL

    1. I was privy to this kickstarter before it went live, and there was a $120 reward for a kilogram of resin. Also a $60 reward for a half kilogram. I have *no* idea why Kickstarter took that down, and if I were Michael I’d be pretty damn mad about that.

      The kickstarter says the eiffel tower cost $1.20 to print; or $0.10 a gram. That’s in the ball park of about 4 times as much as RepRap/Makerbot filament, but I’ll be damned if you can get that quality on a reprap.

      1. The article says the resolution for this is ~100 microns. Typical resolution for a reprap is 200-250 microns, with 100 being fairly easy to calibrate for, and the current record vertical resolution on a reprap is 10 microns.

    1. Having just made a rather accurate wood 3 axis CNC for a college project for $250, the main cost to build one of these yourself would be the stepper motors, the precision threaded rods to move the platform, and the projector. If done right, this shouldn’t cost more than cost of projector+ 200$. In the end though, either you get what you pay for, or what you put the time in to design and build.

      1. Yeah im waiting on a Shapeoko CNC. But i will of course build another. I have quiet a few steppers on hand, even a DLP (projection) 50in TV. Im sure i could build this with less than $100 worth of bought items. Its just the time and effort part.. I dont have time to put effort into something like this.

        But as you said something like this wouldnt cost too much.

        I think Approx. $500-$600 would be enough to make one… With $1000 i would be able to make a nice High quality one. I tried to start a Kick-Starter project about 2 times and they shot me down. So i say screw them! perhaps its because im neither white or black :) (hispanic)

  3. I really like the idea of resin printers, but TBH I still think they will be prohibitively expensive for a while yet.. AFAIK the Ultimaker (which IMHO is expensive, if you have time and DIY skills to build your own 3d Printer..) can print all the way down to 20 microns for less than half the price? whereas this can only do 100microns. or am I missing something?

    1. This project currently has two focal positions 100 micron (xy) and 50 micron (xy) and the Z axis is software adjustable from 10 microns to 100 microns.
      Show me an fdm machine capable of depositing a volumetric pixel 50X50X10 microns.

      Better yet show me an FDM machine that can deposit two such volumetric pixels separated by only 50 microns.

      When you say a FDM (reprap, ultimaker etc) has done 20 microns you are only commenting on the thickness of the Z layer. Even if you lean on the xy step resolution…ultimaker claims 12.5 micron you must take into account the 400 micron (ultimaker) nozzle diameter.

      Additionally since the exposure is done layerwise your time per layer is significantly less than if you were tracing out the image from a single point, an advantage this technique has over FDM as well as traditional laser based SLA systems.

      Obviously the downsides to this technology are:

      total build area is limited by projector resolution/intended accuracy (ie 50 micron XY build accuracy is half the projected size of 100 micron XY)


      Price…1024X768 DLP projectors are more expensive then some teflon, brass, and nichrome wire. 1920X1080 projectors will cost more than your reprap by themselves.

      Ultimately, if a reprap can produce parts the size and accuracy you require then there is no need for you to consider this technology. If you work in small scale high detail ie jewelry, dental, small scale modeling etc…this is for you….and an extremely good deal.

      The closest comparable commercial system has a build area of 60 x 45 X 100mm, 43 micron XY 25-35 micron Z and is only 20X the price.

  4. I don’t have much use for models of Yoda or a Goldberg sphere, but I DO have use for functional mechanical components. Which 3D printer could I used to make good quality functional worm drives or planetary gears?

    I have seen some pictures of a pretty chunky herringbone gear on thingverse, but even so, I do wonder if anybody makes anything other than test prints and ornaments on these things.

    1. The thingiverse is full of various mechanical things, but there are not a lot of designers working on engineering projects they are willing to post.
      That said, you can totally print pretty small gears, provided your printer is either well tuned or built for you. I’ve printed a pretty decent clock, and I’m happy with the quality of the workings.

    2. I’m working on one of the software packages for FDM printing (RepRap like printers).
      And while I do everything OpenSource, I do get requests on email for help. Usually these people are printing prototype cases and stuff that they don’t want to share. To help them I usually can get the model, and delete it afterwards. But people are doing pretty cool mechanical things.
      But as an example, printed lathe:

      The quality is also a bit “random” if you look online. Many older RapRap models are just outdated and do not get the quality that you see from other machines.
      I haven’t printed mechanical parts in a while, but these cube gears are always cool: (they turn)
      As for quality… well, see for yourself:

    1. The short answer is “maybe”. The long answer is: Sure, but it’s an extra few layers of work while you design/develop the mountings for new lenses and whatnot. You’ll also need a new lamp, so it won’t be that much cheaper.

      However, if cash is in much lower supply than time and engineering experience, go for it!

    1. Multiple reasons. For one moving parts wear and lose alignment. Also they cost more than what a projector will cost including the modifications to it.

      Closed loop galvos are what are normally used in a situation like this and is what is normally used in SLA machines.

      Then there is the issue of the laser. If they use a laser this machine will then have to meet FDA requirements for laser product.

      A projector is just so much simpler.

  5. I want a 3D printer so I can manufacture an inkjet printer that doesn’t suck and doesn’t use expensive ink refills! That would be a prime use for a 3D printer!

      1. “what is so shady about it?”
        1. Junior leaned heavily on others when he was developing under the guise of “open sourcing” the idea, until he got results.
        2. Junior has been promising updates in a week that either never come or arrive MONTHS later.
        3. This pattern of silence has continued through the indiegogo campaign.
        4. The indiegogo campaign, until today, only showed images of juniors prototype machine, and prints and videos from the same….Juniors prototype machine was printing @ 50 microns XY. The campaign is for a hypothetical machine with an XY of 147 microns. Larger build area sure but no where near the capability the images imply.
        5. Today feeling the pressure of the B9, they finally unveil their machine….or rather a computer rendering of what they think it might look like if somehow they raise another $200k


  6. To people complaining about the price, please bare in mind that this is made from laser cut aluminium which is then bent on a cnc press brake and anodised. The maker bots are just plywood. If you had this design made as a one off by a laser cutting company it would cost you a pretty penny. This is almost pro quality so expect a slightly more pro price. If you want the cost to be like a projector plus 200 bucks then wait a few years and you will see the ply wood versions.

      1. Hardly the same thing. Jesus, are people not even looking at this thing?

        The solidoodle is made from some trimmed up angle iron. This has a fully custom, enclosed, UV light-blocking case made out of anodized aluminum. It’s not cheap.

        If you want 600 dollars, go buy the Solidoodle.

        I think people are missing the plot here.

        If you are the cheapest person on the planet and really love hacking things together just to print Yoda heads on, this printer isn’t for you.

        Mike has gone the extra mile and put the product placement where noone else has gone yet. The semi pro or hobbyist that wants a solid working product with very little tweaking.

        This isn’t for the robotics geek that just enjoys the thrill of making the CNC machine. It CAN be once Mike open-sources the specs and people start playing with bigger build platforms, etc, but it’s just not for the guy trying to get everything in life as cheap as humanly possible even if it’s made out of scrap cardboard.

        The resin model is also nice for folks craving higher detail, intricate features, overhangs and easier post finishing. So you can sand and tap this without worrying about pockets, etc and it winds up being a much cleaner model in general because there will be no stringing, etc and you can print complex overhangs, etc much easier.


      2. Yeah Solid doodle looks fine for the money but far from an industrial product as it is just just a box spot welded from thin steel angle. The material costs alone would be much smaller as it’s tiny and steel not ali, plus no finish on it, just bare metal. The cube by 3D systems is $1200 and very professional obviously but I bet at it’s heart is some cheap folded steel sections then covered in plastic injection moulded parts (just like your inkjet printer), this is something they can do because of the numbers they will sell. This printer is closer to custom one-off type manufacture so the absolute cost is high however I think realistically it is good value.

  7. There is a lot of talk about resolution, but hardly any about the actual material qualities. Veloso claims their 150USD/Kg resin to be “like ABS”. Whatever that means? Here I see no information for the structural qualities of the resin once cured. 40USD per pound for the resin sounds good, but what type of material is it actually? How does it really compare to the typical FDM materials ABS and PLA?

    1. It does look pretty good.

      The shutter is not the slow part here. The curing of the resin is.

      The other concern I have is the projector position. When dealing with something with fluid I would rather not have the projector below it where it could get into it. Second, most projectors are not designed for vertical operation. Some you might get away with if the lamp electrodes are parallel to the axis of rotation. If it is not and you are tilting the lamp it will kill the lamp life significantly. And at probably $300 a pop that will get expensive fast.

      I think I would also use at least a 1280×1024 projector or maybe even jump to a 1920×1080.

  8. After looking though the info on the Kickstarter I have found a huge issue. It used a Vivitek D535 which has the lamp in a position that will cause it to be outside its intended operating position, horizontal +/-15 degrees. What happens here is the heat from the arc is focused on one electrode and causes a temperature gradient on the lamp. Two things can result from this, one the electrode is slowly eroded away sputtering the gasses under the deposited metal causing the pressure in the lamp to drop and significantly reducing the lamp life. Second, the lamp can explode. Not pretty, I have seen it happen.

      1. It is possibly wrong to assume that the guidance in the manual is more trustworthy than his actual experience. It could be that the problem you mention goes away at 90 degrees but exists between 15 and 70 say. Adding a 45 degree mirror may seem to be the way to do it but it increases the projector mount overhang, widens the printer potentially and might increase the cost. When open-sourced you are free to show how he should have done it in more detail.

      2. No, I am basing this off the lamp orientation itself primarily and the manual recommendations secondary. I have years of experience in the lighting field.

        The project being open source does not remove exclude himself from product liability. He is selling a product wether it be plans or a finished unit with serious flaw. If, in a worst case scenario, the lamp explodes and ruins the projector, who is going to replace it? The warranty will clearly be voided, not only has it been modified for a shorter focal length but it has been operating in the wrong position.

        1. I’ve suggested that he do some lamp lifetime tests to allay fears but he has obviously not run into any problems with his projector and he has been running it since November (not continously obviously.) I’m certainly not discounting what you are saying but would prefer to work from empirical evidence rather than claimed experience (of course for you it is not just claimed.)

    1. Just to be clear, we’ve checked this with Vivitek and we are OK. The Vivitek D535 bulb axis is parallel to the pitch axis and it does not care what angle it is pitched. The +/- 15 degree warning in the user manual is to prevent unmounted units from slipping and falling when on a desktop, etc. We are told to expect normal (~3000 hour) bulb life.

  9. Just to be clear about this, understand that you can get touchy feely models but nothing useful. I work (sometimes fortunately, mostly unfortunately) with commercial 3d printers and autodesk inventor 2013 day in and out.

    When you are trying to do something like tweak a snap feature, if you are not printing in an acceptable medium your model can look cool, but it won’t work period.

    For example on some dimension printers, we now know after much trial and error how to model and print a part on that in ABS that will translate to an injection molded part in nylon 6/6 with a UL94VO rating.

    It’s cool and all, but does not translate to the real world. Print up as many 3D crescent wrenches as you like, but my injection molded 3D crescent wrench will outperform your repcrap day in and day out.

    1. You seem to assume the world revolves around prototyping for injection moulding. Everything has limitations and pros and cons but to say you get nothing useful is short sighted. I’ll stick with my drop forged crescent wrenches thanks but will continue to use my 3d printer very effectively for prototyping. Oh and you might find it interesting to know that the parts produced on these dlp machines are higher res and more homogenous than the dimension stuff, you can also try different resins to get specific properties.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.