Also, I'm not sure if you have seen this: http://hackaday.com/2013/04/26/cube-...bulk-filament/ or if it is even valid on the 3rd generation?
Printable View
Also, I'm not sure if you have seen this: http://hackaday.com/2013/04/26/cube-...bulk-filament/ or if it is even valid on the 3rd generation?
The Cube 3rd gen is a rather polished piece of consumer equipment. It certainly was designed well, with fully enclosed mechanicals, contoured plastic panels and LED workspace illumination. Dual extruders and closed source everything. There's a USB port for the flash drive (A?) and an unused USB printer port (B?) that is marked in the manual as "future use." With the general consensus that 3D Systems is giving up on this item, I doubt there's any future use involved.
Ahhh, has anyone done a tear down? Once you tear it down, and figure out the controller and can replace it you can probably do anything with it. Seems like there are some hacks (not sure if they still work as there was talk about Cube doing firmware updates to prevent them from being used).
I don't know anyone else with a Cube 3rd gen, and the forums for this item seem a bit quiet. The older models have some hacks but nothing new on the Cube 3, which is mostly the latest consumer grade printer.
The Pegasus never really made it to consumer grade, did it? It was a Kickstarter hack job that's lost the following it developed when it first appeared.
Yeah I don't think the Pegasus really made a name for itself mainly because of all the trouble people had during the kickstarter and I think people don't really want proprietary printers (materials/resin/filament/etc) these days for many reasons. I wouldn't buy it again today. I wouldn't even consider it. I *might* buy one from someone if it were really cheap and I get my code working well enough that I could use my software on it.
I have a continuing search/email trace on ebay for this printer and have seen prices as high as KS level and more recently, down to US$1600, which is still rather high for a printer with such a poor track record.
The Gizmo line of printers look promising, but I was burned with this KS campaign, so I'm not going to invest in that one.
If someone had developed an ideal release surface, do you think this printer becomes viable again?
My Pegasus only shined when I started using non-FSL resin. Here are a few prints I managed to get out after figuring out the right recipe for success:
Oblivion drone:
https://groups.google.com/forum/#!se...A/EQ1HSe2UK4IJ
Human skull:
https://groups.google.com/forum/#!se...g/xeK77-Y_AWEJ
Arc reactor:
https://groups.google.com/forum/#!se...s/SYLqXtaEi40J
My printer still actually works today, it's just that I have to use PDMS coated vats.
Also, I'd like to refine and take out all the shortcomings of the printer that I need to compensate for when thinking about using it.
Cooch, I see you didn't specify which non-FLS resin you've used. Was it a matter of luck that it works with the Pegasus? Is the stuff you're using less expensive than the name-brand resin?
I used Madesolid's MS V1 black resin... it was $100 a liter on Amazon, which is less than FSL's. Not sure Madesolid still offers it, as they have had MS v2 for awhile and now they have Vorex resin, which is about the same price as FSL, but the Vorex is probably stronger.
Cool. I wasn't aware other resins worked in the printer! I'll maybe have to try it at some point. It would be nice if they gave us better control over calibration. For my Z axis I have four parameters I've set up for the lift: Speed0, Speed1, Speed2, SpeedL speed. You program the speeds for how fast or slow you want each step to be, on a lift it'll do speed0, speed1, speed2, then you can have it pause, then it'll lower at SpeedL. You can also control how far it moves up for each of the raise speeds. The lower distance is based on how far it moved up and where it goes on the down.
I'm trying to figure out how to calibrate it. I ripped up my PDMS on one of my prints; I'm not sure if I moved it too fast or too slow, I'm really hoping I didn't screw up the VAT itself. I actually tore through the PDMS coating to the VAT, so there may be stuck on resin in that area, which I'm not yet prepared to try to replace yet.
Was playing around trying to work on my GUI last night and created a nice looking calibration feature for the Z axis. It works for me, but might be a little confusing. I'll try to make a video to get some suggestions. I enabled some flexibility in adjustments beyond what FSL provides. They have a fixed 5micro per step, I enabled it so that you can adjust that slightly by using how many steps per MM (200=5 microns). In the process of trying to develop it and using some "precision" measurement tools I found that I am not controlling the Z axis well. And possibly this is why I destroyed my PDMS layer. I need to make some modifications to how I move the stepper so that I take into account the acceleration and deceleration of the stepper and build plate. I need to found info on doing this and coded it up, but haven't had time to try it out. I need to figure out how to fix my VAT as I don't have a warm enough location and probably won't for at least a month or two (likely the stable temp will be between 40 and 60). I also need to clean it a bit better because it is very messy even after wiping it down with IPA.
I also know my calibration on the x/y galvo isn't really correct either, that one will take a bit more effort, but I think I should be able to get prints with what I have regardless of that calibration.
I've been unable to get my GUI to work while printing, but I have a command line method for running the galvo. I'm pondering attempting to enable FSL software on my BBB to see if I can do anything with it. It would greatly speed up comparisons/debug if I could.
I had read that cleaning the PDMS type vat results in cloudiness and possibly other problems.
If your ambient temperature is a problem, have you considered to start up your cooking oven at the lowest temperature, placing the vat inside after powering the oven off? Obviously, you'd want to do so after the oven reached a reasonable temperature to avoid damaging the vat. Once the oven has warmed and then cooled to a reasonable level, it would retain enough warmth to combat cold, I would expect. I'd try that as long as my wife was not home, of course!
My original Z axis code was off by 20micron, oops! That is A LOT! And I think why I destroyed my vat and possibly why I was smushing the print. My new updated code is off by 1 micron; better, and maybe within a margin of error of my measuring tools. I was basically moving up and down 1cm at a time and measuring how much it really moved. I need to order material to redo my vat and then I'll probably try another print.
I'm having issues with github, but if I can work through it by the weekend I plan on releasing my first code stream to it. I have one change I want to make to the code to maybe make things work better. Right now I fill my buffer with about 8000 points before sending it to the printer, I think I want to change it to just doing each line from the gcode. It'll make it smoother and it'll look more like the FSL's print code.
Ugg, software isn't working today. This is what I get for getting busy and letting things sit. I have no idea why it would work worse today than it did last time...but that's software. Laser isn't running properly. Ends up being spotty looking and periodically stays on when it should turn off.
Ok, fixed things up and pushed all my latest code out to the githib repo. https://github.com/doobie42/pegasus3dprinter/
I don't recommend using the FSL BBB, grab a different one. Don't plug the 24v supply into the BBB. Put tape or glue in it so you don't do it.
I need to get around to recoating my VAT and I'll hopefully be able to get printing again. That is unless my last attempt got resin on the vat bottom and I cannot remove it or I goofed and scratch it; I didn't realize how touchy this all way for finger prints/dust/scratches.
Anyone good at mathematics? I need to come up with a formula to go from the X/Y MM position to the galvo position? I'll basically have at least 25 points defined, similar as the FSL3D calibration, which I'll use to transform from Xmm x Ymm to Xpos x Ypos. It's definitely not linear, close to linear in the center, but not as you move to the edges.
Do you have a diagram to better explain your objective? I enjoy math, trig, geometry, the whole package, but the word problems can be challenging. If it is what I think you mean, it's very much trigonometry and relatively simple math as a result. I have a machinist friend who has effectively memorized a good number of such functions and can almost do the math in his head. The drawing to represent the goal is the key.
I don't have a diagram, but if you have the FSL3d Calibration page it has 25 points, imagine each of those 25 points has an X,Y associated with it. The values are roughly ~325 per mm.
so maybe something like this:
0,0 330,5 655,10 980,5 1120,0
0,320 335,315 650,310 985,315 1115,320
....
etc
Not sure how well this will come out.
But each dot is about 45mm apart, so depending on where you are it'll transform into the above matrix, so 135,0 would be 980,5, 90,45 would be 335,315.
Maybe this link will help a bit:
https://docs.google.com/spreadsheets...it?usp=sharing
The projection of the laser isn't linear, but more spherical based on how the laser bounces off the mirror. I'm going to work on a few gcode files to test my calibration to see how close things are with different algorithms, once I have a few formulas or ideas to try. It's physically possible because FSL3D does it, so just need to figure out this transformation. If we had more points it'd be more accurate, I'm wondering what FSL3D does to make it accurate. I'll have to look into how laser projection systems work to keep their images scaled properly.
I had hoped to be able to create an image representing your objective in a pseudo-3d manner. I picture a calibration grid projected on the ceiling of a room. The last mirror in the beam travel is the zero point of each triangle. You would need one very precise measurement from the mirror reflection point to any point on the grid. You already know the travel of the beam on the calibration grid, so you have two adjacent triangle edges. I suspect it is possible to determine via code/experimentation the angle created between the first known length measurement and the second point on the grid. Now there's three pieces of data for the triangle. All other calculations can be done with any number of triangle calculators available on the internet. I like this one, as it requires less thought and much less demand on my brain cells: http://www.1728.org/trig4.htm
How does this fit in with your direction?
It might, I'm not sure. I found this link which is interesting:
http://www.open-electronics.org/how-...ampaign=buffer
"In addition, in the printers using galvanometers, if the size of area to be built is very large, an issue is introduced, due to ovality of the spot of the laser beam in the extreme parts, this ‘aberration’ is to be corrected by the manufacturers through software artifices."
Not entirely sure how to properly correct for it. Going to try to ask around.
I would be very surprised to learn that there is much of an ovality of the spot for the Pegasus Touch printer. My Emblaser laser cutter when well focused provides for a 0.05 mm cut. That would indicate the the laser dot is smaller, as the burn-off of material provides for the increased width of cut. For sake of argument, we could consider that my laser presents a 0.03 spot. Even extending it at an angle sufficient to elongate such a dot by a factor of two, that would mean a 0.06 spot. As the laser cutter creates burn-off increase, so would a resin laser present a burn-on increase in resin, I suppose.
I don't think it's severe, if it's even measurable.
That is very possible the ovality isn't a big deal, but the position is definitely an issue, moving the galvo 500 points in the center is very different that on the outside. I think I have an idea of what to do, just need to figure out how to do it.
Are you able to calculate or measure the angle of deflection/change of the final galvo? For the sake of argument, I'd prefer to consider that you're objective is to calibrate specific points on a line. That would mean one galvo does not move, while the other moves in a precise manner. I'd also expect that with appropriate trial and error, one could calculate the distance from the mirror to the vat. I'll make a guess that you can do the mirror movement and calculate angles, otherwise my method vanishes instantly.
Another aspect of this task is to determine the three inputs needed for the calculator online. You can get your "top" distance on the calibration paper and the angle of the galvo movement from the electronics (?) but the third measurement appears to be a tricky one. Can you think of a way to measure the distance from the mirror surface to the vat? Are there indications that at a specific point, the laser is directly below the vat and that the beam would form a right angle to the vat? That would provide a second angle to the calculation, rather than a second distance but any three works.
Another thought is that the angle of the laser movement through an arc on the vat creates a chord of a circle, but at that point, the math goes well beyond my comprehension.
I don't have a good way to measure the angle. I'm trying to under stand how the FSL3D software and calibration does it. They user that 25 PT grid to solve how calibrate and align the points.
Does the method of moving the mirror involve commanding the stepper motor to move a specific number of steps? I'd expect that the grid method takes into account the number of steps executed for movement from one point on the grid to another and calculates an adjustment factor for any movement within the box described by four points.
Not a steeper motor. It is a galvo. It uses a voltage to position itself. I can position it anywhere, but just not accurately yet, unless I move it manually to a specific spot.
ah, that does complicate things, doesn't it? I'd been seeing that term all this time and never made the proper connection in my alleged mind. Is there a specific value one can assign to a deflection measured in degrees?
ah, that does complicate things, doesn't it? I'd been seeing that term all this time and never made the proper connection in my alleged mind. Is there a specific value one can assign to a deflection measured in degrees? I found this link:
https://www.thorlabs.com/newgrouppag...tgroup_id=3770
that indicates these devices have an output of sorts, but the tech escapes me.
I'll take a look at that later. I have a friend who uses galvos for laser shows, I'll try to talk to him about it. Of course he doesn't need it super accurate, but on theory should be close.
My VAT looks maybe dead. Mighty be good enough to keep playing with, but I accidentally tore the silicon with my last print and have resin on the acrylic VAT. Washed most of it off but not sure if I maybe scratched it or not. Using IPA to clean it then will recoat.
Nah, not recently, have not yet had the chance to re coat my VAT. I've got two things on my To-do list though. I'm still worried my z axis code might lower too much even after calibration.
thanks so much
Ahhh been so busy, newborns and all :) Anyway I saw this guy: http://hackaday.com/2016/06/13/open-...m-the-browser/ and I think I'll have to check it out. I'm hoping that in the next month I'll find the time to recoat my vat and start making progress again.
Children and hobbies? How can you manage?
This might help us out here:
http://hackaday.com/2016/07/21/formlabs-form-1-api-now-available-on-github/
Cool, will have to check it out. I *finally* recoated my vat this morning, so I guess in 24-48 hours I can try to see how well the coating came out. I have some time this week and next so maybe I'll make some progress or at least get some videos of what my "problem" is.