Are you sure it’s the tank and not the tank holder. I have never seen a tank that far off but I have seen 3 printers where the tank holder was that far off.
Pretty sure, because the tank itself is aligned with the platform and the edge of the printer.
Unless the tank holder, the edge of the printer and the platform are equally misaligned (which is entirely possible, although less probable) 
It’s no big deal, though.
It’s 1mm over the span of whole width, it doesn’t amount to too big a skew unless i were to print something with (flattish) surfaces spanning the whole width of the tank.
Wow I have never seen a tank anywhere near that bad off. I’d be interested to know if your other tank is better, when you get around to using it. As for 1mm difference across the build area that’s a lot to me.
Also how is it alighned with the platform and the edge of the printer?
I wouldn’t use the printer’s platform to confirm it’s the PDMS, I would use a caliper’s depth gauge to see how far it is from the top of the tank to the top of the PDMS. Checked on both sides of the tank.
I’m not using the platform. I’m using the platform, the tank and the printer surface.
Like i’ve said - it’s possible all three are misaligned the same way and the PDMS is level, but… That is practically the same as PDMS not being level, isn’t it
Not to mention way less probable.
I would be very interested in some more info on the tank with the PDMS that is 1mm off of level.
Do you have any 1-2-3 blocks and a height gauge? Doesn’t need to be a 1-2-3 block, just something that is a decent area so it does not deform the PDMS much, and has parallel top/bottom.
Put the tank on a flat surface, put the block in the corner, and measure with the height gauge. Zero on the first corner and record the measurements of each other corner. Avoid the very outer edge of the PDMS surface.
Thanks!
That “1mm” might have been an exaggeration, as i’ve just realized (while watching a print) that i’m seeing double the displacement due to reflections off the interfaces between the resin, PDMS and the vat.
I’ll see about measuring it when i empty the vat.
Don’t have a 1-2-3 block, but i’ll try and find something suitably flat.
In the meantime…
Did a print of a rook (supportless, upright) to test for curing issues.
Still on clear, so not that good for judging the surfaces and overcure, but useful to see any issues in the internal wall structure. It appears i do have some jello and “bubbling” “(hollows” like the ones @Steve_Johnstone had i think), but fairly limited.
Both occur on the same side (haven’t correlated with the direction of the flare yet, i set it to print at 4am or so, and took it off the platform just before going to bed - wasn’t thinking straight and forgot to check the orientation), but the jello is limited to the upper-facing surfaces, so it seems to be simply overshoot (fairly common occurrence with transparent resins). The “bubbling” does not affect the surface, it’s embedded in the wall, and is clustered around the bottom (vat) facing surface edge. It might have been due to blocked resin flow, as it’s right at the base of the rook (right where the wall starts).
BTW, i’ve printed it at 0.05 - it was my understanding that there are issues with clear at <0.1mm layer settings with the latest PreForm and firmware.
Umm that version of the rook suffers from blowout due to creating a closed cell in the base. You need a copy with a solid base otherwise you get holes drilled through. I can send you one that should print right if you like. That is likely the source of your bubbling and if it is it is normal behavior.
When you say jello what exactly do you mean? Do you mean a sticky slightly soft surface, or flakes of cured resin, or blobs of partially cured resin? If you mean slightly soft sticky surface that is normal. Otherwise that is not normal.
So it sounds like what you are doing is looking into the tank while printing in clear at the gap between the PDMS and the platform, and eyeballing the difference in the gap you see. First you should not see any real difference in the gap. Last I heard from Formlabs the tray housing and platform should be parallel with their current calibration techniques. If they are and the tank PDMS is level of course you will see no difference in the gap. Now in reference to your eyeball method I understand how you are seeing the difference in angle between the tank PDMS surface and the platform surface. However I do not understand how you are establishing the tank PDMS surface is not parallel with the printer surface. Anyway due to the platform being pushed into the PDMS the eyeball method is not great. To actually get a realistic view of the difference in angle print this @100um:
All four squares should be the same thickness. Mark them then measure them with calipers. Then adjust the platform height so they come out about 2mm thick. Then mark and measure again. This should give an idea of difference in distance for all four corners and help find the appropriate platform height for your tank.
The cell isn’t closed, it has a vent leading to the bottom as well as the “door” that goes all the way down to the beginning of the solid base.
Also, there was no blowout, but voids within the wall (not breaching the surface of the wall) as i’ve mentioned, and that occurred much higher up.
I’m eyeballing at the point (in time, during a print) where the platform separates from the PDMS. Re: printer surface - if you move your vantage point low enough, you can line up the edge of the printer with the top edge of the vat. Another method is - take a suitable object (ruler, for example) and put it at the front edge of the printer and align its opposite edge with the vat or the platform.
But yes, that is just eyeballing the difference. I’ll print the test squares as soon as i catch some free time, as i want to check for dimensional accuracy as well.
Uncured, gellish, deformed flake where a flat surface should be.
You can pick the flake off with a toothpick or tweezers.
In fact, if i held it in IPA longer, it probably would’ve fallen off on its own, as it was already half-hanging off the properly cured portion below.
Like i’ve said, it’s a fairly small bit and it’s where i’ve learned to expect it when using transparent resins - on the upward facing surface, where the laser overshoots the current layer(s) and cures too deeply (and partially).
Sorry I saw rook and Thingiverse and assumed you meant the original one used as a sample by Formlabs. The one you did doesn’t have that problem.
Try getting a surface level app for your phone. Use it on the front edge of the printer and the front edge of the tank. That will tell you roughly how parallel the printer surface is to the tray body, in a far more reliable fashion. Make sure to maintain the orientation of the surface level for most accurate results. For your line it up eyeball method to work you need to have the front edge of the table and the front face of the printer be parallel. Then you need to make sure your angle of observation is perpendicular to one of the axis represented by the horizontal top edge of your desk… Umm… so it’s not super reliable.
Both of these are bad and not standard behavior for the printer (and unacceptable behavior as far as I’m concerned). Do you have pics?
Also the squares only give a measurement of how far off of parallel the surface of the PDMS is from the surface of the platform, they are quite useful for learning the profile of the PDMS for a given tray which only tells part of the story. The method given by @Aaron_Silidker of course gives a rough measurement for how far off of parallel the PDMS surface is from the tray body, The surface level app or a good digital inclinometer can be used to measure how far off from parallel the printer surface, the tray body, the platform surface, or any number of other surfaces are from each other. However I would not suggest trying the inclinometer or surface level on the tray housing itself as it is fairly thin metal and is usually not flat so gives unreliable results. There is another model I will put up later for testing if your functional z-axis is perpendicular to the PDMS surface (the printing surface), and if it is perpendicular to the platform. These can tell you roughly what level of skewing you have.
I’m saddened and somewhat surprised that Formlabs might quite possibly have sent you a pretty crappy printer. 
I have been running on the assumption the majority are good, and therefore the likely case would be you would get a good one. I’m also really disturbed by the fact that your printer my be poorly aligned when support just reassured me they have adopted tight and precise tolerances and techniques to prevent that.
Anyway I don’t know how much you care about all of the measurements and details but anything you do measure and any results you get from prints, and other tests are good to know. I look forward to what results you get.
1 Like
@Ante_Vukorepa and @RocusHalbasch,
Functionally what is important here is that the PDMS is parallel within reason to the build platform across a range of build platforms and fresh tanks (not recoated), not that it is level or parallel to the rest of the machine.
Eyeballing the PDMS is not a great way to measure this because of the meniscus at the edges of the tank surface from PDMS/resin.
I would be very hesitant to trust results from the cell phone inclinometer. Cell phone IMUs are not the greatest. Assuming an error of 0.1 degrees, over the 125mm build area, you only have a resolution of at best roughly 200 microns.
What you need to measure is if the PDMS is level to the bottom of the tank. Using a known flat surface, you can measure different spots on the tank using a height gauge or indicator. Alternatively, you can use a calibrated level to check the surface you are putting the tank on, and the PDMS itself to see what the difference is between the two. Machinists levels are extremely accurate, easy to calibrate, and are not very expensive on eBay.
Lastly, if you are getting consistent adherence at all points on the build platform, and your parts are dimensionally correct within reason, then empirically you know that your PDMS to build platform relationship is correct.
1 Like
I’d like to stress that any claims i make re: levelness of PDMS are to be considered “eyeballed first impression” until i actually test this - i.e. until i print on all extremes of the vat - and/or until i measure a discrepancy directly.
Yeah this is true however I’ve found my platform is usually pretty close to parallel with the body of the machine. This may have been luck but it has consistently been my experience. I usually do measure the platform surface vs the tank body as my primary measure for how well aligned a given printer is. I have found that almost all cofigurations there after are quite close to that one measurement. I would do it with the tray housing instead of the tray body but as mentioned previously I find the housing to be way to far from flat to get reliable measurements on.
I agree it’s not ideal but it beats eyeballing by a long shot and those rough numbers can still clearly demonstrate a problem, and he would not need to empty his tank. I personally use a more precise digital inclinometer but most people don’t have those lying around.
Or alternatively you can measure the angle between the platform and the tank body, then use the results of the four squares print to get the angle between the PDMS and the platform, and calculate from the two the difference in angle between the tank body and the PDMS. This won’t be as precise but it’ll likely be more then sufficient if the angle is as bad as you described.
The first of these is only indicitive that things are not really really bad. Just the latter of the two should suffice to demonstrate that all is well, however it is often harder to demonstrate that than it is to take some key measurements.
Look forward to seeing the results of your testing. You vocal customers who dig into all aspects of the machines really do help us out here at HQ. We do plenty of testing ourselves, and more data points are always great.
The platform should be very close to parallel with the machine. That being said, there is a chance for tolerance stack-up to potentially lead someone awry if they were using this as an indication of a problem so I wanted to make it clear.
Cell phone is much better than eyeballing. Digital inclinometer is better than that. Nothing beats a good analog measuring tool. The 15" Starrett level I use indicates all the way down to .0028 degrees 
. We are very concerned with accuracy at Formlabs!
I think your test print with squares at the corners is a very good idea for an easy experiment to learn whether the PDMS and build platform are parallel. I am very interested to see this data, in particular!
Remember that by the time the support base is finished printing the surfaces will perfectly parallel anyway. This is because the first layers will print very thin on one side until things level out. All the layers that follow will be a consistent thickness.
So as long as you use the PreForm supports it’s all good.
@JoshK that is only partially correct, the layers will be a consistent thickness, however if the surface of the PDMS is not perpendicular to the z-axis you will still get skewing, unless it’s corrected for in the software by changing the layer height, and drifting the x/y position of each layer. The platform is usually quite close to perpendicular to the z-axis thanks to the linear motion system and a relatively well designed bracket, making sure the platform and tray are parallel is basically just to help ensure that the tray is close to perpendicular to the z-axis which is hard to test for.
Speaking of which here is a test for just that. Print the included .form file at 100um:
Note the arrow points toward the front of the printer. When the print is done turn it upside down so the arrow is on the bottom and put it side by side with an engineers square. Slowly rotate it and keep looking at how it lines up with the engineers square. If the side of the cylinder diverges from the side of the engineers square your z-axis is not perpendicular to the surface of the PDMS is not perpendicular to the z-axis, if you rotate the cylinder to maximize the amount of divergence the divergence is representative of the maximal amount of skewing, and you can use the arrow to figure out the direction of maximal skewing.
You can do this with the other side to get the angle of lean caused by the PDMS and the platform not being parallel, and what angle the lean is in.
Technically yes, but we are only talking about a 0.3° angle.