How flat should my heatbed be?  

Re: How flat should my heatbed be?

ive been fighting for two days with mine. ive done just about everything i know to do but cannot get it level. my first layer looks terrible. im pretty sure the bed is warped and no amount of adjusting is going to fix it. i think im going to contact support and see what they say. the downside to all this is my back right screw is stripped and i can not longer remove it. any ideas on how i can get it out?

ciao,
mile.

Re: How flat should my heatbed be?

Welcome to the MK3 beta test. I had the same with my first and second MK3 bed neither are flat both have a dip in the middle and a rise in the right hand corner. What helps a lot are the TPU Micro springs and the Octoprint PrushMeshMap Plugin here are the links https://shop.prusa3d.com/forum/others-archive--f66/-testers-needed-tpu-micro-springs--t15061.html#p83784 and https://shop.prusa3d.com/forum/user-mods-octoprint-enclosures-nozzles--f65/-help-needed-octoprint-prushmeshmap-plugin-t15333.html#p85988

I had to put 2 springs in the right hand corner 1 in the middle and 1 on the left hand side middle for both beds. Just put the screws in snug then run the bed level plugin and then what ever is yellow slowly tighten a little at a time (very little adjustment goes a long way just take your time you do not want to squish the springs completely or you will have to replace it and start over.After that if you need it you can adjust the mesh bed leveling in software

Now getting the first layer down correctly the PINDA is temperature sensitive so you want to install the latest firmware put in the gcode that waits for the correct temp in this thread and follow the instructions to manually calibrate your live -z layer between (0.800-0.950) at 35c, 40c, 45c, 50c 55c and 60c here https://shop.prusa3d.com/forum/-f0/1st-layer-problems-in-depth-look-at-software-pinda-t14815.html#unread

Re: How flat should my heatbed be?

do you have to print the micro springs in tpu? i just bought some but wasnt planning to print until i reolved the bed issue. its kinda a catch 22 though.

did you adjust at all using the bed level micro calibration? my live z is at -1.15. i think the right needs to go up+20 and the left down -20. i have scrapping on the right but then evens out in the center and has a few gaps on the left. this is the most frustrating thing. do i need to raise my pinda since its above -1? ive read multiple peoples account on this topic and its not clear to me if it matters or not.

my variance on the prusa plugin is .6. pretty big difference. it goes from under .1 on the left to .6 on the right. i even tried to shave down the y axis rod holders. no luck.

any idea on how to remove a stripped screw? at this point i cannot get the bed off. not sure if i have to drill it out or what.

thanks!

Re: How flat should my heatbed be?

I had an issue with my first layer level that required putting in manual calibration. See pics:
https://shop.prusa3d.com/forum/hardware-firmware-and-software-help-f64/i-have-to-increase-nozzle-temp-10c-to-get-supplied-t18629-s10.html

At work the fitters use a drill bit called easyout.

Re: How flat should my heatbed be?

one thing that is getting overlooked is that the bed rids on two rods that ideally would be perfectly parallel.

The bed device itself could be perfectly flat but if the rods are not parallel it will move in a way that presents a warped surface relative to the machine.

In other words - first check the Y axis rods.

I checked mine by putting the machine on a granite surface plate (QC department at my company has one) and checked the height at 4 corners. one corner was an outlier so I designed a different mounting bracket for the rear of my right-hand Y axis rod. The same can be done with simply cutting some material or adding a shim.

After this adjustment the bed shows up flat in the self-test results although it will have a minor humped shape as it heats.

I have never had real 1st layer problems and I only did this investigation because I had the tools available and was curious about it.

Re: How flat should my heatbed be?

I found about .3 mm difference in the bed (I'm assuming the units are mm), mostly in one corner.
I will try some shims, but I wonder about what's causing it
1) Differences in the spacer thicknesses. Seems unlikely, but not impossible
2) Differences in bed countersink depth. Also possible, but not likely. PCBs are made very precisely
3) Differences in PCB thickness. Extremely unlikely
4) Intrinsic non-flatness in the carriage piece. Seems very plausible
5) Non-flatness due to forces from the rods. Also plausible (likely?), but there are a lot of places to loosen/tighten bolts

The shims don't care where the error is coming from, but I'm not looking forward to installing them.

Re: How flat should my heatbed be?

Re: How flat should my heatbed be?

I found about .3 mm difference in the bed (I'm assuming the units are mm), mostly in one corner.
I will try some shims, but I wonder about what's causing it
1) Differences in the spacer thicknesses. Seems unlikely, but not impossible
2) Differences in bed countersink depth. Also possible, but not likely. PCBs are made very precisely
3) Differences in PCB thickness. Extremely unlikely
4) Intrinsic non-flatness in the carriage piece. Seems very plausible
5) Non-flatness due to forces from the rods. Also plausible (likely?), but there are a lot of places to loosen/tighten bolts

The shims don't care where the error is coming from, but I'm not looking forward to installing them.

6) one of one of the ends of the rods is too high/ too low. This is a very easy place to move a single point and change the flatness.
- this can easily happen due to an accumulation of tolerances

The rods are only held with zip ties so if you have a spare zip tie this can be tested by putting a piece of paper between the end of the Y axis rod and the saddle that it sits in - zip tie it back down. In my case the rod needed to be lowered so I had to print a new part but in a pinch, if you are confident some plastic could be cut away to lower the rod.

Re: How flat should my heatbed be?

Agreed on the rods. What's potentially confusing is that the bed is moving back and forth during calibration (and printing too of course).
So, if G80/G81 reports that the back left corner of the bed is low, I suspect that means that front end of the left rod is low since the bed is all the way forward when the back edge is checked. The other end of the rod matters too, but the bearings are closer to the front when the pinda is checking the back.
Does that make sense?

Re: How flat should my heatbed be?

What matters is the height between the nozzle and the heated bed. Therefore, I'm pretty sure the bed leveling accounts for a difference in height caused by the rods because it shouldn't matter what the cause is as long as it's repeatable.

Re: How flat should my heatbed be?

Agreed on the rods. What's potentially confusing is that the bed is moving back and forth during calibration (and printing too of course).
So, if G80/G81 reports that the back left corner of the bed is low, I suspect that means that front end of the left rod is low since the bed is all the way forward when the back edge is checked. The other end of the rod matters too, but the bearings are closer to the front when the pinda is checking the back.
Does that make sense?

correct. and it does make sense : ) In that case I would have either raised the front of the LH Y axis rod or else lower the back of the RH Y axis rod to twist the entire reference frame.

I made my adjustment on the Right hand side where there is only one bearing. I'm not 100% sure how it would have worked if my reports had shown the outlier data point to be on the left side. Ideally one would want to know which of the two rods is correct (meaning it is perpendicular to the Z axis) but this is very hard to measure.

as I stated earlier I think this correction means is relatively easy to test and verify by using shims, zip ties and G80/ G81.

one more thing - it is best to take care how the feet are laid out and what surface they are on because the frame has the potential to flex and change things a bit. I have my printer sitting on a thick concrete paver that I picked up at Home Depot (a garden center).