Am I using mesh leveling wrong?  

Re: Am I using mesh leveling wrong?

I'm starting to wonder about this as well. I too added wave springs to level out my bed, my reported variance is 0.021mm, and the past few days my first layers have gone to sh!^. I can't get a consistent first layer for anything, whereas before they used to be okay. I'm half tempted to pull the springs and go back to my ~.8mm out to see if my first layers improve...

Re: Am I using mesh leveling wrong?

You both seem to have a correctly leveled bed at the temp that you print. I wonder if your pindas are the variable that’s messing with you. Read in the forum that to be safe you should start your print with pinda at 35c. Latest firmware even added a gcode that you can use to wait for pinda to reach desired temp before printing. Worth a shot and see if starting to print once pinda is 35c helped with this

Re: Am I using mesh leveling wrong?

I have the exact same problem and it’s been driving me crazy. Regardless of how much I level the bed, the front of my bed always prints too squished and the back middle always prints too far away from the bed. Even though the bed leveling starts the opposite. It’s almost like it’s over correcting or something.

I use this code to make sure the Pinda is warmed up and the bed is set to 80 degrees.
G28 W ; home all without mesh bed level
G0 X50 Y50 Z0.15 ; this is a good PINDA heating position
M140 S80 ; set bed temp
M190 S80 ; wait for bed temp
M860 S35 ; wait until PINDA is >= 35C
G28 W ; home all without mesh bed level
G80 ; mesh bed leveling
G81; check mesh leveling results
M140 S0;


And I get these results

Re: Am I using mesh leveling wrong?

You both seem to have a correctly leveled bed at the temp that you print. I wonder if your pindas are the variable that’s messing with you. Read in the forum that to be safe you should start your print with pinda at 35c. Latest firmware even added a gcode that you can use to wait for pinda to reach desired temp before printing. Worth a shot and see if starting to print once pinda is 35c helped with this

I've had my Printer for around 5 months now and have been printing without issue. This is something that has popped up in the past couple of weeks. I'm now trying to remember when I installed the updated firmware in case there's a correlation with it. Also, at the same time my Z adjust has gone from -650 to -760 for me to get a decent first layer so there's another ?? in this.

Re: Am I using mesh leveling wrong?

I've had my Printer for around 5 months now and have been printing without issue. This is something that has popped up in the past couple of weeks. I'm now trying to remember when I installed the updated firmware in case there's a correlation with it. Also, at the same time my Z adjust has gone from -650 to -760 for me to get a decent first layer so there's another ?? in this.

I think the issue has always been present for me, just was using PETG primarily before now so I had to airprint a bit anyway and that really masked it. With ABS+ I need a good squish or get some warping on bigger prints.

Thinking back to before I did the wave springs and even the original disabled-LA series of firmwares, anytime I printed a full plate of ABS+ I'd get some failures on the left side and overextrusion on the right. Back then I was adjusting Z with only a centered 75mmx75mm square and just using a metal rule to get the bed tightened flat-ish by eye and didn't real know any better. I chalked up most of it to a nozzle that had a bad grind and swapped it, then went back to PETG.

Having this happen now again almost makes me thing whatever the pinda triggers on in the PCB isn't planar to the top surface or similar?

Re: Am I using mesh leveling wrong?

..... to fix the full width print issue I ended up having to set the bed level correct to -50 on the left and +15-20 on the right .....

You say that you've put some kind of springs under the bed to adjust the flatness and of course we know that the mesh leveling will adjust for what it measures. I wonder why one would also do firmware adjustments such as these? Doing this just makes things more confusing. I've never used any firmware adjustments aside from the Live Z.

I've done steel springs in the bed to improve flatness though I never really had any problem. The main difference is that now when the nozzle traverses the bed I cannot see any Z-axis movement because it is either adjusting very little or not at all.

back before I started with the steel springs under the 9 screws, I did adjust one of the Y axis rod mounting points to get rid of a low corner.

Re: Am I using mesh leveling wrong?

There are different versions of the pinda probe. Maybe the OP has the old version?

Re: Am I using mesh leveling wrong?

You say that you've put some kind of springs under the bed to adjust the flatness and of course we know that the mesh leveling will adjust for what it measures. I wonder why one would also do firmware adjustments such as these? Doing this just makes things more confusing. I've never used any firmware adjustments aside from the Live Z.

After flattening out the bed using the screws w/ springs on the standoffs underneath until the pinda/mesh output looks good, the live Z only adjusts well for the center of the bed. Left of center and it was like Z was too high and it didn't squish, right of center it was like Z was too low and it oversquished.

Its like the mesh isn't being applied to the actual print/first layer, and the only way to fix was to adjust the bed in the menu.

Re: Am I using mesh leveling wrong?

There are different versions of the pinda probe. Maybe the OP has the old version?

I've got the black tip one. Its consistently triggering about the same heights in the same places based on temp, sometimes it gets it wrong but usually its been doing okay. It just seems like the results its getting aren't reality when you compare the mesh visualization to the needed bed adjustments in the menu to make it consistent.

Re: Am I using mesh leveling wrong?

Installed 3.3.1 and re-calibrated everything using the 35C pinda pause until I got the mesh plugin to show .04 or better variance, and still getting the same result!

Left side needs negative adjust, right positive, and rear some negative too. Hard to get it all to show in one lighting angle but the left side is worse than the rear middle you can see through. After adjusting a few rounds of Z and the bed adjust all the test spots do get pretty acceptable but I'm just baffled why the mesh results are so off from what the probe "sees".

I'm reprinting the new extruder body now, I'm kind of thinking maybe I'll reprint the whole E axis and X ends while I'm at it and just do a full rebuild? Could something be skewed in my assembly so the probe and nozzle aren't both trammed to the bed?

Re: Am I using mesh leveling wrong?

You say that you've put some kind of springs under the bed to adjust the flatness and of course we know that the mesh leveling will adjust for what it measures. I wonder why one would also do firmware adjustments such as these? Doing this just makes things more confusing. I've never used any firmware adjustments aside from the Live Z.

After flattening out the bed using the screws w/ springs on the standoffs underneath until the pinda/mesh output looks good, the live Z only adjusts well for the center of the bed. Left of center and it was like Z was too high and it didn't squish, right of center it was like Z was too low and it oversquished.

Its like the mesh isn't being applied to the actual print/first layer, and the only way to fix was to adjust the bed in the menu.

I have the exact same problem.. I always thought I might be doing something wrong.
One more thing I noticed in my case is that I leveled the cold bed and got a variance of about 0,01-0,02mm. But the heated bed then had a variance of about 0,2mm 😯 😯 (Pinda always at 35°). BTW I got the grey Pinda probe.
I then adjusted the bed a 60°C and got a noticeable improvement but I still have to do the manual bed level correction to get an acceptable first level in all parts of the bed... I am really wondering if there might be a bug in the firmware if the variance is so low (just thinking out loud since the correction is not really what I would expect)?

Is there a way to disable the bed correction? That way we could at least check if the the pinda readings are correct because with such low variances I would expect really good results without any compensation in the firmware.

Re: Am I using mesh leveling wrong?

Co.me t out or remove the mesh bed levellung command in your start gcode

Regards Joan

Re: Am I using mesh leveling wrong?

I'm having this problem too.

I have tried everything but my first layer is high on the left and low on the right side of the bed.

One thing i have noticed is that after a Z calibration procedure (from the LCD menu) the X rods are not horizontal (measured to the motor mount and the ground) when near the bed.

The X rods seems horizontal when they touch the screews in the z top mount but after the home movment the left side becomes 1.2mm higher than the right side.

Do you have any idea?

Thx.

Re: Am I using mesh leveling wrong?

After many hours of testing with bed mesh correction values I eventually gave up on it, mainly because while I was able to get left/right/front/rear squares on a 9 square test, the corners were always out of whack. Another complication is how the center square would have to be corrected live-z after changing bed mesh values. I even created a model in 3d program and used it to measure that center offset after bed mesh value change.

I think the bed mesh algorithm should be a full 9 points or not at all, well 8 at least. In the end I used wave washers and that quickly got things the best so far. Unfortunately I used Loctite on those puny screws and now the heads are nicely stripped. Hex probably wasn't the wisest choice and I should have replaced them with a head type that isn't so prone to insta-stripping.

After some consideration, I wonder what exactly is the value of having a pinda probe. It seems to complicate matters for no real benefit other than possibly protecting inexperienced users from crashing into the bed.

Re: Am I using mesh leveling wrong?

Well I just finished a full rebuild for the new extruder cooling and did a full E/X/Z axis reprint in eSun ABS+ instead of PETG, and need the same exact bed adjustment levels as before to get a good first layer. At least its consistent...

I noticed the tip of my pinda isn't square to the body, its got a slant to it. Does that matter? Figured its just a molded cap over the electronics which I'm hoping are actually squared up.

Re: Am I using mesh leveling wrong?

I have exact same issue with the brand new MK3 assembled a months ago. The bed at left side is way too low and right side it way too high. the first level layer at left side has less then 3rd of the width compared to right side. feels like mesh bed levelling is not apply any compensation at all. tried wave washer to reduce the most of 9 point variation down to 0.05 etc. but the issue remain. also try to modify the front right Y carriage rod holder to reduced the height of the rear right corner. Did not do much either. tried disabled G80, then end with scratch at right side of the PEI bed badly..... Only the tiny area at centre is print able...the only option CLOSE (far from ok) to has a stick able first level at left side is mesh correction at left -50, right +20 and lower the Z further. Try contact support, they only respond email once per week.....been there for month already sent picture and video. haven't gone very far yet...

Re: Am I using mesh leveling wrong?

I have exact same issue with the brand new MK3 assembled a months ago.

What this sort of post continues to make me think is that the Pinda is working against us in trying to level things out.

Re: Am I using mesh leveling wrong?

I have exact same issue with the brand new MK3 assembled a months ago.

What this sort of post continues to make me think is that the Pinda is working against us in trying to level things out.

I achieved better results by printing your three links than if I use the mesh leveling tool with a heated bed and warmed up Pinda.

It would be nice to figure out reliable Pinda leveling across the bed though because the spring steel sheets aren’t perfectly flat.

Re: Am I using mesh leveling wrong?

I have exact same issue with the brand new MK3 assembled a months ago.

What this sort of post continues to make me think is that the Pinda is working against us in trying to level things out.

I feel the same thing. I run a lot of G80, G81 under PINDA temperature way under 35c. the result is consistence. I am suspecting the heat bed or/and Y carriage are warped/uneven in area between those 9 calibration point. PINDA won't see it. So the predicated compensation mesh leaving applied is not what it needed.

I may completely wrong on this. I am not sure if XYZ calibration have anything to do with it as well. if it set some kind of base bed geometry before mesh apply on top of it. The reason I think it is, I noticed the left X rod holder won't go all the way up to the top during Z calibration due to 2 screw on the gantry are block it for around 1 mm. I cut few plastic pieces off. it make to the top. and I can see the different in calibration result. But it did not make much difference in print. additionally, the XYZ calibration is rely one the 9 embedded point in PCB, if it is not sit straight inside. it may affect the calibration result. Then the mesh level is run when steel sheet is on. I tested the steel sheet will trigger the PINDA . with those 9 point in bed out of picture. it's different reference point to the XYZ calibration. who know how Steel sheet even magnet field will affect PINDA. if any of those 2 method predicted the wrong levelling of be bed. then the finally result will be way off even the wrong way completely.