Need to re-calibration first layer for different materials
 

Need to re-calibration first layer for different materials  

  RSS
boxerlc1988
(@boxerlc1988)
Active Member

I have assembled my MK3s for several weeks and has been using it to print a lot of PLA and PETG materials without any problem. Yesterday I got a spool of black eSUN ABS. After a few prints I noticed the problem with first layer calibration:

If I use PLA to do the first layer calibration then print with ABS, the first layer of ABS becomes so thin that it would start peeling.

If I use ABS to do the first layer calibration then print with PLA, the first layer of PLA becomes so thick that adjacent wires won't stick with each other.

My current solution is re-calibrate first layer for different material. It is so inconvenient. Does anybody have the same problem? What is the root cause of the problem? I'm guessing its the distance sensor's value lacks temperature compensation, what do you guys think? Do I need to get new hardware or there would be a firmware fix for this issue?

The firmware I'm using is 3.7.2.

 

Thanks

Posted : 08/08/2019 4:20 am
BillCampbell
(@billcampbell)
Reputable Member

Find which filament gives you the largest negative number (lowest z - PLA) then calculate the difference required and put that difference figure in as a z-offset in PrusaSlicer in a printer profile specific to the different filament (ABS).

 

Bill
Tagaytay City, Philippines
Founder member of Philippines Prusa Printer Owners FB Group
Sponsor Pillars of God Academy in Bacoor...
Posted : 09/08/2019 4:10 am
RAHRAH
(@big-bird)
Estimable Member
Posted by: boxerlc1988

I have assembled my MK3s for several weeks and has been using it to print a lot of PLA and PETG materials without any problem. Yesterday I got a spool of black eSUN ABS. After a few prints I noticed the problem with first layer calibration:

If I use PLA to do the first layer calibration then print with ABS, the first layer of ABS becomes so thin that it would start peeling.

If I use ABS to do the first layer calibration then print with PLA, the first layer of PLA becomes so thick that adjacent wires won't stick with each other.

My current solution is re-calibrate first layer for different material. It is so inconvenient. Does anybody have the same problem? What is the root cause of the problem? I'm guessing its the distance sensor's value lacks temperature compensation, what do you guys think? Do I need to get new hardware or there would be a firmware fix for this issue?

The firmware I'm using is 3.7.2.

 

Thanks

Bill is correct and for each filament you should do the calibration and set the compensation in the filament settings.  Using Prusa Slicer set up one configuration for each known and tested filament.  The offset will correct to get the best from each filament when printing.  Just make sure the filament type includes details about the filament brand and color in the GCODE file name.  The reason for this is that there are even differences between batches, colors, and filament lines that may be serious enough to account for in your prints.  It really depends on the types of things you are printing also.  More detail, (AKA more retractions), equals more scrutiny on the filament calibration.

That's my 2 cents worth since I am still learning.

Robin

This post was modified 12 months ago by RAHRAH
I am the inveterate tinkerer. I can tink up most anything....
Posted : 09/08/2019 4:49 am
RAHRAH
(@big-bird)
Estimable Member

Also,  Make sure you set the original Z offset for each side of you metal bed sheets especially if you use Powder coated and PEI sheets.  The Z offset WILL differ between sheets and especially between PC and PEI sheets.

Robin

I am the inveterate tinkerer. I can tink up most anything....
Posted : 09/08/2019 4:58 am
boxerlc1988
(@boxerlc1988)
Active Member

Thanks guys for your answers. I tried contact Prusa's customer support, they asked me to do many tests to rule out any potential assembly issues. It turn out my printer is assembled perfectly fine.

The last thing I was ask to do seems worked, which is doing a factory reset then going through wizard routine again. After doing first layer calibration with PLA, I printed a ABS part with a pretty good first layer. I don't know if this condition is temporary and I still don't know the root cause of the problem, I will have my fingers crossed.

My thought on Bill's workaround:

First I want to say thank you and I can image that it would be very effective, but it doesn't look like a ideal solution for me.

I think the layer height is only directly affected by the distance between the nozzle and the plate, no matter what filament is used. It's not the filament that properties that cause the first layers to have different height, it's the different temperatures for the nozzle or the heated-bed required by different filament that some how affect the reading of the PINDA sensor. I know there is temperature compensation inside the sensor or the firmware, but the compensation might not be enough, or it might be corrupted by some g-code or operation. So I think this issue need to be resolved by the firmware(provide better compensation) or hardware(make sensor less sensitive to temperature). 

Posted : 09/08/2019 3:13 pm
Tim
(@tim-m30)
Illustrious Member

I think the common method is to place a sheet, find the proper Live-Z for the filament on that specific sheet, and jot down the value. So the sheet has a list as follows:

PLA = -0.875
PETG -0.860
ABS= -0.840

And the flip side of that sheet might have slightly different numbers due to that side having a different lot of PEI adhesive.

The PC sheet would have even different numbers since each side would be coated with a different pass of the powder coating spray nozzle.

Thus, if you have six print sheets, you have 12x3 sets of Live-Z values to track.

 

This is not something I would ever expect the printer (as designed) to track.  There needs to be a serialized sheet database the printer can read as sheets are installed so the printer can set the appropriate Live-Z values.  Which is a great idea for a future printer and sheet design. RFID tags on each sheet for ID, the printer reads then, adjusts live-z on the fly no user intervention needed ... cool!

This post was modified 12 months ago 4 times by Tim
It is always wise to get more than one opinion......
Posted : 09/08/2019 6:09 pm
boxerlc1988
(@boxerlc1988)
Active Member
Posted by: Tim

I think the common method is to place a sheet, find the proper Live-Z for the filament on that specific sheet, and jot down the value. So the sheet has a list as follows:

PLA = -0.875
PETG -0.860
ABS= -0.840

And the flip side of that sheet might have slightly different numbers due to that side having a different lot of PEI adhesive.

The PC sheet would have even different numbers since each side would be coated with a different pass of the powder coating spray nozzle.

Thus, if you have six print sheets, you have 12x3 sets of Live-Z values to track.

 

This is not something I would ever expect the printer (as designed) to track.  There needs to be a serialized sheet database the printer can read as sheets are installed so the printer can set the appropriate Live-Z values.  Which is a great idea for a future printer and sheet design. RFID tags on each sheet for ID, the printer reads then, adjusts live-z on the fly no user intervention needed ... cool!

I think your thought is very possible, if this is the case, then it's hard to solve by calibration. Your solution is very interesting too.

Posted : 09/08/2019 7:18 pm
JadedRedDragon
(@jadedreddragon)
Eminent Member

I think @tim-m30 's thought is partially correct. Each sheet would have its own live-z value, but each side of the sheet wouldn't. Flipping the sheet over would maintain the same thickness of the sheet. It's held down by magnets so you wouldn't have to worry about warping.

Posted : 12/09/2019 9:50 pm
david.a66
(@david-a66)
Honorable Member

actually tim is correct, although the overall sheet thickness is the some on both sides the location of the steel sheet relative to the 2 coatings is likely different so the pinda which measures from the steel sheet will be at different heights on each side

Posted : 12/09/2019 10:17 pm
JadedRedDragon
(@jadedreddragon)
Eminent Member
Posted by: david.a66

actually tim is correct, although the overall sheet thickness is the some on both sides the location of the steel sheet relative to the 2 coatings is likely different so the pinda which measures from the steel sheet will be at different heights on each side

Oh yeah, I didn't think of that. I keep forgetting it doesn't measure the distance to the surface. It measures to the actual points in the sheet.

So, something I was thinking of. As long as you don't move the printer couldn't you run the 7x7 mesh one time then set it so it doesn't redo it every time? That would prevent the printer from throwing itself off due to PINDA temperature interference and/or if you flip the sheet over.

Also, I would think the Z height would be the same for all filaments unless you're using an exotic filament. Temperature could play a role in it because it could change the viscosity, but wouldn't you just need to change your temperature then?

This post was modified 11 months ago 2 times by JadedRedDragon
Posted : 13/09/2019 4:38 pm
bobstro
(@bobstro)
Illustrious Member
Posted by: JadedRedDragon
So, something I was thinking of. As long as you don't move the printer couldn't you run the 7x7 mesh one time then set it so it doesn't redo it every time? That would prevent the printer from throwing itself off due to PINDA temperature interference and/or if you flip the sheet over.
The sheet is both removable and flexible, so there can be significant differences in the surface height between prints. I've got the Prusa mesh leveling installed on Octoprint and have watched the surface of my bed vary between prints. Doing it every print might be a bit of overkill, but then you're about to spend hours on a print using expensive filament, so 45 seconds of insurance is a good idea. I've quit worrying about trying to "save time" up front, only to have to repeat a large print due to something fundamental that I skipped.

Also, I would think the Z height would be the same for all filaments unless you're using an exotic filament. Temperature could play a role in it because it could change the viscosity, but wouldn't you just need to change your temperature then?

PETG tends to like less squish than PLA or it'll pull up and stick to the nozzle. I've seen some variation even between the same types of filament. Not always anything significant, but enough to tweak the Live-Z height a bit. Definitely an "it depends" answer.

My notes and disclaimers on 3D printing and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan...
Posted : 13/09/2019 4:56 pm
Share:

Please Login or Register