RE: Z axis trouble but only when printing
Not sure if anyone else has posted this but outside of adding the gcode I have had some luck with lowering the idle temperature down to around 120c in the filament settings. After reading through how as others have said the issue seems to be related to the heating being on and interfering with the load cell, turning down the idle temp seems to have reduced the noise enough to get some good readings. Ambient is 80F where I am so you might need to lower it further based on chamber temp.
How did you do this? Do I have to create a new filament type for PLA for that to work, or do I do that in PrusaSlicer?
RE: Z axis trouble but only when printing
Just as a point of information, the wire twisting unfortunately did not work for me. Not that it isn't a good idea. As Prusa have said, there seem to be several possible causes and so there will be different fixes (or so I continue to hope). The wires are red and black on my old cable... the new cable is mostly black with a few white wires.
Maybe I should have put more twists into it (and I could have spaced them more evenly) but I didn't want to shorten those wires too much relative to the others. Back to the modified starting G-code for now... cable swap later (still in the box in the background).
RE: Z axis trouble but only when printing
Out of curiosity did you twist 1/2 and 3/4 or 1/3 and 2/4. The former won’t help as 3 and 4 are ground.
RE:
@David H. Brown: It is a little hard to tell from your photo, but it looks like the twisting isn't right. In the photo I see two red wires twisted together and two black wires twisted together, and it looks like the reds are both HEATER_24V and the blacks are both HEATER_GND. If I'm right about the signals in your picture, that twisting won't help reduce crosstalk.
You actually want to twist together one HEATER_24V wire (red?) together with one HEATER_GND wire (black?), and then do the same with the other red/black pair for these signals. That way the current flowing in the two wires in a twisted pair will always go in opposite directions. This is what minimizes the crosstalk. After making the two red/black twisted pairs, I'd also twist them together loosely, in case one of the HEATER_24V wires carries a little more current than the other.
Based on the Love board schematics, the extruder harnesss connector is J13. On this connector, HEATER_GND is on pins 16 and 18 and HEATER_24V is on pins 20 and 22. It is J5 of the xBuddy board on page 11. Here's a screenshot:
Love board schematic is at:
https://www.prusa3d.com/downloads/Electronics_drawings/FDM-MK4-LoveBoard-38.pdf
xBuddy schematic, page 11, top right (A5 coords) has the connector for the extruder harness.
https://www.prusa3d.com/downloads/Electronics_drawings/FDM-xBUDDY-44.pdf
RE: Z axis trouble but only when printing
Hello,
Do you think firmware 6.4.0RC fixes this issue? I can't find any mention of support for this specific issue.
RE: Z axis trouble but only when printing
I solved this problem.
After chatting with Prusa, I received a new heatsink to try out. I didn't have high hopes, and sure enough, it didn't solve my Z-axis trouble.
So I decided to go further, even if it meant taking risks.
I removed the main cable from the printer and inspected it from end to end. I didn't see anything unusual.
I then proceeded to unpin pins 1, 2, 3, and 4. using a needle as indicated by Charles Webb (thanks to you !)
Then I insulated these four cables properly. I noticed that cable 2 was running somewhat randomly along the cable, and was even passing through the twisted white and black pairs...
Then I twisted pairs 1 and 3, then 2 and 4, then twisted 1-3 and 2-4.I put everything back together and... ta-da! The printer now works like a charm. I no longer need a specific startup gcode.
I don't know if cable 2, which was running anywhere, was enough to cause the problem, but I have a feeling that it was.
Below are some photos.
RE: Z axis trouble but only when printing
@charles-web and @tdcsf I expect that I did do the twisting improperly as it was my (mistaken) intent to twist the power wires together and ground wires together. Thank you for pointing that out.
Unfortunately, the forums seem to have stopped sending me notifications of posts in threads I'm participating in, so I didn't notice your replies before I finally found some time to replace the cable over the past two or three days. The replacement is about 1.5cm longer than my kit cable, no colors (just black and white) and seems to have somewhat more fabric wrap. No twisting on the nozzle heater wires.
It seems to be working at least in the first two prints with one full-bed print and one smaller print having done easy homing, nozzle cleaning, and bed probing on one touch per point.
Of course in replacing the main cable, I've also completely redone its routing with the swing arm and tubing, plus I also broke the clip that holds it to the right-hand X/Y motor bracket (I reprinted in TPU). Both of those (routing, clip) have also been identified as possible issues that can cause this problem, so it's not guaranteed that replacing the cable is what helped, but it seems probable.
If it starts causing trouble again, I will certainly try twisting this cable correctly; @emmanuel-r5-2, your thorough photos will be a good help. (I had also tried the 6.4.0 RC firmware and it didn't help. Isn't hurting either, so I haven't reverted.)
Thanks all!