Moving a Z on a duet wifi
-
@zbeeble said in Moving a Z on a duet wifi:
is there code to handle the two zeds separately with a bltouch as the level sensing device?
Yes of course.
https://docs.duet3d.com/en/User_manual/Connecting_hardware/Z_probe_auto_levelling
You still have to home the Z axis as a whole first, but then you can adjust the position of the two motors independently. However, to do this, each z motor requires its own driver. It can't be done with the single Z driver with the series connectors.
-
@phaedrux said in Moving a Z on a duet wifi:
@zbeeble said in Moving a Z on a duet wifi:
is there code to handle the two zeds separately with a bltouch as the level sensing device?
Yes of course.
https://docs.duet3d.com/en/User_manual/Connecting_hardware/Z_probe_auto_levelling
You still have to home the Z axis as a whole first, but then you can adjust the position of the two motors independently. However, to do this, each z motor requires its own driver. It can't be done with the single Z driver with the series connectors.
OK ... cool. Still doesn't exactly answer how to do the initial homing with the bltouch... or does that just happen with a standard "G30" ?
-
-
I've done further experimentation --- mostly on the combined motors. It would seem that the speed 180 is unobtainable in my setup --- I've increased the motor current to 100% (2000 or 2A).
In my testing, with the two motors on the two Z connectors, 100 or 120 performs properly, but 180 exhibits the one motor running, other motor stalling behaviour.
My questions:
- Are there much stricter limits on the difference in movement torque required of motors bonded in this way?
- Is there a difference in this requirement with parallel vs. serial arrangement?
-
I wonder if it's your very high steps per mm value for Z. Can you send M122 after you've commanded some faster and longer movements on Z? I'm wondering if you're getting any hiccups.
-
Thank-you for your patience. I believe I have resolved the issue... with a combination of your thoughts and my experimentation (as it should be). For posterity, here's wazzup.
BACKGROUND
The printer in question was built (originally) mid-2014 ... and with Chinese electronics. It was noisy and it ran hot. It's largely a Mendel from reprap.org ... combining some of the Prussia Mendel with some ideas from the Reprap-Pro Mendel and some custom changes suggested by the helpful fellow who printed the initial printed pieces for me (among them, a fine herringbone main gear set).
This I upgraded ~2016 or so --- the duet2wifi was a "new thing" when I did this, if that helps nail down the timeframe. I swapped the Chinese electronics for the duet2wifi, I added a heated and magnetic bed and I made minor improvements in the geometry.
This printer eventually choked on some filament and I didn't get around to fixing it until just recently. I have been collecting supplies to make something with the e3d tool changer (but different), but I haven't started assembly yet.
THIS ITERATION
So I started this iteration to improve a few things... one of the most major: add a BLTouch to the system. I ended up learning first that I had some really sub-optimal motors in play --- that only took 0.4A of max current (for X, Y and E0). I replaced these with 1.5A motors. Seriously, the documentation about selecting motors is good (although I found utube about parallel and serial that seemed to explain that better) ... but I'm relatively sure it wasn't available when I first did this.
After the upgrades, I found the print occasionally getting out of square. The pair of Z-axis motors were 2.4A beasts --- and so I hadn't replaced them in the first go round, but motors are pretty cheap ... so I ordered some more of the 1.5A ones. Then my Z-Axis problems got worse.
THE SCALE PROBLEM
So... here-we-are. The problem. Have you guessed yet? I think Phaedrux was dancing around it above. The original pair of Z-Axis motors were 0.9 degree motors. It would seem it is much harder to drive these fast. This is why the motor multiplier was 5104. I discovered this because after configuring my new motors to move more slowly, my "test cube" was 20x20x40 (and having trouble with layer adhesion).
Of course lightbulbs immediately lit (I didn't actually finish printing the cube --- I got to around 50% --- where it was 20cm-ish high ... and noted that it was only 50% finished and stopped the print). I cleared the bed and reached for the caliper --- 10 cm of commanded movement ... 20 cm of actual movement. Hrmmm.
This is the point I realized the situation. So I changed the multiplier to 2552 and adjusted the speed of that axis to 300mm/min (150 was working for the old config, so it made sense. 180 wasn't). I re-printed my test-cube. (If anyone's curious, with a nod to Star Trek's "Transport Test Cube", I have a 20mm cube that I use for testing the printer).
