Still struggling with Smoothie

Controllers Smoothie
1

I previously posted about missed x/y steps with no response.
My printer never missed any steps prior to upgrading the firmware and enabling grid leveling. Since then, I’ve had nothing but trouble.

I’m seeing some *very* weird behaviour going on….

The first time I saw the missed steps was directly after I set up the bltouch and grid leveling. I then turned grid leveling off and tried again which resulted in missed steps again. I then turned grid leveling back on and it would not level. I could see the z motor making adjustments to the height, but the nozzle would still crash into the back of the bed (but not the front). Upon further inspection, I could see that the leveling was making adjustments while the nozzle moved in the Y direction, but then at the end of the Y movement, the bed jumped back to it’s original z location. And then upon starting a move back to the original Y location, just before the move, the z would suddenly jump down to where it should have been and then make it’s slow adjustment/correction. I have a video of this, but I can’t post links but I’ll try: photos.app.goo.gl / A7JINLwAv5nhicA62

And there’s even more weird behaviour going on with the z too. This is with grid leveling cleared and off. M375.1 returns all NaNs. If I home all axis (z homes down/away from the nozzle, corexy), and then move up to a probing location and run G30, the first time I get 1.9mm, and all subsequent times, i get 2.4mm. This is 100% reproducible. So I figured, something must be loose with the bed such that when it moves down after having just moved up, there’s play/slop/backlash. So the next test I did was to move to the probe location and first do an up-down z movement before the first probing. This test ended with the same result: First probing is always 0.5+mm closer to the bed. It also makes a very funny noise when moving x and z simultaneously but sounds smooth when moving one at a time. I’ve verified that smoothie is actually moving the z motor to a different location after finishing the first probe. I have a video of this as well. You can also hear the weird sound the motors make when it tries to move both z and x at the same time.
photos.app.goo.gl / 6UbMf2uAQ0CHBhht1

I’m pretty lost at this point…

Imported from wikidot

2

Behavior this weird must have some kind of weird reason behind it, this doesn’t happen in thousands of current Smoothie setups.

In order to help pinpoint what’s going on, let’s start with trying to reduce the number of possible contributing factors : 

  1. Upgrade to the very latest firmware
  2. Start your config back from the example config on the wiki, editing only strictly what you need to edit to get it to work ( and don’t use M500/config-override )
  3. If possible get the bltouch out of the picture for the duration of the debgging ( a simple mechanical switch will do the job )

Thanks.

3

Behavior this weird must have some kind of weird reason behind it, this doesn’t happen in thousands of current Smoothie setups.

In order to help pinpoint what’s going on, let’s start with trying to reduce the number of possible contributing factors : 

  1. Upgrade to the very latest firmware
  2. Start your config back from the example config on the wiki, editing only strictly what you need to edit to get it to work ( and don’t use M500/config-override )
  3. If possible get the bltouch out of the picture for the duration of the debgging ( a simple mechanical switch will do the job )

Thanks.