MMU2S firmware b0rked?  

MMU2S firmware b0rked?

Hi there,

I assembled my MMU2S on a fully functional and problem-free i3 MK3S+ that served me well for quite a lot of spools. And at first I thought MMUS2S would work alright on first test multicolor print, made it through half a two-colored benchy using PLA in Slots 1 and 2. This is where the problems started.

MMU2S went unhappy because it failed to unload on slot 2, flashing fast on LED 2. Opening up the MMU2S unit showed a thickened tip of filament. I removed the filament and cut 10 cm, designed a nice tip and put the filament where it should be after unload: On gear and a bit further, but clear of selector. Hit the right button to continue.

Now things turned really crazy: Selector was still on position 2, FINDA was empty, so MMU2S turned the idler and tried to load from slot 1 (still blocked by selector!!) and made clicking noises. Whywhyhwhy?! The LED 1 started blinking slowly. I can't move the Selector, so I reset the thingy, only to make things worse...
I observed several more erroneous behaviour by MMU2S, and most of them were connected to it loosing position on either the Idler or the Selector, sometimes driving the wrong filament back until it left the gears (unable to load later), not retracting at all the loaded filament, and not able to move the still-blocked Selector afterwards; sometimes feeding the wrong filament, sometimes clicking just for good... The MMU2S' firmware is not sane, it is doing crazy things, and recovering from that seems almost impossible because you can't trigger it to regain sync on its positions.

What the MMU2S should really do: Never loose sync on its positions.
This will mean a lot of troubleshooting and debugging. Good luck.
I guess all recovery routines (what to do after user cleared the reason for a failure) need to be rewritten. It's a hit-or-miss whether it will recover and continue to print or just turn more confused.

What the MMU2S should do to at least is catch the erros: After each failure, once the user presses any button

• resync position of Idler to ensure the right slot is in action,
• check FINDA,
  • if FINDA is detected empty, resync position of Selector.
  • if FINDA is detected not empty, unload and check FINDA again,
• only then do what the button press was meant to trigger.

This would avoid a lot of problems I'm sure.

Please fix this thing, and make it usable. Now it's just a piece of sophisticated hardware driven by a haunted piece of software.

Keep on printing!