I found serious design error in MMU2 electronics hardware
 
Notifications
Clear all

I found serious design error in MMU2 electronics hardware  

Page 5 / 5
  RSS
Stefan
(@stefan-9)
New Member
RE: I found serious design error in MMU2 electronics hardware

I also faced the same problems with the flashing 10 LEDs, as info, my MMU unit is from a time when it came on the market and is board rev.03. I measured the +5V voltage on the MMU-board, which was 4.27V, as already stated, too low for safe operation.

I had a closer look into the schematics ( https://github.com/prusa3d/MM-control-2.0/blob/master/rev.03/MM-control.pdf ) and looked for the best solution of the problem. In my opinion, bridging the diode D1 is the easiest, cheapest and best solution to this problem. After the bridging I can measure 4.65V, and I could not observe the problem anymore since then.
I want to discuss possible problems with this approach:

  • D1 provides reverse protection against swapping +5V and GND: not really an issue, as the MMU is connected permanently to the Einsy mainboard, therefore no occasionally swapping should occur
  • D1 provides reverse current flow protection, so that my by USB connected PC cannot power the whole Einsy mainboard. That is theoretically true, but only when the diode D7 next to the USB-Port is in place, when this diode is missing, its not even possible to power the MMU board by a connected PC, as the only place where the VBUS voltage from the USB connecter is going is to the VBUS-Sense input of the MCU which is only used for sensing if a USB cable is connected or not, BUT not to power anything. On my board, diode D7 is not populated, therefore, there is no reason why bridging D1 could lead to any problems in my eyes.

The interesting part is, that also on all PCB-pictures in this thread, the diode D7 is not populated, so it seams to be common. When this is true for all units outside, the solution with an external USB wall-charger should not help at all. Maybe also the reason, why the USB wall-charger solution also only helped temporarily for most people.

Posted : 03/08/2022 3:48 pm
Robin liked
Martin_au
(@martin_au)
Reputable Member
RE: I found serious design error in MMU2 electronics hardware

Can confirm. No D7 on my MMU, and it won't power via the USB. It reqiures printer power to work/firmware update.

I'm going to bypass the D1 (and make sure I don't mix positive and negative. :D)

Posted : 06/08/2022 8:00 am
bryn51
(@bryn51)
Eminent Member
RE:

I confess not reading every post in this 2y/o thread. But I got my MMU2/s in March 2022 and still I have had the symptom where Idler behaves erratically  and sometimes the other two steppers also (but harder to spot. )

I took a different path to correction: It was noted by myself and others that the condition was influenced by movement of the wiring loom. In operation, the loom can vibrate. It has more mass than the wires supporting it, so it moves up and down, and the lever arm of individual wires to connectors moves also. So the connector is vibrated. It has tin contacts against gold contacts on the board. So vibration can cause noise and intermittent contact.

But likely it would not be such a problem if voltage was correct, so the issue you found with the voltage could be working hand in glove with the contact issue. The contact issue is solved using a contact cleaner and protectant Deoxit. But likely this is temporary fix because the protectant can evaporate over time.

So I plan first to verify that the low voltage condition still affects the MCU board I have from March ‘22. I am avoiding it because it’s not easy to get in there even with small probes.

I have already created and shared with Prussia some documentation about the MMU2, and so I can update that. Modifying the board either by playing with diode or adding a boost voltage is not something I favor, as it cannot be completely reliable.(IMO).

so what is the best solution? The mk3S+ Power supply must already be running close to its limit, the reason we see regulated voltages dropping below threshold. Then there are multiple places voltage can drop toward the MMU MCU board. So, I’d be looking at raising the 5v rail slightly if it can be adjusted. Perhaps it has component tolerances affecting the voltage delivered, and they could be tweaked with high accuracy (1%) components. 
addition of a buck/boost chip could improve operation. 
this is speculation… 

I’ll post back here with results of my validation tests on “current” hardware (unintentional pun).

thanks to all those people that posted here in the last two years. It provides provenance that supports case for Prussia Research to take  some action. 

In the background is the unknown effect development of the XL printer has on the future of MMU. I think the Mk3S+ still has a life because the XL is such a high price. And with it, the MMU. So IMO. it’s worthwhile for Prussia to provide a pathway for correction of this issue which appears widespread. A replacement/upgrade to that MMU/MCU board appears justified.

cheers to all and happy printing.

This post was modified 6 days ago by bryn51
Posted : 30/09/2022 1:15 am
Page 5 / 5
Share: