New Voron instance lost communication with EBB (can bus)

General Discussion
21

Just for future reference, your log was definitely showing canbus issues:

Stats 9381.8: ... canstat_mcu: bus_state=active rx_error=0 tx_error=0 tx_retries=0 ... canstat_EBBCan: bus_state=active rx_error=0 tx_error=0 tx_retries=978
Stats 9382.8: ... canstat_mcu: bus_state=passive rx_error=1991 tx_error=0 tx_retries=0 ... canstat_EBBCan: bus_state=active rx_error=0 tx_error=0 tx_retries=1001
USB CANBUS bridge 'mcu' is discarding!
USB CANBUS bridge 'mcu' is no longer discarding.
Stats 9383.8: ... canstat_mcu: bus_state=passive rx_error=33557 tx_error=0 tx_retries=0 ... canstat_EBBCan: bus_state=active rx_error=0 tx_error=0 tx_retries=1001

An unstable voltage could certainly cause those canbus issues.

Cheers,
-Kevin

22

These messages appeared after ebb mcu reset (checked this by shooting timestamped video with red status led appeared on ebb mcu boot). So the main reason is a microcontroller reset because of unstable power supply. I’ll try to add 47uf ceramic capacitor(s) to 3.3V line (24V line capacitor too big).

Is there any working ebb2209 printer instances with 70W+ heater ?

23

Today I printed a case for nevermore stealhmax (20 hours of printing) without fails. Capacitor works. Didn’t tried a smaller capacitor on 3.3v line yet.

24

I found this topic while searching for solutions to a similar issue. Symptoms are the same for me and I have the same electronics set up (Manta M8P v2.0 + SB2209 RP2040 connected over CAN).

I’ve successfully printed for over 1000h until this issue started happening and the only thing that has changed recently is starting to print with a flexible filament (similar to TPU). I had no such issues with PETG, ABS and ASA. Moreover, the issue seems to have disappeared again after switching back to PETG.

This, and the fact that it might be addressed by adding a large capacitor leads me to think it’s an ESD problem. Others have already experienced it with this toolhead board, though with a different error: TMC 'extruder' reports error: GSTAT: 00000001 reset=1(Reset) · Issue #47 · bigtreetech/EBB · GitHub One person there came up with some shielding and a bypass to overcome it.

I suspect that flexible filaments moving along a PTFE tube would more easily accumulate a static charge for this issue to happen much more frequently. Especially if the tube is long and terminated inside the toolhead. @regressor Does this happen to you with any kind of filament?

25

No. Adding a better (not implicitly a larger) capacitor to the supply voltages (my favorite would be the main supply voltage (24V)) since the 3,3V for µC and other IC’s is made from that, supports the supply voltage.

I try to put in it simple (not technical) words.
If a bigger current is needed in “a certain situation” of the printer with an EBB or SB2209 RP2040 very quickly (we talk about µs), the power supply (24V) is way too slow to deliver that needed current.
That is why capacitors are placed on the PCB (EBBs or SB2209 RP2040) to support the supply voltage. They should be places in many places all over the surface (my way) of the PCB.

EBBs or SB2209 RP2040s come from China and are very low priced. Every component on the PCB adds to the production costs, So less components on the PCB may end in a better margin. Maybe BTT was too cost oriented or they have some charges of BTTs or SB2209 RP2040s with bad capacitors (less capacitance, than they should have).

Are you talking about Electrostatic discharge - Wikipedia
If yes, then no ESD problem.

Does your SB2209 RP2040 has more to do with that filament?

26

I don’t think this is an ESD problem. Do you have special retract settings for TPU ? A higher speed and length ? This can lead to higher extruder motor consumption and to voltage drop.

Think about a can power line as a thin water tube. So if you need more water at the end of the tube for a short time you should add some temporal accumulated volume for water on the tube end.

27

As this thread is still alive, my 2 cents.

I think it is possible to increase the supply voltage to 25V (safe), 25.5V, or even 26V.
This should generally simplify things and allow capacitors to hold slightly more charge.
All of the components, with account to TMC, generally withstand closer to 29V.

On the other hand, if the hotend creates a power issue, I think increasing the PWM frequency can help.
Something like:
pwm_cycle_time: 0.100 -> 0.002 should do the trick.

This should spread out the load much more evenly.

In case there is a need, it is still possible to limit the max_power.

28

Of course possible, but I wouldn’t recommend it. You’ll never know if you cause other implications.

29

Thanks for suggestions everyone and sorry for not responding earlier, but I ended up experimenting for a long while and finally got a reliable printer so I thought I’d leave this for posterity. TL/DR: it’s a hardware issue, either shorting or ESD from the motor casing.

First, I thought it might indeed be extruder motor being more busy with increased PA and retraction for the flexible filament, but then I changed those to match my PETG profile and the issue persisted. Then, I started getting these toolhead resets with PETG too, just much later in the print (about 3-4h) that repeated consistently with every print and later started happening even earlier, but still much later than with the flexible material.

Here’s what I tried and none of it helped:

  • Soldered a 1500uF 35V low-ESR capacitor to the toolhead power input
  • Increased PSU output voltage to 25V
  • pwm_cycle_time: 0.100 -> 0.002
  • Replaced the PSU from 200W to 350W

I noticed though that sometimes, after some changes it would work with PETG for a few days on long (12-22h) prints before exhibiting this issue again. I also noticed that, once the issue happened and a firmware reset was done, I could almost always repeat it by putting my finger on the plastic XT30 connector of the power/CAN cable or triggering the X endstop switch (which is connected to the toolhead board). After that it wouldn’t happen not matter how hard I pressed on it or other parts of the board or triggered the switch. Until another “natural” occurrence.

A few more details about my setup is: Galileo 2 extruder and a Cartographer 3D probe (connected via CAN pins on the SB0000 Stealthburner board). A few things that were suspicious to me in this setup:

  • Extruder motor casing is very close to the SB2209 USB-C port’s body (might be a conduit for a discharge?)
  • Extruder’s metal carrier shaft is pretty close to the pins on the SB0000 board, though not touching (might be with enough thermal expansion?)
  • Cartographer 3D wires are a bit stressed due to the placement of the SB0000 board (might cause resets/CAN issues?)

Also, I noticed that new packages of the SB2209 RP2040 (both CAN and USB) contain a wire to connect any ground pin on the board and the extruder motor casing, which the USB-variant assembly manual recommends to do in order to provide a continuous discharge path.

So in an act of desperation I did the following things:

  • Soldered an 18AWG wire to the GND pin of the SB2209 power input, crimped it with a circular terminal and attached it to one the extruder motor’s screws, secured with a nut
  • Soldered all the Cartographer 3D wires to the back of respective SB2209 pins that connect the SB0000 board
  • Covered those with an insulating varnish just in case too
  • Covered the SB2209 USB port with an insulating tape
  • Covered all the unused pins on the SB0000 board with an insulating tape (especially those close to the carrier shaft)

Something (or all of) this has finally addressed the issue and I’ve been reliably printing with various filaments (including the flexible one) for over a 100h.

I’m still not sure why it only started happening after I printed with that flexible filament for the first time (and only after a few hours of doing that). I had both G2E and Carto for a while before then. Might be a coincidence, but the consistently more frequent occurrence when using that filament even with PA and retraction decreased to the PETG levels (+ much lower flow rate) is suspicious to me.