Octopus Pro board died? No CANBUS connection out of the blue

General Discussion
1

Basic Information:

Printer Model: Tronxy X5SA Pro
MCU / Printerboard: Octopus Pro v1.1, EBB42, MMB, Eddy Duo (all BTT), BTT CEB
Host / SBC: HP ThinClient T630 Debian Trixie

klippy.log-zip.zip (4.4 MB)

klippy.log

Fill out above information and in all cases attach your klippy.log file (use zip to compress it, if too big). Pasting your printer.cfg is not needed
Be sure to check our “Knowledge Base” Category first. Most relevant items, e.g. error messages, are covered there

Describe your issue:

I’m a bit at a loss right now. My x5sa (converted to Klipper using a btt octopus pro 1.1 board in CAN-Bridge mode) has stopped working. It simply lost the CAN connection to the toolhead BTT EBB42 and the MMU BTT MMB whilst idling. A cold start helped twice (disconnecting and reconnecting the 24V and,USB to host), but not anymore.

Stopped Klipper; only the Octopus board is showing up on the CAN bus now. I find it unlikely that both the EBB42 and the EBB MMB would fail exactly at the same time. And the BTT CEB distribution board connecting all devices is purely passive. The cable from the Octopus to the CEB was loose on the board, so I swapped it for another one, which is now secure. Measured cable, it’s ok.

No change. I’ve now run out of ideas.
What a mess – I was just about to carry out the final calibration steps on the ERCF…

What would be the best place to get help how to diagnose the root cause? Just buying a new Octopus pro without knowing it actually failed is a bit expensive …

2

Ok, now it get’s really strange… some measurements later…

  • everything disconnected from CEB board: no connection between H and L, so no jumper set on the CEB. That’s ok.

  • only connect H+L to Octopus Board with 120R set on Octopus to CEB: 122 Ohm between H and L on CEB pins. OK.

  • disconnect Octopus from CEB, connect MMB without jumper on MMB: no connection, that’s OK

  • set jumper on MMB: 124 Ohm.OK as well.

  • disconnect MMB, connect EBB, jumper not set on EBB, but Eddy connected to EBB: 123 Ohm.
    Ups, the Eddy seems to have 120 Ohm register on H/L, no way to change this?

  • disconnect Eddy from EBB, keep connection to CEB: no connection on CEB between H/L

  • reconnect Eddy: 123 Ohm on CEB

  • connect all devices (Octopus, EBB, MMB, Eddy) to CEB: 60.3 Ohm between H and L on CEB.

Wiring is and has been: Eddy to EBB, EBB, MMB and Octopus to CEB. Jumper set on Octopus, jumper not set on MMB, jumper not set on EBB.

This gives 60.3 Ohm between H and L, as it should be.

Wiring seems to be fine.

Still no devices on CAN-BUS…

3

Can you provide a network map? Include approximate branch lengths and where the resistors are located.

To diagnose, I’d try to simplify as much as possible. Disconnect the EBB and MMB branches (eddy has the resistor?) and see if you can get the Octopus to connect with the eddy. Then temporarily set the resistor on the EBB and connect only it.

Have you wired the eddy to the CEB or daisy chained it to the provided pins on the EBB42?
The daisy chain eliminates a branch and puts the resistor at end of the toolhead branch.

4

The Eddy is daisy chained.

CAN problem resolved?
Physically speaking, this is a kind of wire-bound radio, with a bit rate of 1 MHz square wave plus harmonics (otherwise it wouldn’t be a square wave). That’s why you must/should twist the signal lines; otherwise, you create what’s known as a Lecher line – in other words, an unwanted, length-dependent resonator. Hence the 120-ohm terminating resistors at the start and end, and nowhere else.

So I reconnected the cables. The BTT CEB distributor has the connectors arranged in a row, with an XT30 2x2 at one end and XH2.54 elsewhere.

As the BTT Eddy has a fixed 120-ohm resistor that cannot be bypassed (which I couldn’t find in the documentation), and which is connected to the BTT EBB, I had already removed the jumper for the 120R on the EBB.
I plugged the cable to the EBB into the bottom/rear port on the CEB; it was previously in the middle (with empty ports ‘behind’ it). The EBB MMB for the ERCF is connected via XT30, i.e. right at the ‘top’/front. The Octopus board with the CAN bus controller is plugged in somewhere in between using an XH2.54 connector.
I then noticed that the 120R was set on the Octopus, as this was originally the physical start of the chain. The jumper was not set on the MMB.

The H/L ports are all connected in parallel in terms of DC, and as there were 2x 120 ohms in parallel, the (DC) ohmmeter gave the correct reading of 60 ohms.

In the HF range, however, things seem to be different.

In any case, I removed the jumper from the Octopus and fitted it on the MMB.

And, ta-da, the CAN bus is working again.

The catch with this explanation is that the CAN bus ran for months even with the ‘incorrectly’ placed 120R termination and then suddenly stopped… it shouldn’t have been able to do that.

So it could also be that by reconnecting (the identical cables) and repeatedly disconnecting the power for long periods, the contacts are simply OK again and all the chips have been reset…

Hopefully it’ll stay that way now…

5

In the name of science could you try this:

Switch the eddy to USB mode and re-enable the resistor on the Octopus?

The Eddy’s actual data mode is selected when you compile the firmware. My theory is the switch is a SPST device that only enables/disables the resistor.

As for RFI you never know… was your router in 2.4 GHz and decided to use 5 GHz today? Did the cell company add a tower and changed their beam forming settings resulting in a 10dB gain in your workshop? In 2026 in a decent sized town it is a miracle anything ever works.

Also - Please flag your last post as the solution.

6

I like that :slight_smile:

3 Likes
7

Well, yes… I hate USB cables on moving parts. I guess yourxare right about the resistor switching.

Don’t get me started on RF/HF … I’ have a PhD in physics and I am a licensed ham radio operator… (but working in Cybersecurity now).

I try to avoid WLAN as far as possible, so e.g. my 3d printers all run on Ethernet cables, not WLAN.

And totally unrelated: my Internet connection broke down last night (fibre to the home), only ipv6, no ipv4 … bad day in home office, as my company is only reachable via ipv4 …