Intermittent connectivity issue with both MCUs
Hello, I have a voron v0.2 that runs ratos on a pi4b 1g connected to an EBB36 v1.2 and an SKRat v1.0. I have verified that every component works individually. I swapped the SKRat from my minion and it still has the same issues. I swapped in a spare rpi and the issue still persists. Same with the EBB36. I have redone the cabling multiple times for both MCUs and the problem still persists. I have reinstalled RatOS v2.0.2 twice using new sd cards and I am currently on the prerelease v2.1.0 with no change.
The issue is that around 25% into any print (an average I've had prints stop immediately and I had one finish) I will lose connectivity with either of the 2 MCUs. Attached is the klippy.log which I don't know how to read and a screenshot of the error in klipper. Any help with this issue would be greatly appreciated!
18 Replies
Traceback (most recent call last): File "/home/pi/klipper/klippy/gcode.py", line 459, in _respond_raw os.write(self.fd, (msg+"\n").encode()) BlockingIOError: [Errno 11] Resource temporarily unavailablemeans it can't write to disk fast enough. Maybe you have a bad SD card, maybe a bad sector, etc
Actually first time i see someone with that scheduling error.
It can pretty much be anything, but it's likely to be a cabling issue, bad strain relief is a very common one.
The IO error is definitely worth fixing, i don't think it's related, but i can't rule it out.
strange that it persists across different rpis and sd cards though. I have been using a SanDisk a1 sd card those are generally pretty hardy
The tophat is closer than I like to the umbilical setup. Maybe that would cause it? Could a strain relief issue on the toolboard cause the SKRat to disconnect?
Yeah prolly not related, but never the less you do have SD card issues
It would cause the toolboard to disconnect
(which is the error you're getting)
toolboard_t0That's just the error I screenshoted it's 50/50 between the toolboard and the MCU
There's only toolboard_t0 issues in your log
But fair enough, that makes it real weird
That was the first disconnect since reinstalling ratos 2.1.0 I may be able to swap back and get a log from my 2.0.2 install
I doubt there's more info in there. Only thing i can think of is some wiring issue, potentially a soft short or something related to movement
I'll check it all, is there anything I should look out for?
Besides wiring? can't think of anything in particular.
Perhaps power delivery
That's what I was thinking. The power supply is the only constant between the 2 boards
could be short brown out problems mayhaps?
if the stars align and all heaters, fans and motors peak at the same time, or just results in a period of high average power draw that exceeds the PSU limits.
IDK I don't think so even if everything was full power at once I don't think it would overwhelm the PSU. What's weird is I can heat up to 110 on the bed and 280 on the nozzle at the same time with no issues. It's strange that it takes so much time to have an issue. Maybe I'll try putting a fan on the PSU it does get a little tosty
It also doesn't do it at idle as far as I've seen so many it's just a bad psu
Plus the motors on top of that which vary wildly depending on load, accel and velocity.
what PSU is it?
It's a meanwell lrs150. This was a Facebook marketplace purchase of an ldo 0.1 that I upgraded to a 0.2. I found some wiring issues with the original owners setup so I would say there's a good chance it had a rough life
I have a manta m5p I could swap in somewhat easy but I kinda doubt that'll solve anything
I personally think the 150W is underspecced, but 🤷♂️
depends on the hardware you put on it and how hard you run your motors (voron configs undercurrents quite a lot by default)
yeah that wouldn't do much when you have dropouts on both the toolboard and the mcu
Totally agree. I may try and cram a uhp 200 in there. I love the v0 for what I use it for but I have some design issues with it
I run it at trade shows on my table it's a great attention grabber
For anyone this helps I figured out the issue. After replacing a bunch of the heaters and swapping the PSU out with a UHP 350 with no improvement I stopped powering the pi over gpio from my SKRat and switched to a btt sksm. No more issues with 50+ hours on the V0 in the last week. I ended up having this issue on my vcore 400 as well and the fix was the same. Not sure why this happened all of the sudden but at least the fix is easy and cheap
How odd. You should've gotten undervoltage warnings in mainsail if pi power delivery was an issue.
(it shows up under notifications)
But yeah, duponts suck balls..
2x24AWG pairs should be plenty.