Under voltage events causing printhead crashes

[HarvesteR]

Hi,

I realise there are a fair few posts about these issues, but I wanted to see what others think about my case, and if it warrants something like getting a new PSU.

I started getting consistent print failures here, which usually almost never happen. After sitting in on a print, I got lucky and saw the failure as it happened.

There was an undervoltage event logged to the console. One of the motors didn't move properly, juddered around a bit, and the hotend lost its reference, crashing into the print a very short time later.

This is a SeeMeCNC RostockMAX 3.2 (Delta) here. Had it since 2018, always worked fine. Only mod of note is a PEI print surface. The printer lives in a closed cupboard of its own, so temperatures are pretty stable. I pretty much always run the bed at 90°C, hotend at 215°C (for ABS).

So I started monitoring it as it prints now to see how these undervolt events are happening. I noticed that Vin was sat around 12.3v when idle. Checked it with a multimeter. It agrees with the onboard Vin sensor.

Now, during printing, I'm seeing that the voltage does vary a potentially significant amount, especially in response to the bed heating. When the bed comes on, Vin drops a full volt, to around 11.3 or so.

So just now I tweaked the voltage trim pot on the PSU, to give it an extra half volt, thinking maybe it's just oscillating too close to the threshold... We'll see how that goes I guess... print is underway.

As it's printing, with the hotend and bed holding temperature, I'm seeing a regular oscillation between 12.8 and 11.6 or thereabouts. Hopefully the extra half volt is just enough juice to keep it from spazzing out.

But this is still a bit disconcerting... I've never had these issues before, and I'm thinking that maybe the PSU output should not be varying so much... So I wanted to get opinions... Do you think the PSU is on its way out?

Will follow up as the print continues. Maybe I can run it a bit higher for a little longer as I look at a replacement PSU... or idk, we'll see.

In any case, thanks in advance for any feedback!

Cheers

[HarvesteR]

Well, following up.

The extra half volt seemed to help. The print finished, at least.

Still doesn't leave me very much at ease with this. The voltage is fluctuating more than I would have liked to see... but at least if I can keep it in a workable range, I can live what that.

I'll keep an eye on it. Still have to print the mirrored other half of my thing here anyway.

Another question, in any case... I dialed the voltage up to 12.8v here... but I see that the duet board is 12-24v. I wonder if I'd be better off running the whole thing at maybe 14v then?

The bed element is 12v, and probably so is the hotend... but what would the extra voltage cause there? They'd get hot?

The motors are NEMA17s... idk how the duet regulates their driving power, but if the board tolerates 12-24... I figure there is voltage regulation for the motor drivers then?

Just thinking out loud here, but wanted to see what the general feeling is towards running a little north of 12v.

Cheers

[HarvesteR]

Another follow-up... the new print failed again. The crash wasn't as bad as others, because it actually recovered and continued printing, but there was a large lateral shift at one layer.

I went to check the console, and as expected, undervolt errors again.

One of them was interesting though. It says 12.7v... I wonder if there's something else causing these errors to be falsely reporting... (although most likely it just logged a reading too late or something)

I'm going to try updating the firmware just in case... turns out I'm still running on the 2018 version.

[deckingman]

@HarvesteR If as you say, the printer has been running without issues since 2018, then although the firmware is old, I don't see any reason why updating it will fix your under voltage events. But if the PSU is sized appropriately, then the voltage should remain stable unless the PSU itself is faulty. If it were me, I'd change the PSU.

[HarvesteR]

@deckingman Yeah, that's a good point. The PSU has run up to now without issue, (it came with the RostockMax 3.2 kit, so I believe they picked a good one for their machine)...

So yeah, after all this time I think really the only possible source of trouble could be the aging caps on the PSU... Might be worth a replacement.

The new firmware likely doesn't help, but it is nicer

[HarvesteR]

Well, several days later, tried again... and had the same issues about at the same time into the print (was a very similar model but not the same file)...

And looks like SeeMeCNC doesn't deliver anymore to Mexico...

But thankfully finding a similarly rated PSU was easy enough, and relatively inexpensive.. I was preparing for a lot worse. New one should arrive next week.

[dc42]

@HarvesteR check the connections between the PSU and the Duet because a bad connection could cause that voltage drop. In particular, check that the screws in the VIN terminal block of the Duet are tight and there are no signs of scorching on the block or the wires.

[HarvesteR]

@dc42 I did when I first checked it, the terminals seemed ok. There are some crimp connections that are a big munged... I didn't have the right crimping tool at the time, but the connections are solid.

I replaced the PSU today. Running the same print again, let's see how it goes. It's been running for about as long as when the failure happened last time, so fingers crossed.