A350 Heated Bed temporary drop

[jvilhuber]

Hardware: A350, Dual-extruder, Enclosure

I have this problem, where, when printing PETG, with a bed temperature of 80, the print starts just fine (takes a while to heat up, but eventually reaches 80+ degrees). It prints the first few layers, and then mysteriously the temp starts dropping. It SEEMS to be a temporary drop (5 or so minutes): Sometimes it catches itself, and start rising again, and sometimes not (causing a “temperature runaway” error). After it catches itself (if it didn’t panic the machine), it proceeds to print for many hours just fine. So it’s just after the first few layers.

I’ve swapped out the bed (new cable on the bed), measured the cables with a multimeter (seems within spec of what’s listed in the wiki), swapped out print heads (both are dual-extruder), and nothing seems to work.

At first I thought it had to do with low ambient temperature (garage; 14-16C in the winter), but I have an enclosure, and put an enclosure fan inside (40C) and the problem keeps happening.

Looking at the power-draw on the power plug for the printer I can see that during the warmup there’s a constant draw (< 250W). When it reaches the target temp, it fluctuates up and down (to be expected if it’s just switching on periodically to keep the temp even-ish), and then the temp drops and the power draw goes continuous again (about 300W. More than the original ramp-up), but the temp continues to drop, so it’s anyone’s guess as to where that power is actually going.

I probably don’t need to even mention the uselessness of Snapmaker support here..

My guess is that the power supply (350W-24V, if I’m not mistaken) is too weak (seems weird, though; the voltage used throughout is < 320W (plotted via octoprint), which SEEMS ok for a 350W PSU).

Has anyone encountered this and possibly found a solution? The Meanwell LRS-600-2v is only about 80 Euro, and SHOULD fit in the original case (and if not, I’ll just keep it open or print an enclosure). But if that is known to not help I might as well save the money (I’ve already thrown more money at this printer than I like).

Thoughts?

[xchrisd]

Your gcode is not causing the up and down, correct?
Search for M140 S in your gcode to verify.

Was your replacement bed new?
At first I thought you might have a bad heated bed cable but obviously it’s not sure because of your replacement.

You wrote about a fan and low ambient temperature.
Does the fan blow cold air into the enclosure?
If you blow air on the thermistor in the middle of the bed, this could cause a temperature runaway because of the missing power capacity to proper hold the heat.

Hope it helps you figuring out.

[jvilhuber]

Correct. The gcode is steady at 80C.

As for the fan: I should have used the word ‘heater’ not ‘fan’. Sorry. it’s a small heater that I cranked up to 40C. If I let things just sit (not printing) while the heater is on, the temp on the nozzles and bed show around 40C, so that seems right.

But the temperature runaway happened before I put in the fan. Support said something about ambient air being too cold (possible), so I added the fan to bring up ambient air. Same problem (no change).

[Cadae]

I don’t think ambient temperature has anything to do with it; my printer is in a similarly cold room and has not experienced that problem.
Does the issue only happen during prints, or does it also happen if you just set the temp manually?
If it only happens during prints and it’s not the gcode it might be a loose wire that appears connected when the print starts and during your measurements, but stops working at a certain angle.

[jvilhuber]

It only happens during prints. I raised the bed temp to 80 without a print for half an hour and everything was fine.

But note I mentioned I also got a new bed which came with a new cable. So it (likely) is not the bed/cable (I also check the impedance or whatever of the bed against the published specs and they match).

It could of course be another cable. But which one?

Edit: Note also that after that first dip, once it recovers, it prints for HOURS without any further glitches!

[Cadae]

Since it goes back to normal for hous its sound like an issue with either the gcode or the firmware. Have you tried updating/changing the firmware version

[jvilhuber]

Yea, I looked at the gcode and I see no ups and down in temperature. I tried different slicers, too, just in case. Support (such as it is) looked at the gcode I sent them and didn’t say anything about it (meaning they probably didn’t look at it, for all I know).

Since we’re all presumably running the same firmware (they don’t update it often, and not in a while, so I assume we’re all roughly on the same), I also can’t imagine it’s a firmware bug, but software (and hardware) is complex, so… maybe? Not sure how I could determine it, though. I don’t think there’s been an update since I got the printer (less than a year now). I guess I could try to downgrade.

I also sent the logs from the printer to support, but again they didn’t comment on it.

Shrug. I’m at a loss.

[Simon_Zhi]

https://snapmaker.formcrafts.com/support-ticket

You can submit a ticket at the same time in case the technical staff or fellow users in the forum don’t respond promptly.

[jvilhuber]

Case has been open for a few weeks and rather useless (so far… I could still be surprised later).

[Simon_Zhi]

I’ll try to push the after-sales techs tomorrow to look at your ticket—if you’ve submitted one, let me know the details.

[jvilhuber]

Doesn’t help they’ve been on lunar new year holiday (other companies have call-centers around the world so they aren’t affected by local holidays, but I digress). Let’s see if they answer in the next week. I’ll get back to you, maybe.

[xchrisd]

Because of the mentioned issues I guess it has to be a fault in the controller itself or the connection.

If you are not confident reading gcode you could upload it here or to a third-party service so we can check for temperature commands.

May you share a photo of your machine?
Is the bed connection on the controller side properly seated? Do you sea any color change on the connector pins (controller and bed side)?

[jvilhuber]

I did what you previously suggested with the “M140 S” command (stripping out the comments):

$ grep ‘M140 S’ Support\ Telepeage\ v3_0.2mm_30m14s.gcode | grep -v ‘^;’
M140 S80
M140 S0

If you want to look at the full gcode file, I’m happy to share it (not sure how on this board).

I’ll check the connectors on the controller. Support also just asked for that.

[xchrisd]

You could upload the gcode to a one click hoster like WeTransfer or Google cloud and share the link here.
If it’s small enough, like below 10MB(not sure), it’s possible to upload it here

[jvilhuber]

Oh it’s quite small, it seems (I don’t read gcode much ). This happens on various prints with various settings, but it happened on this one yesterday (I was testing some changes to settings to see if they would matter. They didn’t).

Support is sending me a new controller and heated bed. I hope that resolves it.. Will update when I get them.

Thanks for the kind help so far! Much appreciated! Dankeschön!

Support Telepeage v3_0.2mm_30m14s.gcode (976.4 KB)

[Cadae]

the gcode looks totally normal, are you running these prints through octoprint. Have you tried disconnecting the octoprint device from the printer completely and running the gcode directly on the printer, its possible some plugin or macro is trying to turn off the bed.

[jvilhuber]

I didn’t disconnect octoprint, but I did print directly from Luban and got the same results. I could try with octoprint disconnected, but without it driving the print, I can’t image it’s doing anything (but you never know).

[jvilhuber]

Update: I received a new heated bed and controller from Snapmaker. Didn’t make any difference: I still see pretty big fluctuations in the temp of the heated bed.

Out of curiosity, I swapped out my dual-extruder for the single-extruder, and noticed that in this case the bed-temperature is maintained ALMOST PERFECTLY throughout the whole print.

With the dual-extruder, the temp goes up to almost 82C and down to 79C even BEFORE the large drop. Also note the ramp-up: The power-usage is very jumpy between 220W and 250W.

Compare that to a run with the SINGLE-extruder:

The temp is maintained almost prefectly with no major drop (or overruns). Also note the ramp-up: the power-usage is pretty flat! There’s no up and down at all.

Also the maximum power usage appears to be slightly below 300W with the single extruder, and with the dual extruder if goes up to around 315W.

Two things come to mind (really… they’re just one):

1. why would the print head affect how well we maintain the temperature in the bed?
2. is the power suppply (350W?) being maxed out by the extra draw of the dual-extruder somehow?

I’ve emailed support all this. Let’s see what they say.

(I wonder if swapping out the 350W power supply with a 600W meanwell would help…)

Edit: I also updated to the latest firmware that just came out (didn’t make a difference) and cleaned out all the old files in the controller (I was getting a warning that too much space was being used, which might affect functionality), which also didn’t make a difference

[jvilhuber]

I also just went back in time and looked at a PLA print, where the bed is only at 55C (still dual extruder) and I see similar overruns on temp, and the temperature fluctuations are much bigger than with the single-extruder:

I’m pretty suspicious of the > 300W power-draw (as measured at the power-plug so who knows what is happening inside the power-supply and on the printer-side).

[Boxkite]

Does the dual head have a much stronger blower that cools the bed once it lowers to print, and then after a few layers are laid it blocks the air reaching the bed it starts to warm up again?