2

I've been trying to trace down a mysterious reduction in the flow I can achieve. Since recording this flow test video where I successfully achieved a nominal 30 mm³/s (almost surely with underextrusion, but with no skipping or warping), something happened whereby I'm no longer able to reproduce it.

Shortly after that was recorded, I discovered a burnt-out XT60 connector between the power supply and printer, and replaced it. That made a lot of voltage sag issues (fans slowing down when heaters turned on, etc.) go away, and I'm wondering how it might affect temperature measurement and indirectly flow.

Could the thermistor voltage have been affected by undervoltage/limited current availability from the power supply, and if so, how would that manifest? Hotend (including thermistor) is stock Ender 3. The controller board is SKR E3 Mini V2. Naively, I would expect that if there was actually reduced voltage on the (3.3V?) line feeding into the thermistor, that would appear as higher resistance, and thus lower temperature, driving the controller to heat it well above the requeste temperature, but I'm not sure if that plausibly could have happened.

R.. GitHub STOP HELPING ICE
  • 10,302
  • 4
  • 24
  • 57
  • Heck of a flow test, is that video sped up? – Kilisi May 05 '22 at 07:22
  • 1
    @Kilisi: No, it's realtime. – R.. GitHub STOP HELPING ICE May 05 '22 at 17:53
  • Very impressive :-) – Kilisi May 06 '22 at 02:22
  • @Kilisi: Yeah, now if only I could still reproduce it. Thus this question. – R.. GitHub STOP HELPING ICE May 06 '22 at 02:37
  • @Kilisi: Yeah, if only I could still reproduce it. Thus this question. – R.. GitHub STOP HELPING ICE May 06 '22 at 02:37
  • thermistors likely use 3.3v or 5v signaling. It would take a lot of sag for those rails to drop below set voltage if fed 12v or 24v; all they need is a volt or two above output. As far as the effect of a sag on the 3.3/5v line, it's impossible to say universally because the thermistor could be on either side of the voltage divider, so it could be too high or too low, or it could use a current source unaffected until the rail dropped to <1v. It could even be a thermocouple instead, so the effect would be how the gain amp responds to sag or the MCU glitches. In short (haha), we can't say... – dandavis May 10 '22 at 05:10
  • @dandavis: I've specified a particular board and thermistor, so it's not impossible to say. – R.. GitHub STOP HELPING ICE May 10 '22 at 12:14
  • well i didn't catch the model somehow, my bad. I found a schematic and it's on the gnd side of the divider. So that's half the equation solved. But i was not able to find a part number for the actual thermistor. If you could find out if it's NTC or PTC you'll have your answer. – dandavis May 10 '22 at 18:54
  • @dandavis: It's EPCOS 100K B57560G104F or equivalent. – R.. GitHub STOP HELPING ICE May 10 '22 at 19:54
  • that's an NTC, so it would appear as warmer. That said, I highly doubt your 3.3v "rail" sags, and if it does, the whole controller would freeze up and nothing would operate. – dandavis May 10 '22 at 20:02
  • OK, so it sounds like there are multiple reasons this isn't possibly the issue. Would you like to write the comments up as an answer so I can accept it? – R.. GitHub STOP HELPING ICE May 11 '22 at 01:52

1 Answers1

0

As worked out by dandavis in the comments, this is not plausible:

thermistors likely use 3.3v or 5v signaling. It would take a lot of sag for those rails to drop below set voltage if fed 12v or 24v; all they need is a volt or two above output. As far as the effect of a sag on the 3.3/5v line, it's impossible to say universally because the thermistor could be on either side of the voltage divider, so it could be too high or too low, or it could use a current source unaffected until the rail dropped to <1v. It could even be a thermocouple instead, so the effect would be how the gain amp responds to sag or the MCU glitches...

I found a schematic and it's on the gnd side of the divider. So that's half the equation solved...

[The EPCOS 100K B57560G104F] is an NTC, so it would appear as warmer. That said, I highly doubt your 3.3v "rail" sags, and if it does, the whole controller would freeze up and nothing would operate

It turns out the PTFE in the heatbreak was badly burnt out, and friction from this may account for at least a large part of the loss of flow (I've been able to get back some but not all). But I still have not fully explained or solved the problem. In any case, the power supply issue does not seem relevant, answering this question as stated.

R.. GitHub STOP HELPING ICE
  • 10,302
  • 4
  • 24
  • 57