Tevo Owners
Temperature reading fluctuations - Printable Version

+- Tevo Owners (https://tevo-owners.com)
+-- Forum: TEVO Tornado (https://tevo-owners.com/forumdisplay.php?fid=13)
+--- Forum: Help (https://tevo-owners.com/forumdisplay.php?fid=23)
+--- Thread: Temperature reading fluctuations (/showthread.php?tid=204)



Temperature reading fluctuations - tranmal - 02-15-2020

[attachment=53][attachment=53]TL;DR = Printer bed and hotend temperature fluctuate +/- 3C even at idle only when the AC is turned on for the control.

Please note that the temperature on the hotend AND the bed are involved and it started mid-print (I didn't add/remove/change/move/fiddle with anything between good readings and bad).  I also tried PID tuning on the hot end AND bed with no improvement.  The temperature fluctuations defy the laws of physics.

I use Octoprint.  That's mostly irrelevant...but it's how I noticed the issue.

Normally, the temperature readings are pretty smooth (1/10 degree C here or there) on the print head and heated bed.  That all changed one night when I was printing a lightsaber handle.  It was about an 8 hour print that I started and went to bed.  It turned out just fine and was sitting there in the morning ready to add a blade.  That's when the trouble started.  The normally smooth graph for the temperature was all jaggedy.  The temperature readings were jumping all over the place (+/- 3 degrees C within a second or two on the bed and the print head).  So, I headed to the actual printer display to confirm that it wasn't an Octoprint thing.  It was doing it on the display as well.

The Octoprint will power the controller board but not the servos or the heaters.  So, I unplug the control from the Octoprint and the 110VAC, then plug in the Octoprint.  The board powers up and the temperature readings are stable for both.

Weird.  Thinking that there was some strange firmware thing, I shrug it off and plug in the 110VAC and turn it on.  I head back to start the print through Octoprint, and I see that the temp graph is doing it again...back to the printer and it's jumping all over the place on the display as well.  I turn off the 110VAC again, but leave the Octoprint plugged in (so the board still has power).  The temperature stabilizes again.

It HAS to be that.  I grabbed a spare AC power cord and took a power strip out of the equation.  I turned on the AC again, and BAM!  Temperature jumps around again.

It's the power supply.  I order one, install it, plug it in, turn it on...temperature jumps around again.  Clearly not the power supply.

Must be a cold solder joint on the board in the temp circuit.  I pull the board and grab a magnifying glass.  All the joints are shiny and filled.  It might be a bad pull up resistor or a wonky capacitor...I'll try to fix that later...I order a new board, flash it and install it.  Plug in the AC and...same thing.

Well, that's peculiar.  I try a completely separate circuit with a long extension cord (should have done that before the power supply, I know)...nope.  Same temperature issues.

It COULD be the line power that somehow went noisy...but my other printer has the temperature perfectly stable, so unlikely.  It HAS to be one of the stepper drivers.  I pull all of them (I'll put them back one at a time till I find the bad one).  Nope...it got WORSE.  The temperature was moving +/-5 degrees C between readings.  Obviously not the stepper drivers.

Any ideas?

I have attached the graph from Octoprint showing the difference between AC turned off and on.  The smooth parts are with the AC off, and the not-smooth parts are with AC turned on.  New board and power supply in.