Temperature swings wildly on primary extruder

[insta]

Dual-extrusion M2, only 1 extruder in use (other has the nozzle tucked up and unpowered), while printing I get wildly inconsistent temperatures. Like, with a target of 210C, it will climb to 213C, drop down to 210C and hold that for 4-5 minutes, then drop down slowly to 190, 180C, occasionally down to 160C if the fan is on! During this time the heater LED is full-on, and all the sudden seconds later the hotend is 210C again. I've done the PID autotune with M303 C8 S180 and saved the values to the firmware. Running the autotune gave nearly identical values, and I did the autotune with the bed fan on.

I've replaced the thermistor, reseated the thermistor and heater cartridge in the V4, and wiggled all the wiring to check for loose connections. I've never gotten the MINTEMP error. I just last night flashed the Marlin/Viki 24v firmware for the single-extruder machine and no change in behavior.

Ideas?

[PcS]

Still sounds like a shorting wire . Ohms test the rambo plug in for thermistor and wiggle wires. See if it swings around .

[jimc]

i had the same thing going on with the bed temp. changed the thermistor, wiggled all the wires and i saw nothing but it did end up being a bad wire on the thermistor cable in the harness down to the rambo. it must have just flexed too many times and had a break but not enough to completely break contact. just added resistance to that circuit throwing off the true thermistor reading

[insta]

I'm worried that's what it may be, but I completely replaced the wiring harness when I installed the dual extruder upgrade

[insta]

So this seems to be fixed...

I re-flashed the modified MG firmware (modified for proper dual-extruders with Viki) and the primary was reading random numbers between 170 and 210C immediately after powering up. Wanting to test if it was the firmware or the hardware, I swapped the positions of the 1st and 2nd thermistors on the RAMBo, and it made the second extruder's temp swing wildly. So, I swapped them back, and it immediately read room temperature.

Since we have negative coefficient thermistors, I'm guessing some conductive dust was blown into the RAMBo and providing a lower resistance making it read higher temp. Moving the plugs around cleared this fault and now I'm good for a random time again

[pyronaught]

I just started having this same problem today. I was about an hour into a print job and the M2 just froze up, then the communication commands on the S3D screen were scrolling really fast to the point where I could not read any of the commands. I just rebooted everything and now the extruder temp jumps all over the place. Sometimes it reads as high as 260, other times 160. Currently it seems to be stuck around 250 to 260 range, even when first powering up the machine.

Could it be that the reason the machine froze up is that the extruder temp reading went beyond some upper limit that stops the print job and waits for it to cool?

I've had conductive dust literally short out my PC and cause random crashes, so I know that can be a real issue. Blowing out the computer with an air compressor always fixes it. It usually takes several years of use before enough dust collects to do that though. I've only had the M2 about five months now.

[insta]

The electronics are right above the linear rails, there might be microscopic shavings coming out as they wear in? Try canned air on the electronics and plug/unplug the thermistors a few times from RAMBo and see what changes.

[ednisley]

there might be microscopic shavings coming out as they wear in?

Nope, those beautiful linear guides don't wear like that, plus they should be well-lubricated: no shavings, no dust.

Both of you have failed thermistor wiring: an open circuit will be near 0 °C and a short circuit reads as an absurdly high temperature. An intermittent failure will vary randomly between (nearly) the correct temperature and something absurd. You can have both failures at the same time.

For open circuit failures, the insulation tends to hold the wire ends together and hide the damage, but some detective work with a multimeter will tell you where the problem lies.

Given that you're seeing short circuits, I'd expect a failure at the thermistor, where those hair-fine wires emerge from the sensing element. They're not insulated and, with just a bit of motion, they can rub together. Again, a bit of multimeter tracing while poking around will reveal the problem.

If one of the wires contacts the hot end, you'll see weird temperature variations due to the unpredictable voltage on the metallic parts: the Arduino is sensing a very small voltage, so any interference will produce odd results.

It's as simple as that, really and truly...

[pyronaught]

The electronics are right above the linear rails, there might be microscopic shavings coming out as they wear in? Try canned air on the electronics and plug/unplug the thermistors a few times from RAMBo and see what changes.

My RAMBO is not attached to the usual spot on the M2, I moved it outside my enclosed chamber to keep it out of the heat. So it isn't exposed to glue stick scrapings, hair spray vapor, acetone vapor or any other kind of debris around the machine.

step15.jpg (332 KiB) Viewed 10856 times

[pyronaught]

What should the thermistor resistance be at around room temp? I'm reading values in the 200 ohm range. I read the same value from the connector that plugs into the RAMBO that I read from the connector that goes to the thermistor. If I switch the HBP and EXT thermistors around, the bogus reading swiches to the HBP, so it is not anything on the RAMBO causing the problem.

If I unplug the thermistor while running, the reading goes to zero. If I plug it back in, it goes to 260+. If I turn on the heat, the temp reading actually starts decreasing as the temperature increases. I think the thermistor is just hosed.