Please help UN-confuse an ole man!

Back story, recently, I had to replace the heat block, nozzle, heatbreak on my Ender 5.
As a result, while surfing, ran across a video explaining that you should do a PID auto tune in Marlin which is installed on my mainboard.
(Note of stupidity here: I did NOT save my existing PID values)
But the printer was working and I thought it might improve my prints on this unit which is dedicated to PETG.
So, we ran the autotune and saved the new values.
Now, we can’t print as the printer will show a thermal runaway shortly after the print starts after a layer or two.
After a bunch of trying to troubleshoot, and noticing that we had no problems while running Z probe wizard at PLA temps (which we ran after each teardown and check of the hotend), began to think the recent PID tuning might be a problem.

Long story short, because we ran across the idea that you should run the part cooling fan at 100% instead of the 0% that we had previously done, we ran the test again using the following commands.

m106 s255 ; Turn the part cooling fan on to 100%
m303 E0 S220 C10 U1 :Run the test and save the values based on a temp of 200°C

and to finish up make sure the parameters were saved (which they were)

OK, now the fun begins because being an ex engineer, ran the test again to see how close the results would be.

What a mess!
P I D

Run 1 69.37 13.34 90.19
Run 2 118.55 20.3 173.08
Run 3 48.84 5.36 111.36
Run 4 95.07 16.28 138.8
Run 5 72.21 11.88 109.76

Because the results between the first and second test were so wildly distributed, we ran more!
Based on these runs, I don’t trust any of them! A minor variation of the values could be expected, but these give me no confidence and would like some ideas on where to go from here! (Note:all test run same afternoon within about 2 hours)

One thought is to try the print that started throwing the thermal runaways with each of the above set of values:{(
But IF I can get a print to finish, actually how good are the values and how could I futher tweak them to make it better. Back around 2000, when I set the PID on servos for my cnc, I had a guideline to follow, but for a hotend, have no idea on how to go about it.

So some help for this ole guy from the group would be nice!

Country Bubba

Far from an expert here but to me it seems like your thermistor is loose, not making good contact or bad wire connection. I would just double check them.

1 Like

yeah, @cntbill has a point. If the values are so off the charts it usually is a broken wire or a loose sensor.

If the sensor is loose you are basically measuring the air and depending on the movement and/or the part cooling fan the air gets shuffled in and out of the heater block causing massive differences in temperature.

Even if the temperature is constant you have a heater temperature which is way higher than the printer is showing and your filament will be cooked and as a result clogging is happening after a while.

makes sense now that you mention it and as many times as I have messed with it, might just need to replace the thermistor ;as one of the things I did was check the screw and it was tight!

Will get one on order!

Thanks for your thoughts.

I had a number of Ender 5s with many hot ends. One thing that caused me errors was not having a sock on the nozzle. They will help keep the hot end from cooling too fast. My experience was that after I entered the new PID values, I would rerun and see if they changed. Sometimes you can adjust them and get closer to the optimum setting

I have the sock on. But one thing I got to thinking about was it appeared the hole for the screw that holds the thermistor in is a tad bit more away from the thermistor hole and the screw might not be actually holding it in? I want to check that first before I put a new thermistor in.

Yeah. There also exist thermistors which have the same diameter than the hole.

I never tried these as I bough a 50 part bag of the other ones.

Another good solution is to fill the hole with standard thermal paste before pushing the sensor in. That way the thermal contact is always perfect and the wires have less stress as the thermistor cannot wiggle around within the heater block.

Excellent idea, will be sure to do that.
Thanks for mentioning.

Just as a followup, ya’ll nailed it. Replaced the thermistor and all seems to be working like it should now. Ran a series of auto pid runs and got consistent results.

Thanks for everyones help.

1 Like