I’m binging up a new printer. I asked klipper to calibrate the PID settings for my printer and it came up with some numbers. I manually put them in my printer.cfg instead of asking klipper to do that but that shouldn’t matter.
THe autocalibrated settings cause the temperature to rise to about 15 degrees below target temp and then it starts “backing off”. I see 0.9 or 0.8 for the PWM value. Next second the log indicates 1.0 again, but the heating already slows.
It WILL reach the target temperature but it will take about three or four minutes longer than what I’d expect it to be capable of.
I then guessed that the D value was “high” and reduced it from 178 to 90. And indeed: This helped a lot. I haven’t tuned it beyond this, but to me it indicates that the autotuning found a D value that was way too high.
Thinking about what I saw during calibration: The heater element is close to its maximum at the temperature that I use to print (240). So during the calibration, you have an ON period that is much shorter than the OFF for the 10 degree swing to happen. Could it be that this results in a way higher D value than what is necessary when HEATING (as opposed to cooling to the desired temperature?)
If that is the case… how about: on the first 10 degreen swing, measure the on/off percentage for the full swing. Nuw instead of using PWM=100% and PWM=0%, use duty +/- (100-duty) (the + one will be 100 again)…
