M307: bad model parameters with Beta3.2 LPC

MarMed

With beta3.2 I am getting the following error from setting the model parameters for my chamber heater

M307 H2 B1 A29.9 C1151.3 D55
Error: M307: bad model parameters

on previous versions the parameters would set correctly.

11/19/2020, 11:37:31 AM	M307 H2
Heater 2 model: gain 29.9, time constant 1151.3, dead time 55.0, max PWM 1.00, calibration voltage 0.0, mode bang-bang
11/19/2020, 11:37:28 AM	M307 H2 B1 A29.9 C1151.3 D55

just to note, the parameters used is from previous RRF3.2 betas, but with the bad curve error, due to dead time being negative I used the recommendation here to measure and set the dead time manually.

Additionally I receive a bad curve fit using the new tuning algorithm, is there a similar workaround using the new tuning?

RRF3.2 Beta3.2 on LPC

Thanks

Veti

thin post needs to be moved to the lpc sub forum

dc42

@MarMed, when you ran heater tuning under 3.2beta3.2, what parameters did it report when it gave the bad curve fit message?

The gain of your heater looks very low and the dead time very high.

MarMed

@dc42
This is on beta4.1

Warning: Auto tune of heater 2 failed due to bad curve fit (R=0.030, C=0.001:0.000, D=2.3)

I think the problem may be a combination of a very low power heater(250w), and the situation that the heated bed gets the chamber close to the target temperature already, around 50C and I then use the the chamber heater to reach a stable 60C.

I've tried many combinations when tuning, with hotend and bed heater off, with hotend and bed heater on, and setting the ambient temperature. Each tuning takes 5h+ before it fails.

Any idea why my old heater parameters is giving me an error now?

M307 H2 B1 A29.9 C1151.3 D55
Error: M307: bad model parameters

is there manual parameters that might work with my setup in 3.2?

Help is appreciated.

dc42

I have identified the problem and implemented a fix in 3.2RC1.