Unable to tune chamber heater

dc42

@davidewen this is on my list to look at when I am back in the office later this week.

davidewen

Thank you.

FYI; These are the results from my latest tuning attempt.

M303 H2 P1 Y1 S50 ; auto tune heater 2, default PWM (100%), Tuning Cycle Hystresis 1C 50C target

Warning: Auto tune of heater 2 failed due to bad curve fit (R=0.183 K=3.127:0.000 D=7.70)

Tinchus

@davidewen @dc42 There are several post about this issue with heated chambers.
In my case the problem raised with version 3.3 and up. Something changed on that PID implementation. Before that, my heated chamber was using the old PID very good.

After I upgraded, the problem raised. So I tried to autotune again and I was getting the same errors like you. After playing a while, what I did was: I raised deadtime value A LOT. That way, autotune was able to finish.

BUT then it didnt worked: the values I got from autotune, should have worked but every time I tried to heat up the chamber, an error was there: temp rising too slowly.
Running autotune several times worked again, but then again, using the reported values always failed with the same error.

Then I did a trick: I ran autotune but using ONLY 80% of the PWM. Then to the reported values I changed the PWM to 100%. That is the only way I have been able to use PID on the heated chamber

I hope it helps

davidewen

@Tinchus

Thanks for this info.
For whatever reason, it has only partially worked for me.
I can at least finish the tune, but I still can not run the chamber heater even after updating the M307 with the auto tune parameters.

The heater faults out with this error. The values will change based on the M307 settings, but it's always the same error that the temperature is rising too slowly.

Heater 2 fault: temperature rising too slowly: expected 0.14°C/sec measured 0.03°C/sec

davidewen

@Tinchus said in Unable to tune chamber heater:

I raised deadtime value A LOT. That way, autotune was able to finish.

When you said you raised dead time a LOT. Do you remember what value you used?

Tinchus

@davidewen Remeber that I said that raising deatime only worked in order for the autotune to work.

After that, I had the same problem than you: values reported by autotune and updated on the M307 command, never worked for me, it always fail with the "temp rising too slowly" error.

For chamber PID to work (somehow, it works in my case ONLY for a certain temperature range), the trick I found was doing the PID using only 80% of the available PWM. Then when autotne finishes, notice that this 80% value is reported, change that value only to 100%. This worked for me.

deadtime I used for autotuning was like 240.

Also, my chamber heaters are kind of powerfull: I have 2 1000W heaters.

davidewen

@Tinchus

Thanks for the response.
Same for me. I can trick the system to finish the autotune, but still can not get the system to work in normal use with the parameters autotune gives.

Even changing the PWM back to 100% is not working for me.

I'm not sure what values would even be helpful.

I only have (1) 250w heater, so my chamber heats pretty slowly.
I need about 20-30 minutes to get to 50°C.

Tinchus

@davidewen then I would suggest you to do autoune using only 50% of pwm, or less (probably you will nedd to adjust deadtime for autotune to finish?)

davidewen

@Tinchus

Thanks, I will try that now.

davidewen

@dc42

Have you had a chance to look into this?

I have not been able to get anything to work.

I have been able to get the autotune to pass multiple times, but the parameters don't work after entering them in.

hackinistrator

did you try with R value in m307 at 0.001?
also , maybe your heater never reaches the target temp .

davidewen

@hackinistrator

Thanks for the input. I'll give your R0.001 suggestion a try.

Re: heater not reaching temp. It does, and I had no problem with this issue prior to updating FW as also stated by @Tinchus.

After running autotune again, and kind of passing, here's what I received as settings.

The first line is actual autotune settings that don't work.
The second line is modified with the 0.001 setting as suggested.

I was able to complete a 50°C chamber heat with it.

However, I cannot start an actual print unless I pre-heat the chamber. Printing fails with the typical error of expected this heat time, but got this heat time.

;M307 H2 R1.491 K1.725:0.000 D9.50 E1.35 S0.10 B0			; autotune settings
M307 H2 R0.001 K1.725:0.000 D9.50 E1.35 S1.00 B0			; autotune settings

@dc42 - have you had a chance to look into this problem?

davidewen

@dc42, @Tinchus, @hackinistrator, @RogerPodacter

I finally got this to partially work.
I had to re-run RRF config and get a new config.json file.

This happened to me when I changed from a Z-Probe to a Z-Switch. No amount of changing config.g would work until RRF was run.

I still get errors occasionally, but not every time.

@dc42 , I think this still needs to be looked into. It may also be a good idea to figure out why a new config.json file helps alleviate the problems.

Phaedrux

@davidewen said in Unable to tune chamber heater:

It may also be a good idea to figure out why a new config.json file helps alleviate the problems.

I think that's a red herring. That file is only used by the config tool to read your config back into it again so you can make changes. It doesn't get used by RRF in any other way AFAIK.

Tinchus

@davidewen In my case I never used the config tool. I always do configuration changs directly on config.g

dc42

@davidewen what temperature do you heat the chamber to, how long does it take to get there, and what is the ambient temperature? These figures will give me an idea of the heating rate, so that can see whether the current limits are appropriate.

davidewen

@dc42 said in Unable to tune chamber heater:

@davidewen what temperature do you heat the chamber to, how long does it take to get there, and what is the ambient temperature? These figures will give me an idea of the heating rate, so that can see whether the current limits are appropriate.

Thanks for getting back to me.

I generally heat the chamber to 60C.
I haven’t timed it, but I think it takes about 30 minutes.
Ambient temperature is 70F