Hotend PID tuning method?

New patch, new toys!

New patch lets you tune the hotend pid

  • Support “M301” G-code for setting up Hotend PID.
    • Parameters
      • [P] [I<Value>] [D] for setting up Hotend PID

Lets discuss how to come up with the optimal values for this?

I know Teaching tech has an auto tune guide but apparently this is manually only at this time, per the notes.

I dont really suspect a tremendous difference, but would like to optimize everything i can!

1 Like

Marlin has a autotune routine, this worked pretty fine on my original.

Maybe I haven’t looked through the 3DP topics carefully enough, but I don’t recall mentions of the temperature fluctuating or otherwise being inaccurate.
So even though there might be some potential to improve the PID parameters, I’m not sure if this will really impact print quality.
Perhaps this is more relevant for a planned dual-extruder expansion?

maybe, but i dont think my hotend gets as hot as it purports to.

My cheapo thermal imager can’t ever get a good read on the hotend. :thinking:

Recently i tried pid autotuning marlin command (M303 C5 E0 S240), it doesn´t work.

After that i saw the pid values set up in the firmware and was wondering:
M301 P13.00 I0.02 D106.25
In a older firmware release (where pid was not enabled) it was set to:
M301 P13.00 I0.10 D17.00

There is a huge difference in the two sets of D. The new value makes for me a difference of 3.5°C if i turn the fan on (M106), the old value to about 2°C.

Even the heating needs so much longer with a D of 106.25.
I guess this D and no autotune are kind of a bug.

I set my machine to the old M301 P13.00 I0.10 D17.00 and it works better with this…

I tried both values, the old and the new one. I found no visual differences with PLA+. But the readings on the touch screen were quite different. Printing at 220°C I got values of 216…224°C with the old settings and 219…220°C with the new settings during extrusion. I will stick to the new values as these seem to give me more stable results.

Just to make sure there is no confusion about this:
Old settings: M301 P13.00 I0.10 D17.00 (FW 1.10.1)
New settings: M301 P13.00 I0.02 D106.25 (FW 1.12.2)

2 Likes

Hi all! After years there is still not PID Autotune working as it seems. My printer doesn’t give anything back and I can’t read the values.

In this thread here I discussed the problems with my inconsistencies and I only found the reason to be the fluctuating temperature which was jumping up and down 4deg around the setting point (197-202deg)

Now from this thread I took the D106.25 value from FW 1.12.2 saved them into the EPROM. E voila- the Benchy looks way more consistent. It takes longer to reach the exact temp of lets say 210- it comes to 209, stays for some minutes and then gradually hops to 210 until it almost stays there the whole time

For me this is the reason to say: WE NEED PID AUTOTUNE

When you say it doesn’t give anything back, do you mean in the command terminal? What are you using to send commands?

(Luban terminal doesn’t work for getting replies.)

Oh does it? Yes I was using the Luban Terminal for giving commands.
Which Interface can be used instead? Prompterface or something and then connect via Cable?

Does the normal AutoPID command works in general?

I don’t know about the SM2.0. But I do know they added PID tuning commands to the J1 this year.

And yes, you need to use Pronterface or something, connected to the back USB port. On the J1, at least. I dunno how folks get a two-way terminal on the 2.0.

As you were trying to do, you need to run the command, then set the PID parameters and save them to ROM. This year they later added an actual PID tuning button the J1 screen interface even.

In any case, you will for sure not get any useful responses using the Luban terminal! :sweat_smile:

1 Like

The Luban terminal does work actually, but only for certain commands. Some limited responses can be obtained over WiFi, with more only available when connected via USB.