Temp Readings not properly reported

[boreal trout]

Hello everyone,
i have a problem with my klipper printer (Artillery SW X1 MKS Gen L + Raspberry Zero 2W)
The problem is, that irregularly the temp measurment samples somehow are not reportet/updated in webinterface (and also not in Klipper itself). Actually i suspect the whole Klipper SW to freeze for a moment.

This leads to PWM of heaters not updating accordingly and thus leads to temp over or undershoots. The print then fails once the temp readouts come through again.

You might think this must be a loose connection on tempsensor but the strange thing ist, that once i restart firmware all the previously missing temp samples are present.

Is it possible, that my Hardware is noch up to the task?
PID tuning is done.

Pic is the temp readout after FW restart.

[winter shore]

Ok, sorry to say this is impossibly, the actual PID regulation and readout of temps is done direct in your mcu uC. And then it is send through the Klipper host -> moonraker -> mainsail UI. So if your theorie would be correct all would freeze (movement, both heaters, fans …)

It is more likely that you have somewhere a partial broken cable that at some spots lead to that issue

[boreal trout]

I already resoldered all connector boards. I will now try switching cables between tempsensor and LED. Then try again.

[winter shore]

Wait are the LED wires close to it. Are you using PWM on the LED

Is there any relation beteen switching on/off or changing brightness of the LEDS

[boreal trout]

I am just about to check the wires so i'm not sure yet about location.
Considering PWM I dont really know but since I don't care for variable brightness i didn't implement anything PWM related to LEDs

Only On/Off and colors

[winter shore]

ON/OFF and colors 😉 -> That means you have a higher frequency data line transmitting that informations

[boreal trout]

ok so what I could try is to code out and remove the LEDs. It's only a gimmick anyway. I will also try to find cables from led to use for tempsensor.
Would you also recommend replacing tempsensor itself?

[winter shore]

We are in the debug face. I would remove nothing yet. I would recomand to simple comment out the led out of the printer.cfg and see if the issue still exists

If no -> we have something we can look at
if yes -> we need to go one and check the next thing

I would also try to check if that issue occur at always the same area of head placement

[boreal trout]

these issues are quite frustrating but it's great to get help here 🙂
I will go through your suggestions and come back, thanks!

[winter shore]

Yes I know it could be frustrating ... especial finding cable breaks in moving cables

also your bed curve is a little strange as it raises

[boreal trout]

That was the switch from first to second layer so 100->110°C

[winter shore]

ahh ok is the fan kicking in there too

if yes does it maybe blow to usch on the heater block and cools it down too much

hav you done your extruder pid tune with our without fan on

[boreal trout]

That's right, fan is running from layer 2.
However i already observed inconsistent temps on layer 1. Just not that bad to end the print.

I just finished switching the wires around and the issue persists (no print, just filament swap already showed strange temp behaviour).

I have now commented out LEDs in code and physically removed it.

BUT:
While filament change I now have on screenshot what i tried to describe in words.
The three pics show 3 consecutive samples, however the time between these is up to 1 Minute!
Typical sample rates should be around 1 Second or so.

Strange thing is, that after Firmware restart the curve is completely smooth, all samples are there in a expected sample rate

Maybe something with my Klipper Install is faulty? Is that a possibility?