H1 giving erratic reads while printing.

[pao_wikhan]

seems like the oscillations start at 200degrees, brought the bed down and still the same.

[dc42]

Looks like you changed from using heater 1 output to heater 2. Did you run auto tuning on heater 2, with a target temperature around 200C or a little higher?

What type of hot end do you have, and what are your M307 parameters for heater 2? Send M307 H2 to report them.

[pao_wikhan]

@dc42 said in H1 giving erratic reads while printing.:

M307 H2

tuned to 260, should i tune to a little lower value?

M307 H2
Heater 2 model: gain 347.5, time constant 196.2, dead time 6.2, max PWM 0.50, calibration voltage 24.1, mode PID, inverted no, frequency default
Computed PID parameters for setpoint change: P16.2, I0.462, D70.5
Computed PID parameters for load change: P16.

[dc42]

260C should be Ok for tuning. But why have you set a PWM limit of 0.5?

[pao_wikhan]

@dc42
it gives the overpowered error during tuning.
im using a 24V 50W heater

[deckingman]

@pao_wikhan said in H1 giving erratic reads while printing.:

@dc42
it gives the overpowered error during tuning.
im using a 24V 50W heater

Ahhh. That's most likely the root cause of the problem. I once tried to "tame" an overpowered heater and never did have much success in reducing the oscillations. The solution for me was to use a more appropriately sized heater cartridge.

[pao_wikhan]

@deckingman

i think my issue is more than that..while printing it gives out erratic values from 210 down to 160-140, stopping the print

[deckingman]

@pao_wikhan said in H1 giving erratic reads while printing.:

@deckingman

i think my issue is more than that..while printing it gives out erratic values from 210 down to 160-140, stopping the print

I'd say, looking at the graph you posted, that you have two issues. The general +/- 2 to5 deg or so oscillation is likely just the heater being too powerful. The rapid drops in temperature can't be "real". That is to say, the temperature can't really drop by 50 degrees in such short a time unless you actually remove the hot end and drop it into dry ice or some such So that must be a wiring issue or electrical noise.

[pao_wikhan]

thats the thing. it was working for the past few month, didnt touch anything print print print. now this giving me that erratic motion is weird. so i replaced the sensor same output, and did the stuff recommended here.
just finished rewiring really, so im testing.

[pao_wikhan]

@deckingman
rewiring is the same..oscillations..will try using e3d generic 24v 30W heater..then tune to max P1

[pao_wikhan]

Thanks @dc42 @deckingman @T3P3Tony
i fixed it, seems like the thermistor is getting pulled and something was shorting the 5v rail. still using the new cartridge and ntc sensor (the pt100 was giving 2000degrees, but rtc1 is working when i place a 100ohm resistor) will troubleshoot that in a new chapter.
oh and a new psu too, seems like the psu was heavily grounded.

[DocTrucker]

What's the logic behind having a 0.5 duty limit?

As I understand it this is only really useful if the heater responds too quick for the Duet to control it. If the mosfets fail on the 0.5 setting won't happen anyway and the heaters will be locked on 100%. If the thermistor drops out of its hole and gives a false low won't the heater fault detection pick this up quickly?

I did however have it on my mind an age ago to go seriously overpowered and fuse it so if it went 100% it would pop a fuse. Still want to try that. May need a little coil in the circuit to tame the spikes a little, with a corresponding flywheel diode.

[pao_wikhan]

okay its happening again. now im not sure if my duet is broken

[dc42]

@pao_wikhan said in H1 giving erratic reads while printing.:

okay its happening again. now im not sure if my duet is broken

Are you using the thermistor or the PT100 now?

[pao_wikhan]

@dc42
new thermistor, on that pic..
i replaced it with the one included in e3d v6, seems stable so far. but i didnt connect it to the daughter board as its not giving any values.

[pao_wikhan]

if it happens again, i think the heatblock screw holding the sensor is not gripping correctly, but that is a long shot. (my theory that is.)

[DocTrucker]

If you over tightened that it may edge towards occasional shorts. Too loose wouldn't cause spikes like that. That's got to be an electrical or control board issue. I doubt there are real situations survivable for the temp sensor where it could loose heat that quick - even if yanked out of the heater block.

[dc42]

@pao_wikhan said in H1 giving erratic reads while printing.:

@dc42
new thermistor, on that pic..
i replaced it with the one included in e3d v6, seems stable so far. but i didnt connect it to the daughter board as its not giving any values.

If it's a thermistor (not a PT100) then negative-going spikes like that suggest a bad connection, or just possibly static discharge from the hot end metalwork to the thermistor wire. But I think they are too frequent to be static discharge.

[pao_wikhan]

@dc42

well 4hrs into printing, i think the e3d sensor (blue wire) did the trick. odd though cause when i measure the resistance its like 70k-100kohm and not the pt100 value of below 100ohms. its now cleaner than ever. i also replaced the cartridge and moved back to a newer 24V 50W ceramic type with 100% pwm.