IMPORTANT ISSUE with bed temp controller

jrockland

@Veti you are right, they are ntc 3950. Ill have to find out how to change that

Veti

@jrockland said in IMPORTANT ISSUE with bed temp controller:

ntc 3950

they are not the best and become inaccurate at high temperatures.

jrockland

Anyway thank for all the answers, tones of thing to look at;
like why my board seem to work the opposite way ? (compared to my other 6hc's)
Stop.G files (that files is empty ??)
thermistor: wrong model ? (ill have to find how to change that)
port 0.out instead of 3.out (will move some stuff around)
Also I will add an external fail safe (I should had done that long time ago)

Just to reassure anyone, the relays does shut down as soon as I send a M999/M112 or shut the board off. even if they are written as "!out.3".... maybe the connection to the board is wrong.

jay_s_uk

if you use out0 don't forget to add power in for it

jrockland

@Veti I know, but with what I use them for they are easy to get and cheap..: I prototype parts so I actually destroy parts as a daily task. I basically apply change, then run the project until it break. then make it better. I crashed so many hot ends I buy those sensors in 50 packs XD

jrockland

@jay_s_uk wait.. just want to make sure im not wrong here.. i mean C"io0.out".. my bad. too much power out of OUT0.

Luke'sLaboratory

@jrockland said in IMPORTANT ISSUE with bed temp controller:

the relays does shut down as soon as I send a M999/M112 or shut the board off.

This is positive.

Could you post model #'s or pictures of the relays used? Still want to get to the bottom of why you overheated by a considerable margin

Veti

@jrockland
you forgot the most imporant part. the frequency for the ssr.

jrockland

@Luke-sLaboratory let me check, Ill have to open an other part of the machine for that.

jrockland

@Luke-sLaboratory Actually when I read the code, bang bang mode wasn't deactivate, might explain a lot. it always been deactivated but that specific machine crashed hard before the last update and we had to re write the whole thing.

Luke'sLaboratory

@jrockland If the previous configuration was bang-bang, and instead we're using PWM @500hz, that could be a huge indicator of an issue.

jrockland

@Luke-sLaboratory Ill have to read more about bang bang but if I check on my other machines it is deactivated everywhere

dc42

Using an inverted heater output isn't safe. When you power up, you are relying on the Duet booting up and running that M950 command to turn the heater off. If any of that fails (for example, something prevents the Pi from running, e.g. the SD card goes bad), then the heater will remain on.

What connection are you using that requires an inverted heater output?

Luke'sLaboratory

@jrockland info on the relays would still be useful, but yes, probably should look into disabling PWM control if it was previously operated bang-bang and it was successful.

jrockland

@dc42 well that is the thing, it is inverted so it doesn't turn on as soon as I turn the board on. that is really weird. Ill have the board sent to my tech so he can look at it.

if I remove the inversion the board feed power as soon as it turn on (checked with a meter)

Veti

@jrockland said in IMPORTANT ISSUE with bed temp controller:

well that is the thing, it is inverted so it doesn't turn on as soon as I turn the board on.

this suggests you got the wrong ssr.

jrockland

@Veti ssr's are out of the equation, they are disconnected. Im testing the board whit a multimeter

dc42

@jrockland said in IMPORTANT ISSUE with bed temp controller:

@dc42 well that is the thing, it is inverted so it doesn't turn on as soon as I turn the board on. that is really weird. Ill have the board sent to my tech so he can look at it.

if I remove the inversion the board feed power as soon as it turn on (checked with a meter)

We really need to know what sort of relay or SSR you are using, and how you have connected it to the Duet. If it's a relay board then it may be better to connect it to an IO_x_OUT pin instead; but a SSR is better because it allows you to use PWM and hence PID.

Luke'sLaboratory

@dc42 He's talking about the actual board output, but I don't remember if i'm able to measure V_in across inactive outputs or not, since the ground is switched

jrockland

@dc42 on that specific machine we still use the old SRD-05VDC-SL-C, apparently we haven't blown them up yet. so no SSR'S on that one.

problem is probably in my gcode.