Temperature oversampling

[3poro 9]

For the past couple of months the nozzle temperature reading on my UM2 LCD has started to fluctuate much more than it used to. Now it can show temperature changes of +-10 degrees (!) within a couple of seconds. Obviously the actual temperature of the nozzle can't change that quickly. I think the situation got worse after some firmware update during the summer, but I can't be 100% sure.

 

The problem for me is - sometimes the printer stops and gives me "ERROR - STOPPED; Temp sensor" message. This is obviously a false alarm.

 

I found an interesting note here: https://github.com/Ultimaker/Ultimaker2Marlin

 

"Temperature Oversampling:

To reduce noise and make the PID-differential term more useful, 16 ADC conversion results are averaged."

Depending on how frequently the readings are taken, I should probably simply increase the number of results averaged. Now - how to do this in practice? I haven't yet had the need to hack my UM2 firmware, so is there some "quick-hacking for dummies" intro somewhere?

What is the threshold temperature for the error above?

 

[christopher-luy 0]

Before going to this extreme, be sure to check the themocouple connections. I had a loose thermocouple wire on one of the connectors that caused the same wild temperature fluctuations. Could possibly also be a failing thermocouple.

 

[Daid 306]

The UM2 hotend can heat-up about 3.5 degree per second.

I'm pretty sure you should be looking at a mechanical problem, trying to hide this with firmware tweaks will only make it worse in the end.

 

[gr5 2,173]

It can go up 10 degrees in 3 seconds but it takes much longer to go down 10 degrees.

Is a "couple of seconds" 3 seconds? Or does it jump faster. 3 seconds is a long time.

If it's changing much faster than 10 degrees in 3 seconds then probably a loose wire.

 

[3poro 9]

Thank you for your advice - I really appreciate!

The problem turned out to be caused by myself - I shouldn't have cast any doubt on firmware before investigating the root of the problem more carefully. As I have added airflow for my print head, I had ended up draining too much current from the fan output:

http://umforum.ultimaker.com/index.php?/topic/6307-um2-twin-tornadoes/

I added a note about the potential problem to the end of the thread. Now, after switching the fans to a separate power source, the temperature readings are steady again.

Thank you again for your help. As I had already checked the system for connection problems, your messages convinced me there had to be something more fundamentally wrong.