Heater faults with numerous different heaters and thermistors

Foden

@foden screenshots are here https://imgur.com/a/UDpWWIV
looking at the temp chart its pretty stable until it fails, so could indeed be the fan kicking in. Have included a shot of what the fan looks like, it should blow air onto the part but could also be bouncing back onto the hot end

dc42

Unless you have configured your slicer for the fan to turn on later, the print cooling fan normally turns on at the start of the second layer. But it looks like you printed 10 layers before the problem occurred.

deckingman

@foden It's fairly common for the air flow to be directed back off of the build plate or the printed part itself, onto the nozzle. A silicon sock usually helps though. But until you test with the fan off, all else is just speculation.

If turning the fan off cures the problem, then simplest thing to do might be just to limit the fan speed to (say) 50% using the X parameter in M106 https://duet3d.dozuki.com/Wiki/GCode#Section_M106_Fan_On

Edit. That's good timing - DC42 and I both posted at exactly the same time.

Foden

@dc42 Currently cooling settings are

1 0%
3 50%
5 75%
7 100%

so by layer 10 it may be at the tipping point temp wise

Foden

@deckingman Hopefully that's it and I can start to grow my hair back ;o) will no later today

Foden

Not conclusive, as I haven't had chance to do a long test print. However, after 15 mins with the fan on 50% the printer is 'as happy as Larry' - and so am I. Hopefully I will have chance to do a full test tomorrow evening. Feel a little dumb for not seeing the correlation between cooling fan and the issue...facepalm

Regardless THANK YOU both VERY much indeed for pointing me in the right direction!!!

Hopefully will be able to update fully late tomorrow

Foden

Now it's conclusive - 100% fan back-blow was the issue. Anything above 8% and it would create a heater fault. Have to look to a different fan duct as the semi-circular one I'm using seems to make it worse. Without a fan at all you eventually get a large mass of filament and no print :o(

Thank you both so much for your pointers. I can now move on to the next hurdle, what ever that will be ;o)

OK, so where were the plans for the better fan.....

Thanks guys!

Foden

@foden Ahhh Nutz!

Not completely solved, even on 5% fan i get heater failure after about 15mins. Just ran another text with no fan duct, so air goes straight onto the bed behind the hotend. Fan was at 5% and i didn't use any filament. Heater fault at 1 hr 1 min 14 seconds, will repeat with the fan off and see what happens

dc42

Does the temperature graph show a gradual deviation from the set temperature, or a sudden change?

Foden

@dc42 Hi dc42, it was a sudden change - currnelty into the test, 52 mins, with no fan or filament

Foden

@foden This is what it looks like so far https://imgur.com/a/ycVP0K4. When the fan is on there is a very slight tremor/wave in the hot end temp line

Foden

@foden Now past where it failed last time, will leave it until the end or it fails 32% to go yet

The hot end I'm using has a e3d v6 block and thermistor, with the thermistor at the back nearest the fan. I may tril rotating the unit 180o to see what that does

Foden

@foden completed print run without issue

Phaedrux

Silicone sock?

Foden

@phaedrux Yes, there's a silicone sock on the heater block

Phaedrux

Ah yes, sorry I missed that above.

If it were the fan you should be able to force a fault by heating up and then maxing the fan. If you can't force an error with worst case scenario I think it would be even harder to get a fault in normal operation. That would lead me to think a intermittent writing fault has developed.

I'd try with a new thermistor and writing. If it doesn't help, at least you have a spare.

Foden

@phaedrux I've gone through about 1/2 a dozen heaters and thermisters and several heat blocks all with the same issue. I have LOTS of spares now. If i have the fan at full speed it fails - the link is obvious now but it wasnt until deckingman and dc42 pointed it out.

My hot end assembly is a quick-removal type and I have tried 3 different versions of that, each with their own RJ45 pinvout board for cabling . It's definitely the fan cooling bouncing back onto the head/thermister.

Last test for today is running with fan at 35%, i rotated the hotend 90o and its holding up way better than before. I also up'd the print speed to 200% to stress everything else, loose cables, PSU etc - all good so far

Foden

@foden fan at 45%

Foden

@foden fan at 55% still holding temp

Foden

@foden hmmm, another thought struck me as I'm doing this test. My previous cooling,,that failed, was in steps of 25%. Now I'm doing it in 5% steps, possibly the fan fired up way too fast for the thermister to react? I'm at 75% currently - may try that tomorrow as I'm fed up of testing and its late