LDO Part Fan & Thermal Runaway when Printing PETG  

  RSS
Kaiede
(@kaiede)
Active Member

My MK3S has been modded with a Mosquito/BMG extruder, using the Mosquito 50W heater cartridge. No sock on the heat block.

Things have been fine, and I've been getting really good print output for about 2 months until very recently. The stock blower fan has started to make some grindy noises, so I ordered both a FYSTEC and LDO fan since they are pretty cheap. When they arrived, the LDO seemed to have better tolerances in the build, and so that's the one I installed.

Did some tests with PLA and my 0.4mm nozzle, and it was fine with the stock 100% settings, if loud.

But when I started printing PETG parts again, I'm hitting an issue where depending on the sliced Gcode, I can consistently trigger the thermal runaway detection when the fan comes on during the 4th layer. Nozzle temp crashes from 250C to around 210C, can't recover, so the print fails. But it's interesting it depends on the layout of parts on the bed, so I'm guessing it's dependent on if the air is getting reflected back to the nozzle and the block off the part? When prints do succeed, I notice that the nozzle temp variance is up to around 7C which is higher than it has been with the stock fan.

I also noticed that at 100% fan for PLA, printing using my 0.25mm nozzle at 0.15mm layers, it would partially delaminate the second layer while printing the solid infill.  So clearly this thing is stronger than the stock fan and has to be accounted for. I guess I need to figure out what new fan speed settings I need for use for PLA and PETG. I've set PETG to 30-40% instead and now it looks like variance is down to 4C, and the temp only dips to 247C when the fan comes on. 

I'm mostly just curious if these are expected side-effects of going to a higher power fan that's a little too powerful for the application? It seems reasonable that I should have to drop fan speed to get the same result as the stock fan, but I wasn't expecting it to be able to cause the upgraded 50W heater to beg for mercy. I wasn't expecting it to be powerful enough to delaminate PLA either. 

Posted : 08/08/2020 7:50 pm
Karl Herbert
(@karl-herbert)
Famed Member
Posted by: @kaiede

My MK3S has been modded with a Mosquito/BMG extruder, using the Mosquito 50W heater cartridge. No sock on the heat block.

Things have been fine, and I've been getting really good print output for about 2 months until very recently. The stock blower fan has started to make some grindy noises, so I ordered both a FYSTEC and LDO fan since they are pretty cheap. When they arrived, the LDO seemed to have better tolerances in the build, and so that's the one I installed.

Did some tests with PLA and my 0.4mm nozzle, and it was fine with the stock 100% settings, if loud.

But when I started printing PETG parts again, I'm hitting an issue where depending on the sliced Gcode, I can consistently trigger the thermal runaway detection when the fan comes on during the 4th layer. Nozzle temp crashes from 250C to around 210C, can't recover, so the print fails. But it's interesting it depends on the layout of parts on the bed, so I'm guessing it's dependent on if the air is getting reflected back to the nozzle and the block off the part? When prints do succeed, I notice that the nozzle temp variance is up to around 7C which is higher than it has been with the stock fan.

I also noticed that at 100% fan for PLA, printing using my 0.25mm nozzle at 0.15mm layers, it would partially delaminate the second layer while printing the solid infill.  So clearly this thing is stronger than the stock fan and has to be accounted for. I guess I need to figure out what new fan speed settings I need for use for PLA and PETG. I've set PETG to 30-40% instead and now it looks like variance is down to 4C, and the temp only dips to 247C when the fan comes on. 

I'm mostly just curious if these are expected side-effects of going to a higher power fan that's a little too powerful for the application? It seems reasonable that I should have to drop fan speed to get the same result as the stock fan, but I wasn't expecting it to be able to cause the upgraded 50W heater to beg for mercy. I wasn't expecting it to be powerful enough to delaminate PLA either. 

Since you are not using the original components, you will probably have to experiment a bit to find the ideal fan settings for each filament.
The fans are usually cheap products and have tolerances on speed and airflow. I would use identical fans from the same manufacturer as a replacement. So the 10 WAtt additional power won't bring you much, especially since the performances which are indicated are mostly not reached. I would also recommend you to do a pidtuning.

 

Statt zu klagen, dass wir nicht alles haben, was wir wollen, sollten wir lieber dankbar sein, dass wir nicht alles bekommen, was wir verdienen....
Posted : 08/08/2020 10:58 pm
Kaiede
(@kaiede)
Active Member

I have run a PID tune twice. Once after installing the Mosquito, since I used their heater cartridge with it. Once again after hitting the first thermal runaway event where the temperature crashed, after installing the fan.  I was more surprised because the 50W heater is meant to allow the Mosquito to print high-temp materials beyond what the stock hot end is capable of, or make it easier to keep temperatures consistent. So the LDO fan being able to trigger a thermal runaway is... odd.

That said, I'm not thinking this is the fan anymore. I just had another runaway event before the fan even came on. After reseting the printer, the thermistor was reporting over 300C, triggering MAXTEMP. The filament was pouring out of the nozzle, rather than oozing, which makes me think the 300+C reading wasn't a complete lie. Especially since I just spent far too much time chipping off the first layer when normally I can get it to peel nicely off the bed once the PETG has cooled. 

So it seems like either:

  • I screwed up the thermistor connection in the electronics box when I installed the new fan.
  • I screwed up the boron nitride application and it's starting to fail weeks later, causing the thermistor to read the wrong temperatures as the heat block warms up and expands.
  • The thermistor is faulty. 

This would explain the delamination of the PLA too, if it was printing too hot. Either way, I'm gonna have to find time to check it all, and possibly just tear down the hot end and install my spare thermistor. 

Posted : 09/08/2020 1:16 am
Kaiede
(@kaiede)
Active Member

For anyone else that follows up on this thread, here's my conclusion:

First, I checked the connections. Nothing obvious, and it was seated properly on the board.

I went ahead, removed the heat block, and installed my spare heater block, cartridge and thermistor. It was faster than dismantling the current heat block, and I can do a fix-up on the existing heat block later to make it my new spare. When I loaded the same PETG spool that had been giving me trouble before, I could immediately tell the difference in how it was extruding. So far, my variance has dropped down closer to 1-3C as well.

Also, when I did get the heat block off, there was definitely the tell tale signs that things had gotten very hot. I am like going to have to keep on eye on things in case my backup thermistor does the same thing. I bought it from the same source as the one I've been using at roughly the same time. 

It does look like it was having issues getting good thermal conductivity with the heat block, would start to under-read, causing the control board to crank up the power. And it's obvious now that it was also printing at a higher temperature than intended when not triggering the runaway protection, which is bad news since who knows what the actual temperature was. 

I'll have to keep an eye on the new setup, just to make sure it doesn't repeat the previous issues.

Posted : 11/08/2020 9:49 pm
Karl Herbert
(@karl-herbert)
Famed Member

@kaiede

Glad you were able to locate the error!
Once again the thermistor - not untypical.
If it delivers wrong values of course, the temperatures or the heating power are calculated wrong.
I would lubricate all components around the heating block with copper or ceramic paste before you mount it, including the screws. Then everything can be disassembled more easily.

good luck!!

Statt zu klagen, dass wir nicht alles haben, was wir wollen, sollten wir lieber dankbar sein, dass wir nicht alles bekommen, was wir verdienen....
Posted : 11/08/2020 10:20 pm
Kaiede
(@kaiede)
Active Member

The Mosquito requires you to use thermal paste to get the cartridge and thermistor mounted properly, since it doesn’t use clamping force like the V6 heat block design. Slice Engineering recommends/sells Boron Nitride based paste, which is what I used. The catch is that the stuff sold by Slice solidifies nicely at room temperatures once set, acting a bit like “loctite blue”. It’s just easier to prep a new heat block and swap it once the new paste is set, and free the old stuff from the paste later.

I am left wondering that since it does solidify at room temperature, if the temperature cycle you get with each print might have let things shift. The thermistor once I removed it was still pretty much covered in paste, telling me it was probably wholly dependent on the paste to be able to read the temperature of the heat block properly.

 

Posted : 12/08/2020 1:02 am
Kaiede
(@kaiede)
Active Member

Hmm, the first print was fine, something's still causing the new thermistor to mis-read badly when the print fan comes on, even at only 40%. Need to revisit. The spike at >260C is after the reset. 

 

Posted : 12/08/2020 5:51 pm
Kaiede
(@kaiede)
Active Member

So, the only thing I can think of is that I somehow got two bum thermistors from the same seller. One lasted a couple months, the other got one successful print off before it too was providing false readings causing the printer to pump far too much heat into the heat block. The attached file shows some of the issues I've been seeing. A nearly 40C temp drop during the purge line, which shouldn't even be possible.  Then normal behavior for a while, and then another sudden drop an eventually a thermal runaway error. And always, when I reset, the temp reads higher than target. The only thing I can think of is that the movement itself is exposing wiring issues in these thermistors. Since it seems to start with these partial drop-outs of temp readings, followed by more dangerous problems when it thinks it is behaving correctly:

I managed to somehow get the heat block on the Mosquito hot enough to discolor the nickel coating during a PID tune, because the thermistor thought it never got hotter than 239C when tuning for 240C, and clearly it was much hotter than that.  

I also managed to get PETG baked onto the textured build plate. It's bad enough that I am not going to get it off without damaging that side of the plate. And it will affect printing on the back side of the plate until I can remove it. I wasn't even aware it was possible to bond PETG to the textured plate like this. 

I'm currently breaking in a thermistor from Slice Engineering, where things are looking much better. It's 100% overkill since I don't need a 450C-rated thermistor, and the connector is placed in an unfortunate spot that makes it tricky to mount without placing pressure on the thermistor wires (potentially causing the problem I'm trying to solve), but so far it is looking good. I mostly grabbed it since it should be a match for the Mosquito heat block, and if I was to rule out another issue (like the Einsy board itself), it was faster than getting a Prusa thermistor replacement. I'd attach a comparison graph, but it won't let me attach more than one file. But I'm seeing very clean "bounce" when reaching temperature, and the fan coming on poses no problem, like this thing used to behave when I installed the Mosquito/BMG originally. Fingers crossed it continues to behave this time.

I'll probably add a couple thermistors to my next order with Prusa, so I've got spares that are a little more affordable than the Slice HT thermistor.

Posted : 16/08/2020 7:06 am
Talaminator
(@andrea-t)
Active Member

I had the same problem when i switched from stock to LDO fan (and the "Bear upgrade" nozzle fan duct at the same time) , doesn't matter the filament type.

i have to say that from the first moment the new setup seemed to blow much more air, and in fact i was consistently getting thermal runaways too.

Considering that after reinstalling the silicone sock that never happened again (and everything is working fine) i really think it was just a matter of air-flow amount.

Posted : 30/08/2020 11:55 pm
Share:

Please Login or Register