M2 Fire Report

[ednisley]

the MOSFET itself sticking on

Pre-zackly!

Although you'd prefer transistors to fail open, the usual failure welds the source to the drain as a dead short: whatever's connected to the transistor will be stuck on forevermore.

If that's a heater, it runs at 100% duty cycle until something else fails / burns through / gives out, because the firmware has completely lost control.

The other common failure has the firmware locking up with the transistor turned on. After the firmware vanishes into its own navel, it doesn't pay attention to anything else and won't turn the transistor off no matter what the sensors report. From the outside, both failures look pretty much the same, with full-on heating until something else burns out. With a firmware lockup, however, the controller survives: reboot that sucker, it looks around, goes "Who? Me?", then starts running like nothing happened.

[hybridprinter]

this thread is beginning to make it obvious that someone needs to design/build/sell an add-on safety solution!!!

[ednisley]

an add-on safety solution

-- my overtly biased opinion --

Having lived through the Thing-O-Matic years and watched the debacle of their "safety circuit", with iteratively wrong versions loosely based on my design (*), I wouldn't touch that liability with a ten-foot insulated lawyer. IIRC, none of the MBI "safety circuits" operated as described; they removed the circuit from later Replicators.

AFAICT, you can't sell a safety device to be installed by an unskilled owner, because any subsequent failure / injury / catastrophe becomes your problem, no matter how obviously wrong the installation went, because your device didn't prevent a bad situation from getting worse.

(*) Which wasn't practical for mass production, but, yo, neither was the Thing-O-Matic:
https://softsolder.com/2011/03/14/thing ... t-circuit/

[rpollack]

After internal testing and working with a forensic lab, we have not been able to find a cause for the fire. We will continue to investigate.

Rick

[innkeeper]

what part caught on fire?

[sthone]

The printer owner wrote "The entire 3d printed drive cover and fans melted and were in flames." so I assume it had something to do up top with the extruder.

Rick also noted "We do know that the printer was not working properly prior to the fire as it had shut down multiple times with a temperature error. This means the temperature regulation system was not working correctly. The customer did not contact us for assistance but ordered replacement parts for the machine. While waiting for the replacement parts to arrive the customer continued to use the printer and let it run unattended. If there is a problem with your printer, unplug it and contact MakerGear for assistance and absolutely do not let it run unattended. We will continue to investigate and future updates will be posted on the MakerGear forum."

What ever caused the initial temp problem it sounds like the fire was the result of a lack of common sense in continuing to run the printer with a problem like that.