ESC caught fire, possible reasons?

[mek]

I have the quad-copter attached to a test harness, tried to get the quad just off the ground for testing. Was at aprox. 50% throttle when I noticed a motor slow down and the ESC LED connected to that motor turn red. After returning to zero throttle, I then tired 50% again and the ESC caught fire.

Quad-rotor Specs:

Tarot Ironman 650 Frame
Venom 5000 mAh 35C/60C Lipo (3S1C)
Afro 20 Amp ESC
Turnigy Multistar 2216-800 Motors (800 rpm/V)
APC Electric E 13'' x 4 Props
Turnigy 9x Transmitter Reciever
KK2.1.5 FC (Stevies newest)

​

A little more background:

No soldering was done. I bought connectors for ESC to Lipo.
The point of the Quad was to lift a payload (only for 5 minutes and does not need to get far off the gound). The test done DID not have an extra payload on it when it caught fire. During first test, realized the extra mounting screws on the motors were causing two motors to stutter (touching coils?). Replaced them with correct size/smaller screws. The motors that stuttered were NOT the one attached to the ESC that caught fire. Tried this same configuration on an old frame a while back and the quad would drift/flip on take off but no ESC caught fire. Scrapped that frame for this new Iron man one.

Not sure if the Props are too big for the motors? Or should I up the amp to 30 on the ESC? Or is this just a case of a bad ESC? All oher ESC were not hot after incident.

Any help is welcome!​

View attachment 2570 View attachment 2571

 

[mek]

Forgot attachments!

 

[GJH105775]

• * Looks almost like a short across two phases involving the black wire
  • For a short (seems the most likely) can you attach pictures of the connections from the ESC to the motors, and how long were the motor screws as well as the thickness of the mounting plate?
• over current (normally more than one FET gets hot for that)
  • Not as likely with your setup unless your motors were trying to spin while stopped
• FET lockup.
  • For the FET lockup did you by chance do a firmware upgrade?

EDIT, just noticed the screws part, if some were too long is it possible that the motor you used could have shorted acros the coils with a screw going just too far?

 

[Charles Linquist]

This has happened to me. Most ESCs (especially the cheap ones) are not capable of the stated rating (at least not for very long). And full throttle is easier on the ESCs than 60-85% throttle. I have a thrust stand and a thermocouple and have tested lots of units. Over the past year, I have had TWO ESCs catch fire. One in the air. Both were "no name" ESCs. One was rated at 30A, the other at 40A. The bottom line is that the ESCs were not capable of handling the load that was put on them, and they failed. I'm a EE and can guess the exact failure mode, but I won't go into that here.

 

[mek]

This has happened to me. Most ESCs (especially the cheap ones) are not capable of the stated rating (at least not for very long). And full throttle is easier on the ESCs than 60-85% throttle. I have a thrust stand and a thermocouple and have tested lots of units. Over the past year, I have had TWO ESCs catch fire. One in the air. Both were "no name" ESCs. One was rated at 30A, the other at 40A. The bottom line is that the ESCs were not capable of handling the load that was put on them, and they failed. I'm a EE and can guess the exact failure mode, but I won't go into that here.

I see (EE here as well). So basically you think getting a higher/better grade ESC would be the best route? The Afro's were cheap and I am assuming I paid the price there.

 

[mek]

• * Looks almost like a short across two phases involving the black wire
  • For a short (seems the most likely) can you attach pictures of the connections from the ESC to the motors, and how long were the motor screws as well as the thickness of the mounting plate?
• over current (normally more than one FET gets hot for that)
  • Not as likely with your setup unless your motors were trying to spin while stopped
• FET lockup.
  • For the FET lockup did you by chance do a firmware upgrade?

EDIT, just noticed the screws part, if some were too long is it possible that the motor you used could have shorted acros the coils with a screw going just too far?

Actually, I DID do a firmware upgrade....with the old frame I had an older version of Stevie's and with this new frame I upgraded to a newer version ( Stevies 1.9S1).

When I went to swap out the screws I noticed the screws looked like they almost/barley touched the coils. I attached pic with the long screw in. Second Pics shows the smaller screw I replaced it with before firing up the quad.

Also, If I do end up replacing the Motors/ESC do you recommend a better setup where I can still keep the 3S battery? I feel like te 13'' props were pushing it on the Multistar 2216-800 motors.

 

[GJH105775]

Actually, I DID do a firmware upgrade....with the old frame I had an older version of Stevie's and with this new frame I upgraded to a newer version ( Stevies 1.9S1).

When I went to swap out the screws I noticed the screws looked like they almost/barley touched the coils. I attached pic with the long screw in. Second Pics shows the smaller screw I replaced it with before firing up the quad.

Also, If I do end up replacing the Motors/ESC do you recommend a better setup where I can still keep the 3S battery? I feel like te 13'' props were pushing it on the Multistar 2216-800 motors.

Do you know how thick the motor mount was? Looks like that could be your problem, it only needs to rub the plastic coating off the winding on the stator, and barely touching while screwing in is enough to cause that. I burnt a whole set of motors and ESCs a while back.

Also when you say that you flashed it with Stevie's do you mean you flashed the ESCs or the KK board?

ESCs can be upgraded to te wrong version of the firmware and that would cause the MOSFET problem. 13'' may be a little big for 800kv/12amp motors, but probably not enough for a fire.

Do you happen to have a clamp meter?