Flashed XP12A fried its FET

[DoubleCH]

Just last wednesday my friend finally got all his brushless order from Astroid Design (hp06v2, 8T, XP12A non-flashed, frame brace, RRC prop, etc.). I flashed the XP12A and built him the brushless setup. Everything went well until just 3 days flying his new brushless setup my friend had a crash and the ESC wouldn't arm any more. Motor and rotor still spin freely. Only physical damage was a split stock hollow mainshaft. I managed to take a look today and motor/rotor still spin freely. Without power, I tried shorting out any 2 motor wires temporarily to work as electric brakes and I could feel resistance alright turning the main rotor so I don't think there's anything wrong with the motor. I noticed the 3 beeps on battery plug-in seemed normal. But right after the low/high beeps from arming, blue light on ESC dimmed momentarily and 3-in-1 rebooted and gave out the 3 beeps again, exactly as my friend described earlier. That's an indication of a short that brought down supply voltage for the 3-in-1 board, which was confirmed by warm FETs. I pulled the battery to check voltage and rechecked everything again. Nothing seemed unusual so I tried plugged in the battery again. This time I didn't arm the ESC. Just let it finish the 3 beeps. I left the battery plugged in and wait a bit to see if the FETs get warm. I took the time to get out my toolstick planning to read the ESC next. All of a sudden white smoke came out and I quickly unplugged the battery. By this time, at least 1 FET has been fried. Upon closer inspection, I found 1 of the 3 motor wires (the one that goes to the fried FET) was desoldered from the board and there must've been so much intense heat that the solder on the board metled (i.e. no sign of motor wire ripped out of the board).

I've heard of a few cases of fried Oversky 10A ESCs but I don't think I've ever read any case of fried XP12A ESC. My friend has been running his hp06v2 with 8T and governed at anything from 70% to 80%, He's still experimenting different but very reasonable settings. The setup is only 3 days old! This is bizzare! Did my friend just happen to get a bad ESC to begin with?

EDIT: The ESC and wires are pulled out specifically for these pics. They were NOT from the result of the light crash.

[wadders248]

I had exactly the same thing happen last Thursday! I'd just finished adding the bits from Dylan for a brushless tail, ran a few test hovers and on plugging in a fresh battery, the XP-12 went up in smoke.

The middle motor wire had come away from the soldering pad. I'm pretty meticulous with my soldering so I was quite surprised to see the pad had a fresh shiny layer of solder that looked like it hadn't been touched. I can only think I didn't use enough heat when soldering the wires, though they'd survived the hard tugging test!

Oh well, new XP-12 please... I did think of having a go at the one cell specific fets, though I'll need to find them first!

[Vizzion]

Had a similar issue with mine after a crash, different symptoms but my ESC would arm but wouldnt turn over the motor - replaced the ESC and she was fine...

I could still read the firmware using BLHeli and the like, but no matter what I couldnt get the ESC working again.

[DoubleCH]

Is the XP12A really this fragile? I've run a few different ESCs and have crashed with each of them and I've never killed a ESC just from a crash [knock on wood]. My other friend crashed his XP12A/hp08/16.5k so bad the main gear stripped at least 8 teeth and it still didn't kill the XP12A. I mean, this friend with the dead XP12A has only flown his new brushless setup with the ESC for 3 days. And it was just a light crash. And no, he's not a novice he flies pretty good with 450s as well. Makes me rethink going with XP12A. Maybe I'll just triple-FET my double-FET DP3A.

[coredump]

I don't think the XP-12 is really fragile. However, the placement of the 3 motor
wire pad is not idle. I study your friend's setup, that 3 motor wire, on impact
will pry against the FET. If it lift the pad off the board a little bit, it might create
damages.

That is why I chose to left a section of the original motor wire along, let them
lead to out side of the board. Then I make my connection there. Haven't have
any problem with it so far.

It is a picture of the front of Bruce (aka the shark). On impact, there will be
no prying against the board. Also I use silicon servo tape, it works really well.

[DoubleCH]

Originally Posted by coredump

I don't think the XP-12 is really fragile. However, the placement of the 3 motor
wire pad is not idle. I study your friend's setup, that 3 motor wire, on impact
will pry against the FET. If it lift the pad off the board a little bit, it might create
damages.

I'm absolutely sure the placement of the wires on my friend's bird is not a problem at all. The wires route below the ESC and 3-in-1 board next to the battery to the back of the 3-in-1 and the motor and is not taut and will not get caught by anything in any kind of crash. There's absolute no way a crash will yank the wires unless the 3-in-1 board is totally separated from the frame and fly more than 4 inches away with the FET side facing the frame and with a great force. There's just no way. The wires and board are pulled the way shown just for photos. The setup was nicely done [cough... cough...] and was definitely not in a way to jeopardize the ESC. BTW, I recommend you install the ESC FET-side out. You don't want to trap the heat close to the 3-in-1 board and with gyro tape. Plus I don't recommend routing motor wires on the side of the 3-in-1 board. Your wire routing is much more prone to damage and yanking.

[coredump]

Thanks for the recommendations. The motor wire is not going from the side.
It is going from the top when the canopy is on. I use jig to flash the ESC so
I want to have front panel access of the flashing pads. The ESC is not hot,
I have close monitored it.

About your friend's ESC. I don't buy it will just fry by itself. There must be a
reason for it, especially other people also confirm fry after a crash. It seems
there is highly correlation between crash and frying the ESC. I would think
that there are some where shorted during the crash. More detective work required.

[Zaneman007]

My xp7a just caught fire? Two of the FETs turned red and caught fire?

[DoubleCH]

I do believe the light crash did something to the ESC to trigger the fry, but I do suspect it's something internal to the ESC, and most likely gave the last push to a cold solder joint to the P-FET driver circuit.

I did detective works before it smoked. The ESC became progressively worse before it smoked. At first when my friend handed me the bird on-site, it managed to arm but on spoolup the motor stuttered and stopped. I immediately felt temperature on the ESC and motor wires with my finger and interesting enough only the top rows of FETs (P-FETs) were warm. The motor wires were not warm so I ruled out the motor. Tried a few more times and same result. I then unplugged and temporarily short out any 2 of the 3 motor wires and cranked rotor by hand. I could feel equal resistance on shorting any 2 of the 3 motor wires. Together with what I found earlier with no temperature on the motor wires on unsuccessful spoolup I'm sure the motor is good. When I got home and did more testing, same result with the electric brake test and spoolup test. Temperature was felt only on the top row of FETs. I did a very close look at every tiny spot on the ESC and the motor wires but couldn't spot anything. After a few tries it couldn't even get to spoolup stage because on the low/high beep on arming, 3-in-1 dimmed blue light and rebooted. And eventually the situation got so worse the FETs smoked after plugged in, even without arming. Remember that these 2/3S-capable ESCs (XP7A and XP12A) have their P-FET signals driven not directly by the MCU because MCU cannot be driven higher than 1S. I suspect the transistors and/or its related components that drive the P-FETs was bad or had a cold solder joint and resulted in my observation of the P-FETs getting warm but not the N-FETs or the motor wires and eventually smoked them.

Note that the ESC is secured to the 3-in-1 board with servo tape on the MCU side not FET side and there's no sign of physical damage to the ESC. Besides, it was a light crash and the ESC was so protected by everything else around it that there's no way it has been impacted. I wonder if there was a cold solder joint in the P-FET driver circuit that finally gave way from the crash. Remember, the ESC didn't smoke right after crash or on 1st spoolup right after crash. It got progressively worse and eventually smoked.

[DoubleCH]

Originally Posted by coredump

The ESC is not hot,
I have close monitored it.

Enter program mode with full throttle on startup (stay full throttle). This will give out constant beeps and now feel the temperature on the FETs. You'll be surprised how hot the FETs get.

[coredump]

I take a look at your picture again. From the picture it seems indicate there
are 3 FETS are burned. Can you confirm which one of the FETs burn?
It is very odd that motor did not even get to the arm mode however the
FETs burn. The MCU shouldn't turn on the FETS at all. The very first thing
it does is turn off all FETs and wait for the throttle signal.

I agree it should be some thing happen on the MCU side. Do you have
close up picture of the MCU side?

[coredump]

Constant beeps is some thing will get the FET very hot. Because the
motor is not turning. It can have very high current when the FETS are
on, more so than the motor at full throttle running.

The FETS are not stick on the board. The motor wire lead and the big
capacitor touch the 3 in 1 board before the FET does. However you do
have a point that, if the FETS caught on fire, I don't want the FETS blast
against the 3 in 1. The ESC is much cheaper than 3 in 1.

Originally Posted by DoubleCH

Enter program mode with full throttle on startup (stay full throttle). This will give out constant beeps and now feel the temperature on the FETs. You'll be surprised how hot the FETs get.

[DoubleCH]

Only the bottom FET on the 1st pic was blown (0 ohms). The other 2 P-FETs don't have 0 ohms (but could also be dead). The 3 P-FETs "looked" to have blown but they all have some kind of dirt on each when the ESC arrived from Astroid Design (or maybe that's not dirt, hmmm...). I'll try to take better pics.

[coredump]

Dirt? The ESC should have shrink wrap.
Do you bought it flashed?

[DoubleCH]

It came with shrink-wrap, unflashed. Upon removal of the shrink-wrap, I found the "dirt" spots already there. The ESC flashed and flew alright so I didn't think that would be a problem. Here're some better pics front and back. In case you wonder, the communication wires were soldered on good. No extra exposed wire and no chance of short. With hot-glue squeezed between them and the board, the wires won't pull out the pads and they were nicely positioned when the ESC is stuck to the foam tape.

EDIT: Before I flashed the ESC, I ran it with my OrangeRx and 2g tail motor. It performed exactly as expected before I proceeded with flashing it.

[RedJoE]

Well, fried my Xp12a last night. I had just got done flying 12 packs with it. I got home and was going to run a bench test, but it wouldn't start the main motor and I couldn't hear the arm tones, odd. I started inspecting and found the middle bottom fet had moved off it's pad and looks like the motor wire's solder had been reflowed on it's own.

Now that I think back, the last couple packs, there was some kind of power issue, it still flew fine, but was kind of sluggish and the pitch was spongy feeling, I even landed a couple times and checked the main gear to make sure it didn't slip, so I think that was an indication that it was going south.

Two more things I noticed recently as well.
1) The last 15-20 flights or so the tail would blow out slightly, where before that it was rock solid. I have a bent tail motor bell, so I thought this might be it.
2) On huge direct descents of 100 feet or so with negative pitch, towards the bottom the chopper would rotate clockwise on it's own, sometimes a full or more than one revolutions. I thought maybe the tail motor was shutting down or something.

Also, upon inspecting and as you can see in the picture, there is a capacitor missing on the main 3n1 board, anyone know what this guy does. I wonder how long it's been missing, perhaps this has something to do with the tail. I pulled a cap that looked the same off an old broken msr board and put it on the mcpx, but it made no difference, the motor still won't work.

I attempted repairing the Xp12a by reattaching the fet, however, I still don't get the motor tones, it does arm silently, but the motor will only turn over a couple revs and then dies.

Already ordered another one.

[Skyroo]

Redjoe, I have a flashed v3.4 xp-12a that exhibits the same problems you described but with all the fets still intact. My motor will still spin with the main gear dismounted. With the main gear mounted, the motor will only twitch a little and stop.

I have another unflashed xp-12a that has a mind of its own like making the motor spin at a low speed despite TH and despite turning the TX off. I have to do frequent rebinding to make it work to a certain extent.

Meanwhile, all my xp-7as are rock solid.

[RedJoE]

Yeah, no problems with my much older are more used Xp7a. I wonder why all these Xp12a controllers are burning out.

Does anyone happen to know the specs on that missing ceramic capacitor on my board, or just what it does for the mcpx board. I checked Mega's site but it isn't listed there, looks like it's an uncommon item to break off on these, searched the forums and also couldn't find anyone else with issues on this component.

[branmuffin]

Headed to the park to get some stick time in today, first pack went fine no issues. Decided to adjust my linkage and check zero pitch and smelled that smell we all know to well. Pulled the canopy off and the fet on my xp-12a was melted and the wires obviously reflowed themselves. Not sure why it burned up, possibly just end of its life. Been using this esc for 4 months or so with probably about 800 flights.

[Shake and Bake]

Burned an fet on my xp-12 too and it took off a motor wire as well! Sucks... Wonder if these are really only meant for 1s sPec fets, or just a seriously bad batch going round! I've never crashed mine and it did this..