Disco Pro taken down in flames by Jeti esc!

[henrysj]

Been trying to sort out why when yawing the disco I've been getting a huge wobble - it's alright on small rudder inputs but put the stick all the way across and it only just remains in flight before recovering itself into a nice flat pirouette.

About 3 minutes into a flight this afternoon one engine stops for a beat then restarts just before the quad hits the deck. 20 seconds later smoke starts billowing out of one side of the quad and then the engine cuts again before I can get it down.

So I'm thinking I've found the cause of why I was having trouble yawing!

Now was I asking too much from the 40A esc or was it just a faulty esc? I had noticed on a previous flight they were quite warm but not to the extent of combusting on me midflight!

Setup:
Disco Pro
6s 2700mah
Jet Hicopter 40A opto
Tmotor U5
14x5 Carbon props
View attachment 17909View attachment 17907View attachment 17908

 

[jes1111]

Sorry but I don't think it was the Jeti ESC that let you down - I think you abused it and it gave up as a result, i.e. any other brand of ESC would have succumbed too. From the images, I see that you had the ESC stuffed (quite brutally) into the fuselage space rather than mounted on the arms. The capacitors (and probably the FETs too) would subsequently have been deprived of cooling airflow and, probably most significantly, the power wires look like they've been forced against their will to conform to the chosen layout. Soldered multistrand wire is prone to breaking strands internally when you do that. The breaks diminish capacity, produce internal arcing and the result quickly becomes what you have there - a fire.

 

[henrysj]

Hi Jes1111,

The capacitors not getting airflow could possibly be a factor, although (from my admittedly limited knowledge of capacitors) they should not get warm enough to need a constant airflow over them to keep them cool, also this position has been used by other disco pro users the no issue what so ever. With the power wires would the internal arcing cause a breakdown/fire on the wires themselves or could it cause it internally infront of the capacitors? The Aerial Mob arms allow quite a lot of space under the boom for the wires so dont think the wires were too "abused" but I want to eliminate every possible idea before trying again - thanks for the info

 

[econfly]

Now was I asking too much from the 40A esc or was it just a faulty esc? I had noticed on a previous flight they were quite warm but not to the extent of combusting on me midflight!

The U5 with 14" props at 6s is pulling, at most, 13.7 amps according to T-Motor's spec chart. So that 40A ESC is certainly big enough -- maybe even too big, a 30A ESC would more than do the job. Without tearing it down and examining everything it's hard to say what caused the fault, but I would rule out the idea of excess amps UNLESS you have an evident fault in the motor or anywhere else downstream from those capacitors on the ESC. For these reasons, I also doubt heat alone was the issue.

 

[jes1111]

Hi Jes1111,

The capacitors not getting airflow could possibly be a factor, although (from my admittedly limited knowledge of capacitors) they should not get warm enough to need a constant airflow over them to keep them cool, also this position has been used by other disco pro users the no issue what so ever. With the power wires would the internal arcing cause a breakdown/fire on the wires themselves or could it cause it internally infront of the capacitors? The Aerial Mob arms allow quite a lot of space under the boom for the wires so dont think the wires were too "abused" but I want to eliminate every possible idea before trying again - thanks for the info

I'm no electronics engineer, but I do know that what's happening inside those ESCs is the electrical equivalent of "hell on earth" - the current is being "chopped" into a PWM square wave to feed the 3 phases of the motor. This chopping produces massive back-EMF spikes/ripples on the power input side. The capacitors' job is to absorb those spikes/ripples - which is what causes them to get hot and (eventually) break down. Broken strands inside the wire, poor soldering, mechanical cracks/stresses - basically anything that's not right in that area can cause what is already barely-controlled "electrical bedlam" to turn into "electrical armageddon".

In your first image, the positive wire is bent back over the capacitors at a very acute angle - solder will inevitably wick up inside the insulation with this multistrand we use and causes a brittle junction where the solder stops and the bare copper starts. This is (one reason) why the manufacturers apply heat shrinks that bind the +/- wires to the capacitor barrels - it supports and therefore protects that vulnerable point of the soldered wire end. Add vibration into the mix and you can see why anything less than "perfect" treatment and arrangement is a weak spot that will give up at some time.

 

[Bartman]

Hi Jes1111,

The capacitors not getting airflow could possibly be a factor, although (from my admittedly limited knowledge of capacitors) they should not get warm enough to need a constant airflow over them to keep them cool, also this position has been used by other disco pro users the no issue what so ever. With the power wires would the internal arcing cause a breakdown/fire on the wires themselves or could it cause it internally infront of the capacitors? The Aerial Mob arms allow quite a lot of space under the boom for the wires so dont think the wires were too "abused" but I want to eliminate every possible idea before trying again - thanks for the info

henry

did you have anything under the ESC like a layer of foam or silicone sheet? it's possible that the ESC got hot but a full blown meltdown like that was probably either a short or a component failure like a FET or something. putting the ESC's up against the tube mount like that wasn't a great idea and if the ESC was sitting on a hard surface then the vibrations probably got the better of it.

 

[henrysj]

hey Bartman,

There wasn't anything under them apart from the cable that connects the top and bottom plates together at the front end and another couple of wires for the FC power at the other - these made quite a soft bed for them though. Also it was the top esc that burnt out. The photo is a little misleading too, I had pulled out the esc a little in the first photo to get a good shot of the burnt out section - the positive cable passed in a relatively gentle bend under the tube mount back to the solder points. There weren't any forced bends in it but it was looped all the way around so as jes1111 mentioned this might have just been too much.