Kosmik 200/770 loss

[abanihani]

We still don't know if Kosmik is to blame or not, we are running higher power motors 4540, stc on Kosmik and it asks for more. The batteries could have shorted out and caused the fire and not vise versa. I noticed most of the peolple just jump into conclusions as if they want just to blame Kontronik for that. If i can see these batteries upclose I think they are pulse batteries which are not of good quality anyway. I am also so sorry about this incident but we all have to be patient and wait for Kontronik to give us an update.

[felborn]

Looking at the later pics, no way that's because the batteries shorted, all the damage are from the esc. A crying shame, I honestly feel for you man. Hope you get it sorted soon!

[Wh1teAfr1can]

Originally Posted by felborn

Looking at the later pics, no way that's because the batteries shorted, all the damage are from the esc. A crying shame, I honestly feel for you man. Hope you get it sorted soon!

It is still under investigation nothing has been rulled out yet. Hopefully by next week we will have an answer.

[Stolla]

I bet running the fuse would have prevented this........maybe i should install mine

[smsodhi]

I don't understand the fuse, It's a 200amp fuse. We are peaking in the 260-320+ amp range. Will it blow then?? or is there more to it than I understand? Would not want it to blow prematurely, I think this is possibly why a lot of people have not installed it.

[The Dentman]

Yes the batteries and ec5 and balance leads are fine.. damage on frames is focused on lower part near esc. None near batteries and all cells still okay. Just little burned from framed.I'm only posting pics to help in possibly discovering the failure. The attention Charles Anderson is showing me and compassion is very appreciated. And also the reps who've contacted me. My first priority is to help them understand the failure. Esc already on way to guys here in US.. if they continue to handle this in the way there reps are now ill be more confident then before to install there equipment in a new 770 build. I've made no demands for new kontronik equipment. I'm staying positive and confident till i see otherwise that Germany will follow up with assistance of kontronik equipment for the 770.

[KSHeli]

Originally Posted by smsodhi

I don't understand the fuse, It's a 200amp fuse. We are peaking in the 260-320+ amp range. Will it blow then?? or is there more to it than I understand? Would not want it to blow prematurely, I think this is possibly why a lot of people have not installed it.

You would possibly have to run a larger fuse, it just depends on how long the fuse can sustain current spikes above 200A. At 260-320 a 200A fuse would blow quickly.

[abanihani]

Unfortunately fuse or not fuse it does not explain it. I saw a flight by a Banshee team member using pyro 800-850 at over 2200 rpm, i am sure he logged over 400 amp on this thing and nothing blown

[Stolla]

The fuse doesnt blow at 200, read here https://www.helifreak.com/showpost.p...6&postcount=13

[abanihani]

Here we go, thanks Stolla for clarifying this matter, I am now convinced to install the fuse, I rather have a burnt fuse and auto than a burned heli

[waxman]

I think I will install the fuse on mine.

[ccpecot]

Wow....sorry to see that happen to that beautiful machine....

[4cyclic]

Sorry to see this.... cringe at it is more like it.
But hard to believe it's esc related. The unit has really high amp capacity.
First suspects would be pack itself or faulty connectors or shorted motor.

[Vinger]

To me it looks like the wire with the piece of circuit board still attached to it, somehow broke in flight and shorted on the case. That would explain the case being burned like that.

Still not good to see this on THE top of the line ESC, but I am sure Kontronik will sort it out.

[Stolla]

For what its worth i had a perfect low flight time powerjazz blowup with similar damage, the blown leads and all. In my case it was due to a "faulty"scorpion 5020 motor, it would start on the jive then cut randomly so we did 1000v test and decided it must be jive idiosyncracy, out came the pjazz, up in flames went the x7, also on maiden btw. We knew it was motor related as both jive and pjazz didnt work but we never found the issue as we sent it back being a new motor. So sometimes an outwardly normal motor can be the cause of some serious destruction. However the jive would cut but pjazz burnt so the questionable issue here is why the kosmik protective devices did not work

[dl7uae]

Strange the overheating cut off protective function did not function to prevent it going up in flames.

so the questionable issue here is why the kosmik protective devices did not work

protection == sensor + intelligence + action
action: That's the problem.. Just try to keep a car with a broken axle on the road.

Rather easy (but fortunately rare) to cause a worst case scenario like here:
Bad commutation failure --> very high voltage --> FETs on both ends of a bridge go broke by extrem overvoltage, unfortunately both continuous conductive now --> short circuit --> burning --> other FETs join in solidarity --> bonfire

No chance for any protective function to act against.

Unfortunately fuse or not fuse it does not explain it.

Does not explain, that's right, but would have been the only working protection against fire.

And before anyone asks: No chance to kill highly energetic overvoltage spikes like that by protective means. Big motor, much stored energy, high supply voltage to add on.

It's time to move to sensor-equipped brushless motors.
Sensorless commutation is potentially dangerous and the potential rises and it rises above the critical level now, thanks to better and higher voltage batteries.

Tom

[rossk]

Originally Posted by dl7uae

It's time to move to sensor-equipped brushless motors.

Whats the downside of these motors and why are we not using them already?

[dl7uae]

No downside. Only that the market is flooded with sensorless motors. It even wouldn't be that difficult to add sensors to sensorless motors (Hall elements).

The problem with our sensorless motors especially is the madness of all of these overpowered motors, the iron goes into saturation, and we get no clean zero crossing anymore (EMF pulse), - among other things. On top and in bad combination: High and stray manufacturing tolerances, non-equidistant glued magnets, non-central bore in the stator, worn out bearings etc.

Strong and cheap is really just an absurdity. It's only a license to kill.

And thus: a strong battery always wins in the end.
Impossibility for a FET to always survive.

AND: The same for a BEC, wether internally or externally. An overvoltage spike by commutation failure will kill him also. Thanks to the high drive voltage no electronic eye will stay dry, receiver, servos, FBL etc.
At least a protective device like a ZVS diode should be provided to rise chances for electronics to survive, applied on both ends, primary (battery side) and secondary (BEC output). But such a thing would be no absolute security. -- A BEC galvanically decoupling primary (battery side) from secondary circuit is the only way to get security against the worst case scenario of battery voltage appearing on output. See the HV²BEC from my mate Linus.

Well.., things are not as simple and risk free as we like to put down. The "Motor Madness" clearly shows our wrong philosophy of "How much is it?".

Tom

[Stolla]

Tom absolutely agree with you! please when you get time show us an example of a setup you think would allow some safety without risk of premature failure. A diagram showing components needed would help electronic noobs like myself to get a better understanding of how to incorporate these safety margins in existing setups if possible. Maybe you should add the "safe log"module to "jlog" franchise

[ElectricSheep]

Originally Posted by Stolla

Tom absolutely agree with you! please when you get time show us an example of a setup you think would allow some safety without risk of premature failure. A diagram showing components needed would help electronic noobs like myself to get a better understanding of how to incorporate these safety margins in existing setups if possible. Maybe you should add the "safe log"module to "jlog" franchise

+1!

I'd also like to see a setup example that Stolla is requesting...

Thanks!