Second ESC burned within a minute

bb42 · Oct 20, 2009, 11:15 PM

My SJM400 just burned it's second brushless ESC controller - the first (Otter BS35A) lasted a couple of short (1 min.) hover flights after which the ESC cut down, seemingly due to overheat. Before it was neccessary to change to "hard timing mode", otherwise the motor would stop beyond half throttle. The current I measured was max. 13 A (w/o the blades attached).
The last flight then left the Otter ESC with dark smoke...

Then I got another ESC from Pulsar (a rebranded Hobbywing) with 25A. Here I took off the plastic coating to feel the warming up in test runs. It turned out that the ESC gets quickly hot after pushing the throttle over a certain level - similar to the Otter ESC before I changed to "hard timing". So I changed this timing to "high" - and the ESC sent smoke signals on the first test run.

Folks, what's wrong here? Is there some issue with the motor that it burns one ESC after the other? Are there not any (a bit more) intelligent ESCs that sense what the motor is doing instead of sending smoke? Or do I need a more powerful ESC (this is a 400 heli, definitely smaller than a 450)?

arbilab · Oct 21, 2009, 12:29 AM

Can't say definitively, not familiar with the exact products.

Generally, 25A should be adequate for a 400. My 450 is 30A and has no problems.

You read half the ESC capacity without blades? That might be a clue. Without blades there is no real load on the motor other than initial acceleration. Plus possible mechanical binding, gear overmesh, knurr around the shafts, etc.

Again generally, if you have a thermal problem, retard the timing rather than advance it. Could you have too few pinion teeth?

Bad motors can eat ESCs, brush or brushless. If the motor has been hotter than you can touch, it could very well have shorted turns. Any shorted turn eats 3 turns worth of motor efficiency and turns the rest into excess heat and current consumption.

I can't nail what's going on with your set, just offer things to check and consider.

bb42 · Oct 21, 2009, 12:07 PM

Quote:

Originally Posted by arbilab

Generally, 25A should be adequate for a 400. My 450 is 30A and has no problems.

Exactly my thinking.

Quote:

You read half the ESC capacity without blades? That might be a clue. Without blades there is no real load on the motor other than initial acceleration.

Ok, but I felt this could be okay as idle current, if the motor is designed like this. All gears and mechanics move easily. What would be (more) reasonable figures?

Quote:

Again generally, if you have a thermal problem, retard the timing rather than advance it. ... If the motor has been hotter than you can touch, it could very well have shorted turns.

The motor was not too hot, only the ESC went hot. And, before I set "hard timing" (on the Otter ESC) the motor would not turn above 1/3 throttle. After this it worked fine, the heli hovered well and had plenty of power left. Only the ESC got too hot.

Quote:

Bad motors can eat ESCs, brush or brushless. ... Any shorted turn eats 3 turns worth of motor efficiency and turns the rest into excess heat and current consumption.

As the performance was ok, I don't think there's a shortage (my Voltcraft VC820 multimeter did not show any resistance on each of the three coils, but I assume the coils resitance is just below its range).

As I understood the load on the ESC is not highest with the highest throttle but somewhere in the middle (because it has to "throttle more"?). Still an ESC should not smoke up on half throttle.
Also there seems to be some magic with different motors and their number of coils (?) etc. I searched on the net for a complete howto on these things but did not find any.

Simple question: If the ESC has about the correct size - on an idle run (w/o blades) it should easily handle all throttle settings and motor speeds, without getting hot - correct?
In this case I'd only conclude the engine coils were manufactured wrongly, something like this.

arbilab · Oct 21, 2009, 01:26 PM

Disclosure: I've supported (not designed) motor drives for 30 years. 3-phase electronically-commutated motors were in a great many things, BUT they required very little support, as they basically just worked. These were lowpower motors, such as VCR and diskdrive, AND the motors and drivers were perfectly matched to each other and to their loads. The same is not necessarily the case in heliworld.

13A is a lot of current to just be spinning an unloaded motor. Strongly suggests something is very unhappy. Rather confirmed, when the smoke came out at half the ESC's rating. I'm just not well-versed enough in heli powertrains to say what the problem is. If I still had access to lab equipment I could give you an appropriate 'idle' (unloaded) value for a 30A drivetrain. I would guess, after acceleration, about 1/4th of what you saw.

You probably know, the motor is a current source as well as a current sink. CEMF and all that. Running the motor is a wrassling match between the ESC source and the motor source. Reason I raise this, is that you noted the motor stopped beyond halfthrottle on one setting.

Makes me think you've gotten your hands on a voodoo motor. Pretty sure there is such thing in heliworld, just from reading. Some of these designers go way overboard on one principle, compromising the others to the extent the motor becomes unusable except under very specific conditions. Example, note that it flew OK but got a little too hot, but unloaded it smoked right away.

I'd ditch that motor for something a little more mainstream/conventional.

TMorita · Oct 21, 2009, 01:31 PM

Quote:

Originally Posted by arbilab

...
Again generally, if you have a thermal problem, retard the timing rather than advance it. Could you have too few pinion teeth?

I think you mean "too many pinion teeth".

The symptoms sound consistent with having too many pinion teeth. If your hovering throttle is, let's say, 20%, and your average hovering current is 8 amps, then what happens is the controller is forced to deliver 40 amp bursts at a 20% duty cycle. This will easilly kill a 25 amp ESC.

This is analogous to driving at 10 mph in 5th gear in a car...the gear ratio is wrong for the speed you're trying to drive, and it places a huge amount of stress on the engine.

Toshi

arbilab · Oct 21, 2009, 02:01 PM

You're right, I meant too many. Though you can both overload and underload a BL that's designed for a narrow load range. Note that his failure occurred unloaded--without blades.

Note also, the A rating is RMS. Any given switch cycle can exceed it. It appears the problem is the source current bucking the CEMF. Thus at half throttle, the RMS A rating was exceeded.

bb42 · Oct 22, 2009, 04:46 AM

Quote:

Originally Posted by TMorita

The symptoms sound consistent with having too many pinion teeth. If your hovering throttle is, let's say, 20%, and your average hovering current is 8 amps, then what happens is the controller is forced to deliver 40 amp bursts at a 20% duty cycle. This will easilly kill a 25 amp ESC.

Argument understood, maybe also my pitch curve is not correct, i.e. too much pitch for a given throttle (trying to make my first 3D heli behave like a simple one for a first try).
BUT the second (25 A) ESC did not even see the blades turn, only the head with the paddles. And with the first ESC the system drained 13 A from the battery also without blades. Then the first ESC was the one originally fitted in the package, as well as the motor, so they should be fitted to each other.

After that, I'd assume something's wrong with this very motor, wrong winding of the coils (or one of them) etc. Seems plausible to me that as arbilab explains "the motor is a current source..." it can destroy the ESC.
So I'll get another motor and ESC and report my findings at the time.

arbilab · Oct 22, 2009, 12:56 PM

There's another difference with heli BLs that throws me. Every BL I ever encountered was Hall timed. No heli BL is, they derive timing from CEMF and I have no lab exposure to how that works.

norcalheli · Oct 22, 2009, 03:57 PM

If you're trying to tame down the heli by lowering headspeed, it is much more desireable
(mostest desireable???) to do this by going to a smaller pinion, not by setting a lower throttle speed in your pitch curves. I've found, in very limited testing, that it is best to keep the throttle at or above 80% for all flying. Lower than that just seems to build up heat way too fast. For some helis and motors, closer to 90% is better. I found this out on my Swift with the 600 + 1110kv motor. On 4s, hovering at around 82% the motor came down very warm. Raised this to where hover is between 87% and 90%, the motor is noticeably cooler.

bb42 · Oct 23, 2009, 10:30 AM

Yes, I can confirm that lower throttle does not mean less losses.
And right, heli BL motors do not not have a hall sensor, this might be because with the proper pinion the speed does not change that much, only the load (pitch).

Only it does not seem to be the problem here, the second ESC burned just seconds after starting - so it seems that quite some amount of energy is fed back from the motor into the ESC which is not build for this.
I ordered another motor and ESC and will report my findings at the time.

arbilab · Oct 23, 2009, 12:41 PM

Surmising CEMF shorted the buck diode.

bb42 · Oct 28, 2009, 04:04 PM

Funny, with a new motor of exactly the same type everything works fine.
Idle current reaches just 3 A, braked it drains not more than 7 A.
What I did _not_ test was the old motor on the new ESC, for obvious reasons.

The only perceivable difference between the two motors is that the new one has clearly stronger 'latches' when turned, on each (of 9 during one turn) latch it clearly turns the same angle - this looks normal behaviour to me.
The other motor shows only minimal 'latch force', and there are now 16 positions on one turn, more difficult to sense.
So I assume it was assembled with the wrong type or quality of magnets.
Anyhow, it's strange to see that such a motor ate two ESCs but was still powerful enough to hover the heli.
And then ESCs should be able to detect such a faulty motor behaviour instead of going up in smoke.

richard55 · Oct 28, 2009, 05:06 PM

This is interesting to me because my Trex 500 just ate two ESCs. The second one was a brand new Align 60A. Plugged it in, wiggled the servos and POOF smoke pouring out of ESC. Before I could unplug it the RX also started puffing smokle. Argh. I put the motor on a scope and it shows a nice sine wave on each phase when i spin it, no noticeable damage.

I am a little hesitant to plug in another ESC. I'll try a cheep one on 12V I think. Just to see if it will start spinning.

Side note: I once flew the 500 on 12V for a minute. Even pulled a loop! Almost could not recover from the loop though. Landed safely.

arbilab · Oct 28, 2009, 05:14 PM

Whuddeye say in the second frame, about eating ESCs? And the 4th frame, about 3A? OK, 'nuff toldyaso.

Magnet may have cracked, introducing more poles than were supposed to be there. Still "runs" but nowhere like it was designed to.

richard55 · Oct 28, 2009, 05:31 PM

Quote:

Originally Posted by arbilab

Whuddeye say in the second frame, about eating ESCs? And the 4th frame, about 3A? OK, 'nuff toldyaso.

Magnet may have cracked, introducing more poles than were supposed to be there. Still "runs" but nowhere like it was designed to.

Ah, yes, this is very interesting.

Oct 20, 2009, 11:15 PM	#1
bb42 Registered User Join Date: Oct 2009 Location: Berlin Posts: 10	Second ESC burned within a minute My SJM400 just burned it's second brushless ESC controller - the first (Otter BS35A) lasted a couple of short (1 min.) hover flights after which the ESC cut down, seemingly due to overheat. Before it was neccessary to change to "hard timing mode", otherwise the motor would stop beyond half throttle. The current I measured was max. 13 A (w/o the blades attached). The last flight then left the Otter ESC with dark smoke... Then I got another ESC from Pulsar (a rebranded Hobbywing) with 25A. Here I took off the plastic coating to feel the warming up in test runs. It turned out that the ESC gets quickly hot after pushing the throttle over a certain level - similar to the Otter ESC before I changed to "hard timing". So I changed this timing to "high" - and the ESC sent smoke signals on the first test run. Folks, what's wrong here? Is there some issue with the motor that it burns one ESC after the other? Are there not any (a bit more) intelligent ESCs that sense what the motor is doing instead of sending smoke? Or do I need a more powerful ESC (this is a 400 heli, definitely smaller than a 450)?

Oct 21, 2009, 02:01 PM	#6
arbilab Unemployed Engineer Join Date: Jan 2009 Location: DFW, TX Posts: 3,329	You're right, I meant too many. Though you can both overload and underload a BL that's designed for a narrow load range. Note that his failure occurred unloaded--without blades. Note also, the A rating is RMS. Any given switch cycle can exceed it. It appears the problem is the source current bucking the CEMF. Thus at half throttle, the RMS A rating was exceeded. Last edited by arbilab; Oct 21, 2009 at 02:09 PM.

Oct 28, 2009, 04:04 PM	#12
bb42 Registered User Join Date: Oct 2009 Location: Berlin Posts: 10	Motor ate the ESC Funny, with a new motor of exactly the same type everything works fine. Idle current reaches just 3 A, braked it drains not more than 7 A. What I did _not_ test was the old motor on the new ESC, for obvious reasons. The only perceivable difference between the two motors is that the new one has clearly stronger 'latches' when turned, on each (of 9 during one turn) latch it clearly turns the same angle - this looks normal behaviour to me. The other motor shows only minimal 'latch force', and there are now 16 positions on one turn, more difficult to sense. So I assume it was assembled with the wrong type or quality of magnets. Anyhow, it's strange to see that such a motor ate two ESCs but was still powerful enough to hover the heli. And then ESCs should be able to detect such a faulty motor behaviour instead of going up in smoke.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
ESC burned up	Hap02	Power Systems	4	Jan 08, 2004 02:43 PM
Fusion esc burned up?	ProfitOfDoom	Electric Heli Talk	3	Oct 13, 2003 11:53 AM
6 amp Esc burned in stock GWS Beaver	Wheazel	Power Systems	2	Jul 13, 2003 12:07 AM
Using a second ESC to control lights etc. ?	Bill Glover	Electric Plane Talk	8	Sep 26, 2002 01:27 PM
GWS esc burned out: repair question...	burgan	Power Systems	1	Jul 15, 2002 11:16 PM

Oct 21, 2009, 12:29 AM	#2
arbilab Unemployed Engineer Join Date: Jan 2009 Location: DFW, TX Posts: 3,329	Can't say definitively, not familiar with the exact products. Generally, 25A should be adequate for a 400. My 450 is 30A and has no problems. You read half the ESC capacity without blades? That might be a clue. Without blades there is no real load on the motor other than initial acceleration. Plus possible mechanical binding, gear overmesh, knurr around the shafts, etc. Again generally, if you have a thermal problem, retard the timing rather than advance it. Could you have too few pinion teeth? Bad motors can eat ESCs, brush or brushless. If the motor has been hotter than you can touch, it could very well have shorted turns. Any shorted turn eats 3 turns worth of motor efficiency and turns the rest into excess heat and current consumption. I can't nail what's going on with your set, just offer things to check and consider.

Oct 21, 2009, 01:26 PM	#4
arbilab Unemployed Engineer Join Date: Jan 2009 Location: DFW, TX Posts: 3,329	Disclosure: I've supported (not designed) motor drives for 30 years. 3-phase electronically-commutated motors were in a great many things, BUT they required very little support, as they basically just worked. These were lowpower motors, such as VCR and diskdrive, AND the motors and drivers were perfectly matched to each other and to their loads. The same is not necessarily the case in heliworld. 13A is a lot of current to just be spinning an unloaded motor. Strongly suggests something is very unhappy. Rather confirmed, when the smoke came out at half the ESC's rating. I'm just not well-versed enough in heli powertrains to say what the problem is. If I still had access to lab equipment I could give you an appropriate 'idle' (unloaded) value for a 30A drivetrain. I would guess, after acceleration, about 1/4th of what you saw. You probably know, the motor is a current source as well as a current sink. CEMF and all that. Running the motor is a wrassling match between the ESC source and the motor source. Reason I raise this, is that you noted the motor stopped beyond halfthrottle on one setting. Makes me think you've gotten your hands on a voodoo motor. Pretty sure there is such thing in heliworld, just from reading. Some of these designers go way overboard on one principle, compromising the others to the extent the motor becomes unusable except under very specific conditions. Example, note that it flew OK but got a little too hot, but unloaded it smoked right away. I'd ditch that motor for something a little more mainstream/conventional.

Oct 22, 2009, 12:56 PM	#8
arbilab Unemployed Engineer Join Date: Jan 2009 Location: DFW, TX Posts: 3,329	There's another difference with heli BLs that throws me. Every BL I ever encountered was Hall timed. No heli BL is, they derive timing from CEMF and I have no lab exposure to how that works.

Oct 22, 2009, 03:57 PM	#9
norcalheli My other addiction! Join Date: Jan 2007 Location: Marysville, Ca., US Posts: 546	If you're trying to tame down the heli by lowering headspeed, it is much more desireable (mostest desireable???) to do this by going to a smaller pinion, not by setting a lower throttle speed in your pitch curves. I've found, in very limited testing, that it is best to keep the throttle at or above 80% for all flying. Lower than that just seems to build up heat way too fast. For some helis and motors, closer to 90% is better. I found this out on my Swift with the 600 + 1110kv motor. On 4s, hovering at around 82% the motor came down very warm. Raised this to where hover is between 87% and 90%, the motor is noticeably cooler.

Oct 23, 2009, 10:30 AM	#10
bb42 Registered User Join Date: Oct 2009 Location: Berlin Posts: 10	Yes, I can confirm that lower throttle does not mean less losses. And right, heli BL motors do not not have a hall sensor, this might be because with the proper pinion the speed does not change that much, only the load (pitch). Only it does not seem to be the problem here, the second ESC burned just seconds after starting - so it seems that quite some amount of energy is fed back from the motor into the ESC which is not build for this. I ordered another motor and ESC and will report my findings at the time.

Oct 23, 2009, 12:41 PM	#11
arbilab Unemployed Engineer Join Date: Jan 2009 Location: DFW, TX Posts: 3,329	Surmising CEMF shorted the buck diode.

Oct 28, 2009, 05:06 PM	#13
richard55 Registered User Join Date: Jan 2007 Posts: 61	This is interesting to me because my Trex 500 just ate two ESCs. The second one was a brand new Align 60A. Plugged it in, wiggled the servos and POOF smoke pouring out of ESC. Before I could unplug it the RX also started puffing smokle. Argh. I put the motor on a scope and it shows a nice sine wave on each phase when i spin it, no noticeable damage. I am a little hesitant to plug in another ESC. I'll try a cheep one on 12V I think. Just to see if it will start spinning. Side note: I once flew the 500 on 12V for a minute. Even pulled a loop! Almost could not recover from the loop though. Landed safely.

Oct 28, 2009, 05:14 PM	#14
arbilab Unemployed Engineer Join Date: Jan 2009 Location: DFW, TX Posts: 3,329	Whuddeye say in the second frame, about eating ESCs? And the 4th frame, about 3A? OK, 'nuff toldyaso. Magnet may have cracked, introducing more poles than were supposed to be there. Still "runs" but nowhere like it was designed to.