What is the problem

[Erfolg]

I have had two ESC's fail recently.

In the latest case, there could be an mode of operation issue, or there may not be.

It is a sports plane, drawing about 225w at max volt from a 3s-20c, 2250 battery. The ESC was a Pentium 30 amp.

Does the ESC work the hardest in that heat is generated at full or half throttle?

I spend most of the flight at about 1/2, with intermittent full.

Does this mode of operation tax the ESC the most?

I have a 40 amp on order, will this operate with less thermal stress.

At present cooling is probably poor, it seems obvious more ventilation would do no harm, yet how necessary is it really.

[Alwyn Gee]

Hi

You should always be working everything harder at full throttle as this is when you draw the maximum Amps & it is current draw that creates heat etc. Still with what you say about the power output you should only be drawing about 20A max. You say that you were using a 30A esc so provided your motor could also take more than 20A you should be within the limits of your power train. Ventilation can be a problem but this is usually more evident when running at the top end of your power trains capability and that doesn't seem to be the case here.

Not sure what you mean by "mode of operation issue" but does not sound to me like you are over-stressing the system if your figures are correct. Could it be an issue with cheep pre-installed ESCs in RTFs

Alwyn

[PETER BRUCE - Eastchurch Gap]

Hi. Just one point and that's related to the LOAD you put the motor under - if your using a prop that's creating a bigger load (too big or too much pitch) then the motor will draw more current and that can cause problems - only a long shot in your case but worth a mention if it helps you. Regards Peter

[Garbo]

Have you a watt/power meter, this has to be the best tenner I spent.

Old wisdom was that running at half throttle put the esc under more strain as it was "switching" constantly and full throttle lets the power flow straight through. I'm not sure if this statement still holds true.

[Allan Bennett]

Are you sure it's not the BECs in the ESCs that failed? Did you still have control of the model after the failure? The spec. for the BEC says, "4 micro servos on 3 cell lipo, and 3 micro servos on 4 cell lipo", so how many servos have you got, and how hard do you work them in flight?

Simple logic tells me that the ESC works hardest at full throttle, for then it's handling a continuous 20A (say) instead of the intermittent (several thousand times per second) 20A that it's handling at part throttle. But I have read that part-throttle is more taxing -- don't understand why though.

Either way, a 30A ESC should have no problem at 20A, and a 40A ESC should have even less problem. If you really are drawing only 20A or so. The 40A Pentium also comes with a switch-mode BEC, which should be able to handle more and bigger servos.

[Frank Skilbeck]

I'm not an expert but I have put a temperature logger on an ESC in the past and the temperature was higher at part throttle and then decreased at full throttle (but this could have been due to increased air flow).

I believe they get hotter at part throttle setting because of the switching losses.

I've had problems with overheating ESCs in the past and the most important thing is to make sure that the hot air can get out, Your batteries and motor will also be generating some heat so everything will be getting nicely cooked if there is no way of removing the heat, especially in a foam model which is an excellent heat insulator. All my good foamies have a much larger air outlet than inlet and none of them have ever sufferred from overheating ESCs.

[Bob Cotsford]

Don't the FET's that do the switching have the highest voltage drop when they are actively switching current? I can't remember enough electronics theory now but I would have thought that they would absorb very little power at full throttle and at closed throttle with a significant increase at part throttle.

[Erfolg]

The mode of operation is reference to the approx 1/2 the flight at half throttle, some 30% at full, the rest ticking over as coming into land, stall turns, chandelles etc.

The motor is a Keda Thumrun 2834/18 rated at max of 260w

I have a watt meter max draw in the past was 225w on full charged Lipo, before volts sagged.

I have not tried it again, as I have only just got everyyhing running again.

With the failed ESC (in both cases) I retained full control of the surfaces, so not the BEC part. It seems the timing part has failed, as the set up timing noise is absent, of course the thyristors could also have failed.

I have 4 servos, although the rudder is seldom used, mostly elevator. aileron, although no sign of the BEC part failing.

The Lipo can at the hottest be said to be just above ambient after a flight.

Edited By Erfolg on 04/09/2012 15:07:32

[Garbo]

Its one of these I use for setting up my electric give's Watts and Amps so I can balance between the best prop for power and stay within the limits of the ESC.

[PatMc]

Posted by Garbo on 04/09/2012 18:38:25:

Its one of these I use for setting up my electric give's Watts and Amps so I can balance between the best prop for power and stay within the limits of the ESC.

An ammeter would do that.

Erfolg has already said he has a wattmeter & I'm sure if he isn't able to measure the current directly he knows how to calculate that he's within the current limit of his ESC.

[C Norton]

You are quite right Bob, the FETs in the ESC are dissipating the most heat at part throttle settings although many of them have thermal protection circuits and SHOULD shut down if over temperature. I wouldn't bank on it

Edited By C Norton on 04/09/2012 19:19:06

[Erfolg]

Could the burbing sound be the moor cutting in and out as a consequence of a problem with temperature protection device. Finally failing with the consequence of not running at all.

The only thing that perhaps contradicts this scenario is that the motor no longer plays a tune when setting up the timing. It could be the algorithm just overrides the start up sequence if the boolean logic prevents start up if over temp is effectively open.

I can work out the A = W/V, in our case 225/(4.2 * 3) = 17 amp

What I am really asking why not have the max head room possible, with respect to ESC amp. Most things are not all gains, particularly that the difference in cost of a 40 amp esc is not very different to say a 20 amp.

[Simon Chaddock]

I now put small finned heatsinks on my ESCs protruding into the external airflow on my 20A+ foam planes and despite this the heatsink still gets warm to the touch no matter how much (or little) throttle is used.

However after say a 10 second glide to a landing the heatsink is practically cold which suggests that quite a bit of heat is being dissipated when they under load.

Unless this heat can actually be removed even a 40A ESC running at 20A will overheat although with a bigger thermal mass it may take a bit longer.

[Erfolg]

As with many things, this discussion has rekindled memories.

The Burbing sound I mentioned at the beginning, is reminiscent of a similar noise I heard from a model previously, although essentially forgotten.

Having read that there is a thermal cut out, I do wonder that is what I have heard, cutting in and out.

Anyway, I have decided for sports types models cooling is necessary to the ESC, which are controlling the motor current continously. Unlike gliders which typically run for less than 30s, thereafter being shut down for 5 minutes or there about, or even longer.

I have rearranged my Lipo and ESC in my Nobler

The Lipo has gone to the bottom of the model, the ESC is now sitting on top. I have cut the PVC around the heat sink on the ESC. I have made an inlet in the top hatch,angled to direct air onto the heat sink, a little further down the top hatch I have made an outlet, angled to assist out flow. The outlet is *3 the inlet.

Without instrumentation and measurements, I will never really know if I have solved an imaginary issue, or heat is the issue.

Edited By Erfolg on 05/09/2012 11:01:12

[Simon Chaddock]

Erfolg

That is one advantage of having fins sticking out into the airflow - you can feel directly how warm it is!

The fins on the 30A ESC buried inside the Depron fuselage. There is no internal airflow.

.

It did not over heat (cut out) during 71 minutes of continuous running.in flight. I checked immediately on landing and the fins were just warm to the touch.

[Cliff Bastow]

i am by no means an expert on electric flight but my understanding s that the esc is working harder at low throttle than full. this is because the esc only supplys full power, but at low throttle settings it is constantly switching on and off wheras at full throttle it is just switched on. this is what generates heat. hope this helps

[PatMc]

Cliff, I thought the same. In fact for that reason I was planning to use a good quaility geared brushed motor in a Jnr 60 instead of a brushless one as I expect it will be flying at low throttle most of the time. However I've since been told that FET losses are the same brushed or brushless ESCs, except for some early brushless ESCs that had a low PWM rate. They were 50Hz, same as Tx frame rates but modern ESC are typicaly 8kHz or more.

[Keith Simmons]

A light bulb springs into my mind, The bulb lasts longer when it remains switched on, however if you keep swiching it on and off it will soon fail, Perhaps it is due to the power surge going through as maybe the surge itself produces more heat. I am no expert on electric theory so cannot explain why.

[Ben B]

The only thing I'd add is that Pentium / Plush speed controllers are pretty robust so if you're killing them you're being very nasty to them!

Unless you've done something really bad like put a different prop on from when you had it on a watt-meter it can really only just be over-heating.

Brushless ESCs are pretty efficient so they don't make lots of heat but they do make some and you'll cook them unless you have enough airflow.

PatMc- as you've said brushless FETs lose just as much as brushed. The only difference going brushed would make is you'ld have even more loses from the motor

[GrahamC]

A bit of latteral thinking - which could be way wide of the mark.... Dodgy motor? Bearings failing or bits of swarf on the magnets could cause it to stiffen up and overlead the ESC? - Does it spin freely? I've had a motor go notchy on me causing the ESC to get very hot. Worth checking I would thin.

[Tony Smith 7]

Surely an ESC on a brushless motor is switching all the time, whenever the motor's running, and always at the same frequency relative to the motor rotational speed. It's never "straight though" as it needs to switch between the phases even at full power. That's my understanding anyway,

Tony S

[Kelly]

Full power is constant, the esc acts like a blip switch on the old ww1 rotary engines, but a lot quicker so the esc generates more heat at lower throttle settings cos its working harder (blipping more)

[Simon Chaddock]

Tony & Kelvin

You are both right as a brushless ESC actually switches two things.

First the motor 'phases'. The bit that is equivalent to the commutator on a brushed motor so it happens all the time the motor is running. The switching rate is proportional to the motor speed.

Second the power level. The length of the power pulses vary depending on the throttle setting but the frequency of the pulses is fixed by the ESC's circuit design. Thus the 'throttle' switching losses are more or less constant. Apparently even at full power it is still pulsed but the 'off' bit is so short the motor doesn't notice.

Then there are the resistance type losses. These are of course directly proportional current flowing.

More or less heat at part throttle?

About the only thing you can say for certain is there will be some!

[Erfolg]

At the moment I have some personal problems which will probably stop me doing anything for the moment.

I should find the time to check the watts with the new ESC and revs.

As for the rotation of the motor. The Runthum spins freely, there is not even the cogging I have with some motors I have.

Anyway on the new temp, perhaps permanent ESC it seems to run well.

It could be just coincidence that two different models have had ESC failures. Particularly as one is a small glider, where the run time is in secs, no time to overheat.

[Tony Smith 7]

Posted by Kelvin on 07/09/2012 10:40:12:

Full power is constant, the esc acts like a blip switch on the old ww1 rotary engines, but a lot quicker so the esc generates more heat at lower throttle settings cos its working harder (blipping more)

Surely even at full power the current is not continous, but has to be switched between the three phases in synch with the motor rotation.

I must say I haven't looked at the design in detail, but I assumed that the power control (length of pulse) and the commutator type function would be carried out by the same switching components. Ie you'd have three sets of switches one for each phase, and they'd be switched on and off to both rotate around the phases and to deliver the required current.