Pre- Launch Checking your Electric DriveTrain

[jims123]

I'm not sure if this is the right forum to ask, but would like to know if there are any and proven techniques for pre-maiden flight tests to check your ESC and Motor without over stressing it to thermal overload and point-of-failure, but to be sure the Battery/ ESC , and motor with a given prop as wired in a particular plane.. are not going to over heat in flight..

Can i simply run up to WOT for 5 or 10 seconds or so, and then shut down and feel the components that may be HOT and use that to predict possible failure during maiden flight? .. or are there better ways to pre-check while on the ground?

I am most concerned about fire risk.. and DC/ or AC power connectors and the motor's Bullet connectors and soldering connections that carry a lot of current in general.. Is there a way to be sure my plane's components will all run well within thermal limits in flight.. and that the plane won't crash and cause a potential fire risk as a result of say a bullet or Dean's connector or solder joint that expands under stress or heats up as high amps are pulsing thru it. :o :o

[Acetronics]

Hi,

May be something like that for first check ??? just to verify everything is ... well designed.

http://www.drivecalc.de/

Alain

[Ron W3FJW]

A check with a watt meter when setting up each plane, should provide the information you're looking for. That along with a finger test for hot equipment/connections should ease your mind.

[jackerbes]

I like to tie the fuselage down and run tests in the shop. I usually do it with the Watt's Up connected and keep some notes on the run times and I take some RPM readings with a handheld tach too.

I like to do a test to realistic throttle settings and durations to get a feel for motor vibration or issues and the learn the battery run times. I usually do that by starting with 5 to 15 seconds or so at full power or whatever power level I expect I'll use at launch. Then I throttle back to the level that will sustain flight and occasionally throttle back to idle, and up to launch power to simulate landings or touch and goes.

I cycle the sticks too to see the loading changes on the Watt's Up and move the control sticks and hold them at full travel to see if I'm can spot a stalled servo or anything.

I'll do a few momentary full throttle runs too just to see the load and make sure the ESC is going to handle it after it is well warmed up. I figure the ESC is well enough cooled in the shop and the ventilation in the plane with the wing on is not going to much if any better.

I like to do at least one test run that takes the battery down to the LVC. That gives me a test of the LVC and gets a real time value for battery duration that I can usually count on. And it also give me a feel for any changes in RPM or performance that might signal an impending LVC.

I recently bought a used Phoenix 25 ESC and tested it when I got it and it was working fine. I used Castle Link to update the firmware and set the settings to my preferences. I intended to use it with a 4S A123 pack so I set the LVC to a custom setting of 2.5V per cell or 10V.

When I got it installed in the plane, I wanted to test the LVC so I got it running was just playing with it occasionally. The ESC was in the fuselage, unwrapped and laying loose near the receiver. The motor had pulled up to 18 Amps for a few seconds and I had decided that I would probably limit the throttle travel to make that the full throttle setting for this motor/prop combo. Most of the run was in the 10 to 13 Amp range.

About two minutes into the test the motor was at half throttle and pulling about 12-13 Amps when the ESC self destructed. There was a flash of flame from the motor leads end of the ESC and a puff of smoke. I pulled the battery Dean's connectors apart at the load end of the Watt's Up. The motor had already stopped, and the fire was out.

I felt the battery and it was not even warm, battery and motor leads were cool to the touch all the way to the ESC. The motor leads had all three become unsoldered and fallen away from the ESC and the 3.5 mm bullet connectors were cool to the touch and showed no evidence of heating.

Looking at the ESC it looks like an IC burned and vented and may have taken out an adjacent IC too.

There was no damage to the plane other than a slight scorching of the foam and some smoke residue. The receiver and servos are all working normal tested with a UBEC, the motor has not been tested (I'm out of ESC's) but looks fine.

Interestingly, the Watt's Up showed peak Amps reading of 62.4 Amps. That was apparently a momentary peak at the time of the failure. My 1100 mAH A123 pack is fine, only slightly depleted from the three minutes of use so the brief 57C loading does not appear to have hurt it any.

I had not touched the ESC to check it for heat before it failed. I'm going to start doing that more often.

I'll mail that off to Castle for an out-of-warranty repair ($25) and hope I have better luck with the repaired or replaced one I get back. This ESC was 2-3 years old and had only been used about 5 times when I got it. It was clean and showed no evidence of previous overheating or abuse.

So to me, the shop testing to realistic durations and limits paid off this time. This would have been much more unpleasant had it happened in flight.

Jack

[Acetronics]

Hi, Jack

You used it at its maximum "specified" voltage ... ( 4 LiPos )

What was the failure caused by, then ???

misuse or ESC component failure ?


Might be the failure could be avoided by a little "oversizing" ... don't you think ??? :rolleyes:

a little calculation shows internal resistance * current * current is about 1 Watt per output Mosfet...

This is very,very,very,very .... close ;) to the allowable use of those output transistors.

So ... failure is quite a logical happening, here ...

I've also read 25A ... but calculation show half of that is ... to seriously consider.

Ahhhhhhhhhhhhhhhhhhhhh, Marketing exagerations .... :censored:

Alain

[jackerbes]

The A123 LiPoFe4 cells (aka 18650M1 1100 mAh cells) charge to 3.8V and then fall to 3.3-3.4V within a few hours. So using 3.4V as a "fresh charge" value in the field is about right. The starting voltage on that 4S pack would have been about 13.2V which is slightly higher than a 3S LiPo starting voltage (12.6V) and quite a bit less than the maximum potential of a regular 4S LiPo (16.8V).

Under load the A123 packs quickly fall below 12.6V (the 3S w/BEC spec) to 12.0 V or about 3.0V per cell. Then that voltage sustains pretty well for the duration of the run time. Then the A123s fall off sharply and quick at the end.

So I think I was well within the advertised operating parameters for the Phoenix 25. I'd consider the near instantaneous failure and the apparently very brief spike in amperage, with it running at a 12-13 Amp load with about 12V on the loaded battery, to be more typical of a component failure than a sustained overload.

But I defer to the experts on that. The Castle Creations out of warranty repair process has me filling out a form that accurately describes the battery, motor, and other electronics involved and I'll do that providing accurate and honest figures. Maybe they'll tell me if they think I abused anything here.

I like the Castle Creations Phoenix line and have a Phoenix 10 too. To me, it sort of boils down to my choices be that of buying the cheap imports, dealing with programming limitations that can be pretty complex and bizarre from one to another, and simply throwing them away when the fail. Or buying the Castle Creations stuff because it is much easier to program and has the warranty programs (which is pretty darn good, either in or out of warranty) to fall back on.

I'm not happy about having to spend $25 and wait up to four weeks to get a repair on an item like that but I know that is easier than trying to get a warranty or out of warranty repair from a non-U.S. seller of a cheaper ESC.

But winter is setting in and it is not the peak of the flying season for me and I'll wait out the four week (worst case) turn around.

I'm fairly new to electric RC, this is really my first season. Maybe the Phoenix line is not as good as I think it is. I have been shunning the Thunderbird line because they don't seem to be as well thought of as are the Phoenix.

I apologize if this is getting a little off topic but I think it does go to the topic as far as the reasons for and techniques of doing in shop testing.

Jack

[jims123]

Thanks guy, I had a plane recently go down and catch on fire 10 seconds into a maiden flight, so am pretty interested in the pre-test issue, and eager to hear about ways to prevent it.. :rolleyes: JimS