Here's a new one - at least I have not heard of it before.  I think my experience is worth sharing so here goes:

My Nephew Archie has been flying an Oriental (see thread "Shocking the Oriental)  set-up with a Brodak motor, 3 cell Brodak pack, and a Kontronic ESC and it runs PERFECT.

As a proof of concept, I recently built a Vector(ARF) and converted it to a 3 cell Brodak pack, Brodak motor, but used a Turnigy ESC.  Initial test were promising - PLENTY of power available, but the power could not be sustained.  The Vector runs every bit as well as Archie's Oriental at the start of the flight but will not hold RPM throughout the flight.  Power goes down until at the end it is barely flyable.  Did not help that at Brodak's it was in the 90's and dead calm with the wind was mostly from passing thermals - no wind-flying THERE!

* I tried slower initial RPM's and a cut-down prop to reduce battery loads but could not make an improvement.  Later I determined that I was using only 2100 mAh or so per flight when Archie's Oriental was using around 2800 mAh; something in the system was choking off the energy flow.

* The long and short of it is that the Turnigy ESC is rated at 40A but it would not sustain even a 29A load (on ground) due to overheating.  At the start of a run (on ground) the current draw would be at 29A, seemingly plenty safe for a 40A ESC right?  However, within first couple minutes or so the ESC heated up, the sustainable current dropped to around 20A and was still falling when I shut it off.  By that time it had shed over 1000 RPM!

* These are static loads, but I believe the same thing is happening in the air, albeit over different current levels and a somewhat longer period of time.

* Thinking the shrink wrap made a nice insulator; I cut off the shrink wrap and discovered a 1" square heat sink that had been covered up.  Nice place to hide it. 

* With heat sink exposed and ESC re-installed (internally but with its own air scoop the sustainable load only fell off to around 23A; better but still NFG.

* Relocated the ESC to a surface mount exposing the heatsink.  Still would only sustain about 25A.  However this was the best I could do in field mods.

* Came back at the problem from a different angle, the Eagle Tree data on the E-Oriental with same power showed an AVERAGE flight load of around 17A.  I'm running a similar RPM on same prop, on a similar size and weight airframe and similar (beginning) lap times so I expect similar results.  I re-set the ground RPM to an 18A setting - and the ESC was able to keep up.  Thus I think it will now be OK - depending on if/how often the current goes over 25A and how well it cools down between maneuvers, but did not get to fly it yet.

* Thus, just like IC engines, the enemy is HEAT. 

So what do I THINK I have learned?
* I think a contributor is that it is the Vector is set up with 3 cells - the current loads (and heat?) are higher than they would be with a 4-cell system driving the same prop. I consulted with our fearless moderator on that point, Dean's Reponse (clipped) follows:

"There are 40 amp ratings ... and then there are 40 amp ratings.  There's a whole lot of marketing going on!"

"There are two sources of heat in the power transistors. The first is the ON resistance of the FETs and the current load.  That current will be about 33% higher on 3S than on 4S, so the power dissipation will be about 75% higher. (I^2*R) The power dissipated in the ON/OFF transitions will be higher with 4S than with 3S, but every design will have a different mix of conduction and switching losses. It sounds like conduction is dominant in the Turnigy."


I have an untested 4-cell set-up at home and plan to run the same experiment.  4 cell versus 3 might be the tie breaker I need, but I have already removed the shrink wrap and moved the ESC to the surface.  My Father also has prepared some finned heat-sinks I can add to substanitially increase the cooling area if needed.

Stay tuned!

Sounds like the Turnigy ESC is more like a 25A instead of a 40A.  Maybe they mixed up the labels?

Dennis,
Here is a link to a mod to the Align ESC.  Some useful info concerning attaching the heatsink.


http://www.wavelandps.com/stuff/

Dean,
That's what I really think.  The el-cheapo ESC's are over rated.

First a comment on my chea-pass approach: Guilty as charged - but lemme throw myself on the mercy (yeah, right) of the forum!

* Turnigy 40 is $35, Phoenix 45 is $115(?) or so.  Turnigy 60 is $47.  Maybe its worth "oversizing" so I can keep the following...

* Turnigy uses a $7 programming card, the Phoenix needs an interface and a PC.  The Turnigy (& Will Hubin timer) has to be the EASIEST system to set-up there is - just like setting a needle valve.

* Turnigy enables brake in Governor Mode; Phoenix does not.  Frankly this meant a LOT to me.  It would be so blasted simple if Castle would get off their duffs and release the ECL mode they promised over a year ago!

Crist: Thanks for the info/link I'll check it out after work.

Dean: I plan to look into the thermal compound and thermal epoxies...

Dennis,
Another point, if I understand your setup correctly, is that you are using the Brodak motor (the one sold with the Super Clown). That is a pretty high kV motor (~1500) and is more or less optimized to turn relatively high rpm's (>12k). If you are actually turning a "real"  sized prop (10"?) on your Vector, I am guessing you may be turning at a lower rpm, lets say in the 10 k range.

Anyway that means your ESC is probably set at a quite low effective throttle setting. The other thing to remember is that the current you measure with a Wattmeter (or the EagleTree data rcorder) is the average current. However you are drawing current from the battery only half the cycle or so, so what the motor is seeing and the ESC is supplying is actually more than what you think. Since heat is I^2R, twice the current for half the time still means you have twice the heat than you would calculate for a real steady current.

This may be your problem. I suggest you read this faq from CC.
http://www.castlecreations.com/support/faqs/faq-phoenix.html#gen2

Dennis has a couple really custom heatsinks and thermal compound heading his way. Big hacked the heatsinks out of an (computer geeks will know) old skived aluminum cpu heatsinkk I had laying around. They are about 20x25mm and 20mm high fins. He thinned the base out to lighten it then cut it up into the individual parts. If cooling the esc will help these will cool the esc. Crist, the link said it all, thanks.

Dennis,
Another point, if I understand your setup correctly, is that you are using the Brodak motor (the one sold with the Super Clown). That is a pretty high kV motor (~1500) and is more or less optimized to turn relatively high rpm's (>12k). If you are actually turning a "real"  sized prop (10"?) on your Vector, I am guessing you may be turning at a lower rpm, lets say in the 10 k range.

Anyway that means your ESC is probably set at a quite low effective throttle setting. The other thing to remember is that the current you measure with a Wattmeter (or the EagleTree data rcorder) is the average current. However you are drawing current from the battery only half the cycle or so, so what the motor is seeing and the ESC is supplying is actually more than what you think. Since heat is I^2R, twice the current for half the time still means you have twice the heat than you would calculate for a real steady current.

This may be your problem. I suggest you read this faq from CC.
http://www.castlecreations.com/support/faqs/faq-phoenix.html#gen2

INTERESTING!  Lets try this from another angle - that the ESC is not really at fault at all, but that the way I am running it is.  Put another way, no matter WHOSE ESC I would have used the current loads and switching would lead to LOTS of heat build-up.  I know Dean talks about successfully running his Phoenix's at or above their ratings in some RC apps, but those are (I think) WOT, a near-steady state condition which while hot, may not be as hot as sustained operation at some midpoint setting...

Does that sound plausible?

Yes it is plausible, but my Phoenix 35 survives this type of abuse.

For example I am/have run kV motors or 980 and 1080 with a 4s battery, more or less using the ESC as a "transformer" of the higher voltage lower current into a lower voltage higher current setup. I lose some efficiency, but the CC FET's are up to the task.

If you recall when I did my motor tests (http://stunthanger.com/smf/index.php?topic=9256.0), I compared how they ran with both a 3s and 4s battery (at the same rpm). I didn't give the current in that post (just the watts), but to illustrate what is going on, my Scorpion 3020-12 (kV=1080) drew 17A on a 4s Lipo @13.7V  and 23A on a 3sLipo @10.1V. Since the motor was running with the same rpm and load for both measurements, its actual current in the windings had to be the same (>23A since even on 3s, the throttle was not at full on), even though the measured battery current was different.

So the point is that the current we measure with data recorders and wattmeters is only the average current, not the instantaneous currents. Only at wide open throttle (WOT) are the two the same. Since heating goes as the square of the current, you will underestimate its value if you use only average values. Another detail is that current flows in the motor even when the battery current is not--because there is still a path for current to flow through the FETs when they are turned off.

Best option is to try and match the motor kV to the battery selection--to minimize heating losses anyway. Nominally it would be best to be running at ~75% WOT so that we have the extra capacity to keep rpm up in the overheads. I also think this 75% is at the end of the flight, when the battery voltage has dropped. That's when the overhead eights and 4-leaf clover need that power.

To attach the heat sink, use thermal epoxy.  There is some good stuff, but it only comes in fairly large containers.  Black when mixed.

There is also the Artic Silver (the epoxy version, not the standard paste) that works fine.  While it supposedly non-conductive, I would try to keep it off the MOSFET pins.

The "family joules"?  Is that like when  Micro Henry took Millie Amp for a ride on his megacycle over the wheatstone bride and... (edited for family content)         

I would consider the Nobler and Oriental to be virtually interchangeable with respect to thrust needs.  I do not think his battery usage has been too far out of line with yours all things considered.  A couple possiblities affecting draw:
* Flight time set at 6 minutes (almost 20% longer)
* 63 ft of line eyelet to eyelet, probably 66' handle to center of fuse - roughly 4 ' longer radius, however...
* 5.3 sec lap time (has flown from 4.9 to 5. Given difference in line length this is pretty close to same air speed.
* Battery usage has clearly gone DOWN in the wind (at Brodak it was nearly dead calm)
* Has flown between about 46-52 oz depending on equipment on board (3 diff motors, 3 diff ESC's, with & w/o Eagle tree, etc)
* 10x5 prop, but the one he is using now requires a different throttle setting than one he had (and broke) last fall.

Other "factoids" I do not know how to address but which probably have an impact:
* Brodak 3s/4000 battery versus what ever kind you were using - I think there are differences.
* "Too tall" kV motor throttled down to usable RPM; probably results in at least some current being lost to heat & switching - something you are not having to deal with?
* The Oriental still has the original cowl cut for a Brodak 40 - there is a big old draggy hole in the bottom!  The aerodynamicist in me cringes...

The Oriental as Archie has refined and flew at Brodak's worked perfectly.  I think it serves as proof that success is achievable even when using sub-optimized components, they can be meshed into a solid system.  The Vector experience showed I had another big element to consider - but setbacks like that are OK - if/when they are overcome!

...and THANK YOU (and everyone else) for the ongoing commentary/discussion.  I THINK this is how the state of the art evolves!