Testing made possible by generous support from:

ESC Shoot Out: Cobra 2300kv HQProp 5040GF

Cobra 2300kv and HQProp 5x4 Full Results [CLICK HERE]

For my first test I have decided to use a Cobra 2300kv 2204 motor with an HQ5x4 propeller. This combination is one of the most stressful on ESCs due to the combination of high RPMs and quite high current draw (~20A). This should give a fairly clear picture of the upper limit of switching speeds at 4S. (Eventually I would like to test on a >3000kv ESC, which will push things even higher in terms of RPM, but lacks the current draw to push both simultaneously.) Combining this with the PWM generating code on the Arduino that includes instantaneous throttle jumps from 25% throttle to 100% throttle turns this into a true torture test as a first shot.
The XRotor ESCs are definitely the clear winner in this, especially the XR40A. The KISS came in a close second. They give a tad less average thrust at peak, but their RPM lines are extremely smooth. They give the smoothest running out of all the ESCs tested. The Atmel based ESCs all had a hard time keeping up with the high RPMs of this test setup. You can see jagged RPM lines at the top end of the throttle in all these tests. The Atmels all struggled on top end thrust as well.

I was really surprised in this test by the Afro 20A Race Spec Mini ESCs. I read through the datasheet when I first got them and was quite concerned by the amperage rating on the FETs they used, as they are only rated for 39A peak at 100 deg F. By comparison the FETs on the RG20 are rated 100A peak and the MRM Zuess and SN20 are rated 50A peak. So far in my testing though this hasn't seemed to be an issue. They're close to the same price as the DYS SN20A and seem to run cooler and have basically the same performance without the reliability hit we've been seeing. The build quality is WAY higher. I was a little nervous about the ratings on the FETs, but they did very well on the bench even with my crazy transverse throttle tests! Airtime will tell though, but I'm much more optimistic now than when I was reading the datasheet. The 20A Race Spec and the Afro 20A High Voltage tied for the the top thrust performance in the Atmel based ESCs. See the results above for the whole story.

A disappointment here was the BearHug v2. I've been flying them on a quad for a couple weeks now without trouble, and I was hoping the latest revision had fixed the issues that have plagued the BearHugs from v1 and 1.5. So far on the quad it hasn't had trouble, but in this test it started rebooting with light damped enabled as soon as it hit 15A or so of current. This was even in the slow ramp part of the test before the transverse throttle tests. It hasn't failed outright yet, but the rebooting was a precursor to failure for me on the v1 BearHugs. Myairbot.com did include a 220uf 35V low ESR cap with the ESC, but they seemed to infer that it would only be needed on >4S. I'll solder it on here and give it another go. All this being said, they have been performing fine for me in there air on 1850kv 2205 motors with KingKong 6x4 props. I haven't bench tested those motors/props yet, but my guess is the peak current is significantly less than the bench setup here with the 2300kv and 5x4. Anyway, we'll see what else I can find with these, and hopefully I can get some test data. EDIT: Well adding the cap made no difference whatsoever. Still rebooting at the same point. No data for this one. At the very least, they're not frying outright. Just rebooting.

The last issue on this test was with the XRotor ESCs. It seems at high RPMs and high current they hit about 80% throttle and then motor output starts oscillating like crazy. I'm not sure what's going on, but I've never seen anything like that before. Does it on both the 20A and the 40A so it's not isolated to only one. Steffen has sent me some beta code to see if that fixes the problem. Hopefully it does and I'll have test data for those in the next few days. EDIT: Got the beta firmware on and tested! The results are in the table linked above. I was not surprised by the excellent performance of the XRotor 40A. I was however VERY surprised by the almost exactly equivalent performance of the XRotor 20A . Due to high resistance it the FET switching circuitry I expected much less out of the XRotor 20A than what they gave. They still won't be as efficient in braking, but they nailed the high RPMs, reaching exactly the same limit as the XRotor 40A. Very impressive!

KISS ESCs came in today, and I've added the data for those. Very interesting results. While they peak at the same RPMs as the XRotor 40A, the average RPMs over the max throttle are a tad less, as is thrust on the stand (which isn't fast enough to register the RPM spikes) Overall I would say the KISS performed very well. They were also the smoothest of the ESCs I tested in terms of the sound of the motor, especially at low RPMs. Anyway, browse the data for more info!

On another note, I realized I had some errors in my temperature readings due to my measurement techniques, so I re-ran all the temperature tests. Readings should be as accurate as my equipment allows now.

DISCLAIMER: The RG20, KISS 18A and 30A, BearHug v2, and Afro 20A Race Spec ESCs were all sent to me free of charge by the manufacturers for testing. That in no way impacts my testing or reviews of the products, other than the obvious fact that I have them to test and wouldn't have otherwise!