Motor and ESC Performance Testing

miniquadtestbench.com

MQTB Facebook Page

Mini Quad Power Systems Facebook Group

UPDATE 09/13/2016
I made some changes here to the first post and sub posts, and I will be making more soon. I have also created a Facebook page for news and updates, and a facebook group for further discussion of miniquad power systems. Check them out above!


UPDATE 11/20/2015
I've started a more permanent website for the results of my testing. This project has expanded significantly since I began, so a searchable indexable and otherwise expandable solution was needed. Check it out over at:

miniquadtestbench.com

Summary and Motivation
I've been frustrated for a while that there seems to be a dearth of data out there related to brushless motor and ESC performance on a variety of setups. Some community members have been great about providing thrust data, but I've always been looking for something a bit more comprehensive. As I seem to have collected a pretty large spectrum of Motors and ESCs, I've decided to do a bit of testing myself. This thread will be a location where I will be updating data as I collect it and take suggestions for further research.

Since the site launch I've dramatically expanded the initial test banks, data, and methodology used in testing. The website has taken over the roll of the main outlet of all the data, as well as details on the equipment and methodology. This thread serves primarily as a discussion of miniquad power systems and news and updates on the testing.

SPEEDY THINGS GO FAST!

I'm constantly trying to find this post, so for an interesting discussion on how speed is produced on a quadcopter, check out this post: https://www.rcgroups.com/forums/show...postcount=3268

Reserved...

Reserved.

Nice Work!

I haven't had a chance to install my Bearhug 2.0's yet. I was really hoping to hear some good test results on them. Hey, they aren't frying so i guess thats a good thing.

Originally Posted by freqdrive

I haven't had a chance to install my Bearhug 2.0's yet. I was really hoping to hear some good test results on them. Hey, they aren't frying so i guess thats a good thing.

Yeah, and they're flying fine on the quad. That setup is a lot lower current and RPM though. So I guess the moral is, 4S with light damped enabled, keep it under 17A and under 24000 RPMs.

Thanks for all your work you put on this tests and giving the results to us.

Is the new BlHeli beta only used with the 2 Xrotor ESC, or also with the others?

Regarding the Bearhug ESC when there is a current limitation to 15-17A it might still be used for 1104 motors and 3020 props where the current should be even at 4s < 6A.
Also 1306 motors with GF4045 props might still work with a peak current ~12A at 4s.

But both are high kV motors and if RPM is also limited to 24000 the Bearhug ESC will also not work for these setups with 4s and damped light.

QuadMcFly absolutely respect the effort you have made in this testing.
Looking forward to the KISS18 and 30 test results especially compared to the RG20 and Little Bee.
Results for Max thrust, Max Amps, Watts, RPM etc are not that difficult to obtain with the correct equipment, I like the way you ramp up, then step up the throttle. I wonder if there is a way you could analyse these results to show a deviation from the command signal. Ideally averaging a little and giving a % error from perfect. OR rate of accel value and decel value?

GREAT WORK QMF !!!!

Originally Posted by Exec

Thanks for all your work you put on this tests and giving the results to us.

Is the new BlHeli beta only used with the 2 Xrotor ESC, or also with the others?

Regarding the Bearhug ESC when there is a current limitation to 15-17A it might still be used for 1104 motors and 3020 props where the current should be even at 4s < 6A.
Also 1306 motors with GF4045 props might still work with a peak current ~12A at 4s.

But both are high kV motors and if RPM is also limited to 24000 the Bearhug ESC will also not work for these setups with 4s and damped light.

I'm not sure whether the limit is current or RPMs or whether there is a relationship between the two. It will require more testing to be sure. The highest KV motors I have right now are 3100kv. Might test that out to see what it does. On this setup the reboot seems to happen at about those limits though.

Originally Posted by dschafer

I wonder if there is a way you could analyse these results to show a deviation from the command signal. Ideally averaging a little and giving a % error from perfect. OR rate of accel value and decel value?

GREAT WORK QMF !!!!

Unfortunately right now I have no way of syncing input throttle with the rpm data. That is a goal moving forward. I plan on analyzing accel and decel at the motor though when I get my optical RPM sensor in. That shouldn't be hard at all. My only worry is that my resolution is only 100ms. Some accel and braking may be happening faster than that. We'll see if it is enough. I already know it's missing a lot of the current and voltage spikes. I catch some of them when they happen to fall right on that 100ms sample, but I know there is more to see if I had a higher sample rate! I'm trying to keep the testing reachable and reproducible though, and I'd like to not spend a fortune on test equipment

Originally Posted by QuadMcFly

I'm not sure whether the limit is current or RPMs or whether there is a relationship between the two. It will require more testing to be sure. The highest KV motors I have right now are 3100kv. Might test that out to see what it does. On this setup the reboot seems to happen at about those limits though.

Yes, the 1306 are 3100kV and 1104 are labled 4000kV but i don't if that is real.

In that test the 1104 with 3020 props on 4s show ~ 27250 rpm: http://rotorbench.com/DYS_1104/Summary.txt

I don't know how much poles that motor have, but it maybe comes close to the eRPM Limit of BLHeli with the used ESC.
I think that test is done with SN20A ESC, so maybe already close to the eRPM limit of that ESC. Silabs based ESC especially the ones that run at 48/50MHz (like the Bearhug) should be capable of more.

Originally Posted by Exec

Yes, the 1306 are 3100kV and 1104 are labled 4000kV but i don't if that is real.

In that test the 1104 with 3020 props on 4s show ~ 27250 rpm: http://rotorbench.com/DYS_1104/Summary.txt

I don't know how much poles that motor have, but it maybe comes close to the eRPM Limit of BLHeli with the used ESC.
I think that test is done with SN20A ESC, so maybe already close to the eRPM limit of that ESC. Silabs based ESC especially the ones that run at 48/50MHz (like the Bearhug) should be capable of more.

Yes, I agree, he's hitting the RPM limit of the ESC for sure.

Love it. I just picked up all the parts to make my own data logging thrust stand last week. I'm hoping to be able to chart reaction time accelerating and decelerating, as well as seeing how they hold up to high frequency oscillation.

The source code link in your google sheet seems to just link back to the same google sheet? Are you using the servo library, or generating the raw PWM from scratch as well?

Originally Posted by extent

Love it. I just picked up all the parts to make my own data logging thrust stand last week. I'm hoping to be able to chart reaction time accelerating and decelerating, as well as seeing how they hold up to high frequency oscillation.

The source code link in your google sheet seems to just link back to the same google sheet? Are you using the servo library, or generating the raw PWM from scratch as well?

Sorry, broken link, will try to get it updated tonight. I'm using the 16bit timer to generate the PWM signal.

Your results are very similar to what I see. The XR40s are king of the power but are large and heavy. I use them on my 6S rig. The XR20s have been my favorite ESC so far. The power is very similar to the XR40 but they are small and incredibly reliable. Braindrain and myself have about 60 between us and it's what we've been using for racing. We have yet to have an XR20 fail on us. This smashes the reliabilty record of any other ESC we've ever used. I'm also a huge fan of the F39X series MCU the Xrotors sport. I believe this MCU is going to lead to some really cool features in the future. After this last generation of ESCs I hope the Atmega and F330 MCU finally get retired.

I have a set of MRM20 ESCs I'll be flying soon. I was going to have them on flying last weekend but my last polulu reg was bad. I love their small size compared to the Xrotors. They may be my next go to ESC until something better comes along.

The issues with the XR's throttle on the bench do not show up in flight (At least with OPNG firmware). It's something I've noticed with BLHeli in general for a while. More than just the XRs are affected by it to some degree or another. The dithering helps, but it's not the final solution to the problem. You can really see the problem on other ESCs if you go back to BLHeli 13.2.



Something I'd like to see tested is braking. I have dozens of different types of ESCs and have yet to see one match the incredible braking power of the XR40. I test braking by yawing back and forth quickly in flight. The faster I can yaw back and forth without getting altitude gain the better the braking is. I'd like to see a braking test performed on the bench.

This is how I would rate braking like this based on that yaw test. I don't know where the MRM ESCs rank yet:

Level 1: XR40(BLH14),
Level 2: RG30(BLH14), RG20(BLH14)
Level 3: XR20(BLH14), KISS
Level 4: SN20A(BLH14), BS(BLH14)




As far as testing, I've been wanting to use and STM32Discovery board for a while as a tester. It could generate a set program that it runs every time on each ESC and could measure from several sensors (RPM, thrust load cell, voltage, amperage, temp) and record the data to a PC via serial. I haven't had the time to do something like that though.

How are you measuring thrust? Are you measuring peak thrust or the thrust after the ESC settles down. I've noticed the smaller ESCs tend to settle at a lower RPM than larger ESCs do.