First, thanks for all the nice comments! Before trying to answer all the questions, let me show you the latest color summary table which has been updated in Post #2:



I added another Watts/oz column based on some discussions with Brashley. This column provides a comparison of motor/prop performance for a common thrust level of 16 ounces rather the motor's max thrust. The weight of the motor/prop is also subtracted from the 16 ounces before the final Watts/oz is caculated. This column if very useful for comparision of a motor/prop overall efficiency for typical quad flying.

I also added the published kv value for the motor, the source of the motor and the price. The ranking number is my evaluation of best to poorest motors based on the overall performance across all metrics.

Here are a few notes about some of the motors:

1) KDA20-22L - This motor is just amazing and likely to be the one I end up using in all my quads. It ties with the very popular 2410-09 for flight efficiency, has the highest thrust level and is one of only two motors that had good vibration numbers on both OMM and Rusty test fixtures. While it is almost twice as expensive as the Gold base -09 motors, in a quad configuration it has more that 2 lbs of extra thrust available!

2) DT-750 - I believe this is Rusty's favorite. It looks like it could take a crash better than the KDA20-22L and will run cooler because of the much larger size. This motor has an excellent WQF which should allow it to be more responsive to the controller. This could make for a more stable platform as long as the vibration is low. The high vibration of this motor on my test fixture is a current concern but needs to be verified with another "reference" motor exchange with Rusty.

3) 2410-09 Red base - The -09 motor is a proven high value performer and is what I have used for all my quad videos during the last year. Since we know it works well with both UAVP and Spectrolution quads, it demonstrates that a mid 50's WQF is good enough for stable platform control. (there will be more discussion about what WQF means later in this thread).

OMM

Originally Posted by jesolins

Great work OMM and Rusty! So it appears the 2410-09 series are still the ones to beat for efficiency and cost per weight/thrust/vibration. Now we need to come up with at way to balance the motors themselves and provide good vibration isolation from the motors.
Just wondering, did you use new motors and correct the motors end play and allen screw tightness on the shafts and bases before the tests, or just use them as they came new? Also, how did you make the props consistently dynamically balanced for each motor, in order to eliminate them as a cause for vibrations?? What do the white spaces in the vibration fields represent? No data or no vibrations?
Cheers,
Jim

Thanks, Jim.

You may want to look more carefully at the color summary above since there is little doubt that the KDA20-22L motor currently appears to be a much better overall choice.

Yes, I think the next big thing in quad design will be balancing and vibration dampening techniques. The nice thing about the test fixture is that the vibration and WQF metrics provide a means of comparing techniques.

We tested new motors, used motors, repaired motors, motors known to have low vibration, motors known to have high vibration, motors with various tightness of allen screws, motors at different angles, etc. You can easily tell when a prop is out of balance base on the vibration metric. I use a varible length of tape on the underside of the prop until I get the lowest vibration number.

The white spaces are just areas where the test was not performed or I am missing the data. For example, I have not yet tested the DT-700 motors so there is no Vib-O entry. They will eventually be filled in.

Originally Posted by foliage

OMM,
Great work as usual. I am using Himax HC2812-0650 with EPP1245.
looks lot of thrust at low power, I had to add gromets for vibration with the EPP1245.
Have you tested these motors.I am curious to see were they stand on vibration.
Jack.

Thanks, Jack.

Brashley has been kind enough to send me a few of his motors, including the HC2812-0650. I am also most anxious to see how it compares. We only made a few test runs with the EPP1245 but you can definately see that it reduces the WQF factor (as expected). I think we will try to complete as many different motors with the EPP1045 before testing too much more with other props.

Originally Posted by otlski

... Just how much different are your respective natural frequencies? Considering you guys both used the same Load Cell, the pivot to cell distance will primarily drive the system stiffness. Next comes the rocking inertia of the arms on either side of the pivot. After that, the method of coupling the forces usually dominate. Are the test jig resonances outside the operating range?

Any pictures of the jig Mike?

The wobble tests are most interesting. were the ramp profiles based on data gathered from Eagletree Recorder flight data?

Regards,
Dan

Thanks, Dan.

There are two resonance frequencies that we can see during test runs. One is the quad arm on which the motor is mounted. The other is the test fixture aluminum mount lever arm. The FFT program is most useful for identifying each one. During the development, Rusty and I tried all sorts of different quad arm lengths before choosing a final reference position. While the resonances are mostly outside the area where the WQF and vibration metrics are made, there is still some effect. I was able to filter almost all of it out of the WQF metric but the vibration metric remains a challenge.

I'm not sure what your question is about the wobble tests. My interest here was to determine just how fast a motor/prop can respond to throttle changes. The answer is, not very fast. In fact, it is my believe that thrust changes in response to throttle changes faster than a few milliseconds are lost in vibration noise. Note how small the response is to the 40 ms cycle swings.

Originally Posted by gke

Mike/Rusty,

Nice work.

So I am seeing a rise time on the example of around 100mS is that correct? What is the PWM pulse repetition rate and is there an offset from the trailing edge of the pulse and the motor response? Ditto if I2C control.

Questions relate to control loop considerations of course.

Thanks,
Greg

Thanks Greg,

Yes, you can see the throttle rise time response of the KDA20-22L motor/prop combination is around 100ms. This corresponds to about a 60% WQF metric. The PWM pulse rate in my program can be varied down to a near 2ms repetition rate but I've set it for a 6 ms rate for the testing. However, note that when the throttle is changed, the PWM is updated within a 2ms period. I've tested at near 2ms repetition rates and could see no difference in the risetime response compared to the 6 ms rate. Bottom line, is that almost none of the rise you see in the wobble test is from the ESC. Overcoming the physical rotational inertia is the dominate factor. I suppose there might be a very powerful motor/esc combination where 2 ms thottle changes could significantly change the thrust, but I have yet to see it.


OMM

Originally Posted by Old Man Mike

2) DT-750 - I believe this is Rusty's favorite. It looks like it could take a crash better than the KDA20-22L and will run cooler because of the much larger size. This motor has an excellent WQF which should allow it to be more responsive to the controller. This could make for a more stable platform as long as the vibration is low. The high vibration of this motor on my test fixture is a current concern but needs to be verified with another "reference" motor exchange with Rusty.

Nice additions to the chart Mike! For anyone who doesn't know, Mike is somewhat addicted to Excel

The DT-750 is in fact my favorite motor at the moment. I believe the large diameter gives it a torque advantage, which gives it an edge in the WQF scores. The motors are easy to rewind, and I've wound a few different versions as tests.

The biggest problem is the quality of these cheap motors. I don't think any two are exactly alike physically, though fortunately, that doesn't seem to matter much in their performance. The bearings are pretty loosely mounted, which make it easy to take the motor apart, but adds some play.

It's well known that the winding varies from pole to pole. For example, if the motor is supposed to have 12 turns of wire, you might find anywhere between 8 and 13 turns on various poles. Since the spectrograms sometimes show a prominent component of vibration that could be related to a particular pair of poles, I thought maybe rewinding properly would smooth out the vibration. While it was an interesting experiment, it did not seem to have any effect on the vibration level.

The KDA20-22L appears to be built better, and I'll probably get a set to play with. The only thing I don't like about it is the prop attachment. It has a fairly short, unthreaded 1/8" shaft, as well as a nice prop adapter. Unfortunately, the adapter has a large (5mm I think) prop shaft, requiring a large hole in the prop. I just fear that reduces the already questionable strength of the EPP prop hubs. By default, the shaft sticks out of the wrong end, such that the motor would hang below the mounting plate, and be subject to damage. It can be reversed, but being a better quality motor, it's a pretty tight fit, and the snap ring won't take too many uses. The motor also comes with a spare shaft, which is a nice touch, and makes the higher price easier to justify.

I think Mike answered all the other questions, so hopefully, I'm not missing anything.

Rusty

Mike and Rusty,
Thanks again for your diligence and clarifications and updates. Nice to know that the motor winding errors don't play much in the vibrations. Did you note any other performance reason to insure that the motor windings are exact in number of turns, or is close enough good enough?
Great work Gents!
Cheers,
Jim

Nice work!

Did you see any relation between motor cogging when (when turning motor by hand) and the level of vibration on a test run (either low or high RPM)?

Thanks,

Peter

Very good data indeed!

Regarding your vibration readings. Did you see base RPM vibrations as well? I would have expected to see lots of that from play in the rear bearing and not fully centered rotor and prop imbalances. I ordered 5 2410-09 motors from Hobby City and two of them are so horribly out of balance right out of the box that I'm amazed I can fly the quad at all.

gary

Hi Gary,
Are you sure it is not bent shafts? I've seen that with those brand new as well. Fortunately they are easy to straighten. The end play must be adjusted to no more than .5mm, and all the allen screws tightened with loc-tite. I've always been able to get them smooth, even after "hard landings".
A pre-flight spin-up holding the quad firmly down and checking motor balance is important for good flight performance. Also a pre-flight check of the tightness of the props...or the rottenness of your O-rings whatever the case might be...is important
Cheers,
Jim

Originally Posted by garystofer

Very good data indeed!

Regarding your vibration readings. Did you see base RPM vibrations as well? I would have expected to see lots of that from play in the rear bearing and not fully centered rotor and prop imbalances. I ordered 5 2410-09 motors from Hobby City and two of them are so horribly out of balance right out of the box that I'm amazed I can fly the quad at all.

gary

Thanks Mike. Just what I needed.

Greg

[QUOTE=Old Man Mike]
I'm not sure what your question is about the wobble tests. My interest here was to determine just how fast a motor/prop can respond to throttle changes. The answer is, not very fast. In fact, it is my believe that thrust changes in response to throttle changes faster than a few milliseconds are lost in vibration noise. Note how small the response is to the 40 ms cycle swings.

Understood Mike and agreed you got what you were after. The torque required to overcome rotational drag and MOI is simply not available. I was really attempting to enquire as to how "likely" a test it was. That is to say, if you took your standard craft into a hover and monitored the steering signals to any ESC (via Eagletree), are your recorded flight environment control excursions in the same ballpark as your fastest wobble sweep periods and amplitudes? What does a DF control board ask of its ESC's under flight conditions?

I'm asking out of both curiousity and for the yaw experiments I'm wrapping up.

Regards,
Dan

Originally Posted by jesolins

Nice to know that the motor winding errors don't play much in the vibrations. Did you note any other performance reason to insure that the motor windings are exact in number of turns, or is close enough good enough?

Jim, Best I can tell, close enough is good enough. I never exactly reproduced the stock winding, because they use multiple strands of fine wire, rather than a single strand. I can easily get more power out of the motor, but not without giving up some ground on the other parameters.

Peter, As for the cogging, I've wondered that too, but don't quite know how you could test it fairly. You'd need two motors that are supposed to be the same, but show different levels of cogging. That does exist to some extent, mostly due to magnets not being positioned precisely I think. I had asked about this on one of the magnet threads, and was nearly overwhelmed by the time and effort people have put into trying to measure and explain this. The bottom line was that cogging itself didn't really contribute to vibration, or power level variations enough to notice.

The other thought is timing. Cogging is a static effect, and once you start energizing the poles, that's all out the window. What if the timing was not quite right? If you energize a little early, or late, then it would seem that you'd get a prominent vibration, and power would be wasted as well. I've tried tests with the whopping 3 timing levels, but have never seen much of a difference. There's an ESC called MPower (or similar) that allows very precise timing adjustments, and I've been tempted to get one to use for testing. Unfortunately, it's never made it near the top of the priority list.

Rusty

[QUOTE=otlski]

Originally Posted by Old Man Mike

I was really attempting to enquire as to how "likely" a test it was. That is to say, if you took your standard craft into a hover and monitored the steering signals to any ESC (via Eagletree), are your recorded flight environment control excursions in the same ballpark as your fastest wobble sweep periods and amplitudes? What does a DF control board ask of its ESC's under flight conditions?

I'm asking out of both curiousity and for the yaw experiments I'm wrapping up.

Regards,
Dan

Hi Dan,

I've actually fantasized (for lack of a better word) about recording the PWM signals that go to an ESC in flight, so I could play them back to an ESC on the test stand. It would be really interesting to see what real life is like for the system "in flight".

As for yaw, we realized early on that the test stand can measure yaw as well. Just turn the motor 90 degrees, and run in the direction that puts pressure on the load cell. I did some runs to try to see the difference in yaw authority between one prop and another, but it was just a little side project, and the data is long gone. As I recall, it was pretty darn predictable.

Rusty

Rusty, I sent OMM three of the same motor each with different cogging. The explanation I got from the mfg was that when the magnets get cut from the block not all have the exact same magnetic field alignment and hence if they are all off a bit cogging could be reduced. I am expecting that this will come at a price of efficiency but the one with the least cogging has also been the motor that has giving me the least vibration (or so I think). I was also told that you could reduce the cogging by angling the magnets slightly but I am not a motor expert so I am not sure. These motors will not do well in the max thrust department but that is not what they are good at, they should be good a swinging large props slow. (OMM, it would be interesting if you put an EPP1245 on one, I have not tried one yet on my test stand)

The lower KV motors provide lots of torque at the expense of lower max thrust. I think quad flying puts another spin on motor selection that most thread / discussions are not geared towards. A selection for a plane will be totally different then a selection for a quad and for the most part people talk about quad motors like they talk about plane motors. In my mind I want as much torque as I can get away with while still meeting my minimum Max thrust requirement to keep the quad stable in all flight conditions. (ie I don’t need a 3-to-1 thrust to weight ratio). On average the lower KV motor will provide higher torque and better prop response (OMM’s WQF numbers) and with a smaller RPM range will provide finer control from the ECS. How important is all this, don’t know yet. This is why I am glad OMM and Rusty are doing this type of testing. For my quad, I want the most efficient/highest torque motor at hover thrust…..but that is just me….

Richard

OMM: I am having problems seeing you HTP dirrectory. I tested HTP on some other places and it worked fine. Thanks for posting them by the way

Richard

That's just solid gold. Thanks so much for posting your data guys.

Originally Posted by Mr. Rick

That's just solid gold. Thanks so much for posting your data guys.

Thank you!

OMM