This thread is privately moderated by Dr Kiwi, who may elect to delete unwanted replies.
Nov 22, 2010, 03:15 AM
Registered User
Antony (France)
Joined Sep 2003
3,127 Posts
Quote:
 Originally Posted by Heliosupersix Hi AUG I am still stuggling to understand why six examples of the same motor and manufactured to a fairly comprehensive specification can have any measurable parameter that is so variable. Taking a scientific approach one would expect them to be identical in all respects. OK you call it cogging but say that it says nothing about how the motor will perform. That's the bit I don't get. Ultimately I have an open mind I just need to understand why. Thanks for your patience.
Hi Helio
Try to measure the cogging torque on the same outrunner but changing the position of your rotor relatively to the lever (your graduated rule)
See if your cogging torque results are repeatable (for successive peak values CW or CCW) along the rotor angle.
Louis
 Nov 22, 2010, 03:53 AM Registered User Joined Nov 2010 16 Posts Hi Fourdan If you mean what I think you mean I have already done it. Using a 1cm drum wound with thread. A force was then applied to the thread until the rotor moved one "cog" and then on to the next etc.until it had revolved through a full 360 degrees. Am I on your wavelength? How are you interpreting CW and CCW ? Thanks for the positive suggestion. Jon
Nov 22, 2010, 04:25 AM
Registered User
Antony (France)
Joined Sep 2003
3,127 Posts
Quote:
 Originally Posted by Heliosupersix Hi Fourdan If you mean what I think you mean I have already done it. Using a 1cm drum wound with thread. A force was then applied to the thread until the rotor moved one "cog" and then on to the next etc.until it had revolved through a full 360 degrees. Am I on your wavelength? How are you interpreting CW and CCW ? Thanks for the positive suggestion. Jon
Hi Jon
Yes we are on the same wavelength ..
But I think (like some others) that the cogging (at no current) is not easy to measure with precision and repeatability and so not very representative of quality.
What about your results by figures versus rotor position and repeatability ?
For me if you test Rm (Ohm law) and no load rpm and current versus V, you have a good idea of any motor and can compare.
Louis
Nov 22, 2010, 05:01 AM
Sic itur ad Astra
United Arab Emirates, Abu Dhabi, Abu Dhabi
Joined Aug 2009
1,605 Posts
Quote:
 Originally Posted by Heliosupersix Hi all As requested
Just a few thoughts on this subject.

Looking at your test rig Helio, it would seem to me that there is potential for error if you are trying to measure 'cogging' force. As the distance from centre to the measuring link rod is small, the motor would need to be precisely aligned each time otherwise a small differential angle could be seen between each measurement. Any angle will affect the displacement of the actuator slightly, which I think may have significant effect.

I think an actual load test is the only useful way of measuring motor performance, though I too am surprised at the differences noted between 6, apparently identical motors.

I don't know, but suspect that these motors are not actually made in one factory, there are probably several, but many of them buy in the alloy and steel components from another supplier(s). Thus, the magnets and windings may come from other (possibly multiple) suppliers. As a consequence, the motors all come out looking alike with very similar specifications, but are not necessarily exactly the same.

sparks
 Nov 22, 2010, 06:44 AM Registered User Joined Nov 2010 16 Posts Hi Louis Working on some figures versus rotor position.You didn't say what CW and CCW are. Hi Sparks I am aware that the rig is not perfect and nor is the technique used during the tests. All I can do is to be objective and do the same, warts and all for each motor tested So I could not say for certain that my Gms/Cm are spot on, however the errors are the same for each motor tested and it is the difference between the motors that I am interested in. The "feel" of the motors confirms that there is a difference. These are just my first attempts to put those differences down on paper. Your last para says what I am thinking in a nutshell. You get what you get and the origional spec goes out the window. I have spent some time reading past threads on the subject and also looked at some published papers on cogging.There is a 2002 thread from this forum in which a designer of very large motors describes the attempts to design out cogging (detent) to make the motors smoother to operate. He describes the downside of the changes as follows: "The drawback is loss of torque, high detent motors have higher peak torque and better efficency than low detent motors." http://www.rcgroups.com/forums/showthread.php?t=28021 I can readily understand why this is might be so. My tests are not looking at differences in different motors of different designs they a looking at differences in supposedly identical motors. Therefore I have yet to be convinced that the 8 Grams/cm motor can give the same performance that the 19 grams/cm motor does. Many thanks for you input. Jon
 Nov 22, 2010, 07:11 AM Sic itur ad Astra United Arab Emirates, Abu Dhabi, Abu Dhabi Joined Aug 2009 1,605 Posts Hi Jon, I think we are broadly in agreement and can understand the argument for the higher peak torque for a given winding at a given flux magnitude. But there's the thing that makes me think about the varied suppliers. Assuming that the motors have the same stator, rotor, shaft dimensions and materials (which may be an assumption too far of course), you might surmise that the 'air gap' between rotor and stator should be the same, at least on same models from the same brand. If that is true, unless i've missed something, then it really only leaves the quality, or perhaps the position of the permanent magnets to affect the level of detent. Thinking about that a little further, I guess it is possible for the material of the rotor and indeed the stator to also come from different sources and introduce another variable that could affect performance. In terms of perfomance, lower detent levels can be balanced by increasing voltage/current, ie increasing the magnetic flux density to induce the required torque. I'd be interested to see the results of a performance test between the 8gm and 19gm motors, at a variety of speeds and loadings. Going to sit in a quiet room now 'cos my brain is aching ;-) sparks
 Nov 22, 2010, 07:23 AM Sic itur ad Astra United Arab Emirates, Abu Dhabi, Abu Dhabi Joined Aug 2009 1,605 Posts [QUOTE=Heliosupersix;16616969]Hi Louis Working on some figures versus rotor position.You didn't say what CW and CCW are. Clockwise...I think you'll get the other one OK Jon
 Nov 22, 2010, 08:26 AM Registered User Joined Nov 2010 16 Posts Thank you Sparks And if the motor is increasing the magnetic flux density to induce the required torque then it is going to get hotter, waste more juice = less efficient ? Have to do some more testing. Thanks too for CW....thinking to deeply elswhere. At least Louis did'nt have the pleasure. Jon
Nov 22, 2010, 11:14 AM
Registered User
Blackpool, Great Britain (UK)
Joined Dec 2003
1,342 Posts
If a precise plot of cogging torque were made, I think it would look something like the atachment.

It would show both +ve and -ve peaks, which would be of equal value.

The average torque, over one complete rotation, would be zero. This is regardless of the magnitude of the peaks.

Jon's test indicates only the magnitude of the peaks, and only one way. It gives an incomplete picture.
--------------------------------------------------------------------------

At the conceptual level, the notion that the average is zero should make it easier to accept that cogging does not influence motor performance.
--------------------------------------------------------------------------

To anyone who enjoys playing with test rigs, I suggest you make a torque balance.

Running a test and plotting the measured values helps you to understand your motor.

http://www.rcgroups.com/forums/showt...=645168&page=4 post 52

# Images

View all Images in thread
Nov 22, 2010, 03:05 PM
Registered User
Blackpool, Great Britain (UK)
Joined Dec 2003
1,342 Posts
Quote:
 Originally Posted by Heliosupersix ........ please explain to me how one with 50% less strength in its permanent magnets can operate to the same specification.
Jon

You are assuming that variations in cogging torque [between individual motors] are due to variations in magnet strength.

I suggest that variation is more-likely due to variations in the gap between the stator poles and the magnets.

The variability of the gap is the sum of at least three manufacturing tollerances.

The attached picture shows how tiny variations in three dimensions can produce a two-to-one variation in the gap.

# Images

View all Images in thread
Last edited by peterangus; Nov 23, 2010 at 03:52 AM.
Nov 23, 2010, 01:36 PM
Registered User
Joined Nov 2010
16 Posts
The vertical rig was unsuitable for the rotational measure so I came up with a Mk11.

I found a 10mm drum so the figures need to be halved to get grams/Cm. Gearing up actually made the process easier.

The benchtop stabilisers one axis the guide the other. A 10mm spacer goes in when the thread is wound on the far side of the drum.

The rest was just some steady and balanced thumb pressure!

The figures speak for themselves, ran out of time to do a full test, but you will get the idea, full results to follow.

Jon

# Images

View all Images in thread
 Nov 23, 2010, 03:13 PM Registered User Joined Nov 2010 16 Posts Hi Peter Yes I think magnet strength must play apart whether directly or indirectly. The way you have drawn how the tolerences can stack up I agree must play a big part too. With CNC machining it should not be too difficult to control these but they can't or won't do much about under strength magnets except perhaps use them when they shouldn't? I am trying to get my head around a posting you made earlier.... "At the conceptual level, the notion that the average is zero should make it easier to accept that cogging does not influence motor performance". Yes OK it produces a moment of resistance to rotation then assists rotation and one balances out the other. But am I right in thinking that motor strength is dependent on magnet strength, if there were no magnets at all then the windings could do what they liked and there would be no rotational force since one reacts directly with the other. Maybe cogging does not directly influence motor performance but it just gives an indication of the condition of the magnets that do. A secondary situation ? Thanks for your patience and your drawings. Jon
 Nov 23, 2010, 07:44 PM Curiouser and curiouser Rochester, NY, USA Joined Oct 2005 2,066 Posts For the strength of the motor vs. magnet strength: I doesn't seem to me that an average of zero is important. After all, the ESC keeps the current driven magnetic field rotating at the rpm of the motor and makes the permanent magnets move to keep ahead (or behind) the rotating field - always on the up (or down) slope - never going over the top (or cogging). So I guess that the peak strength of the permanent magnet is important. At less than WOT the height of the individual peaks are not important -- don't even have to equal. BUT, nearing WOT as one or more of the peaks is surmounted slipping will start to occur. Okay, this is rambling - am I on the right track, eh? Walt
Nov 24, 2010, 01:45 AM
Registered User
Joined Nov 2010
16 Posts
Hi Walt

I think you are but then I am only the spanner boy around here? It's the Profs that you have to convince.

"At less than WOT the height of the individual peaks are not important -- don't even have to equal".

Interestingly from my measurements,they are not. I have averaged out the three readings I did for each station, then taken the average of all the readings for CW and all for CCW then taken one from the other.....I never get zero.

For supposedly two identical motors there seems to be rather a lot that isn't.

Jon

# Images

View all Images in thread
Nov 24, 2010, 01:52 AM
ancora imparo
Melbourne, Australia
Joined Jul 2005
7,730 Posts
That's one of the first, and briefest, derivatives of Murphy's (Sod's in the UK) Law.

"Interchangable parts won't."

Quote:
 Originally Posted by Heliosupersix For supposedly two identical motors there seems to be rather a lot that isn't. Jon