Harmonics in a 12 V brushed DC Motor
I got a small 12V DC brushed motor and collected its current data. After performing the FFT of the signal, I found there are multiple orders present many of which I didn't expect. as motor has 12 commutators I saw a 12th order and maybe due to shakiness of brushes I saw some 24th order too. Also, there was a 5th order, which my colleague suggested is due to the sin winding on motor. but there are other orders like 7, 19, 36, etc
It would be great if someone can explain the presence of these orders
Thanks
Please explain what you mean by 12 commutators?
A commutation is created with a pair of brushes and the slotted slip ring. I believe that you really mean 12 slots in the slip ring.
It would be nice if you showed an image of the FFT. Some of it may be related to the poles.
But also it could be leakage from the FFT?s windowing function.
But we are only guessing.
As I know brushed motors have totally stohastic spectra. Spectrum peaks depend on running time, temperature, voltage, loading and everything else. But I am not an expert of those.
There is the commutation frequency and its chop-harmonics,
but there is also broadband emission from the spark-gaps the
commutator has by construction. And these "noise bursts" are
rate-modulated, probably mixing with the commutation products.
Hi,
Without detailed information about your test and measurement steup we can only guess.
RPM, anti aliasing filter, measurement time, sampling rate, windowing function, complete results...
It could be a measurement error, like alias frequencies,
It could be the the balls of the bearing.
But a correct setup will show you a lot of good informations.
There will be noise, but there will be a lot of true overtones.
Btw: FFT on the motor current is done with reliability pumps for example, they early detect bad bearings, defective propeller and so on.
Klaus
Yes I meant 12 slots in the slip ring. Unfortunately I wont be able to share the picture which i really wanted to, however, what orders I have mentioned are harmonic orders with relatively much higher magnitude compared to the many other orders with very less amplitude these happen to be the dominant orders.
hope that gives some more info
Thanks :D
I am sure that there is no measurement error or aliasing, actually for accuracy we have sampled much higher than highest frequency. windowing was done for a constant velocity region so that also is not an issue.
Thanks :D
Hi
can you explain in little detail about chop harmonics and mixing with noise bursts. Your perspective see very relevent.
Thanks :D
Hi,
Only you know whether this is true or not.
My concerns:
* How do you know your "highest frequency"? Like HF noise cause by the sparks. It may be in the high MHz region.
* "Measurement window size" and "windowing function" are completely different things. Both need to be treated carefully.
(For FFT "windowing function" do an internet search. You may use keywords like "hamming" or " von Hann").
Since you hide your measurement parameters ... I'm guessing they are too secret to be shown in a forum.
Then we can can not validate them..
Klaus
My interest was mainly in commutation order or its biproducts. Hence have kept the sampling frequency 100 times of highest motor mechanical frequency.
I understand that as a first attempt everybody will try to validate the signal and see if the components like 7, 19 and 36 could be just noise components but they are not as they are dominant over multiple tests at various conditions. The ideas was to get opinions from the members here if they understand anything about why they could exist if they are not a component of noise or other disturbance. In this behalf, I somewhat agree with the answer provided by @dick_freebird. however, I have requested more information
It would be great if you can provide some points from that thinking prespective
Thanks