Measuring current diferences when universal motor is sparking

[djpiky]

Hey to all!

I am working on project, where i have universal motor ( https://en.wikipedia.org/wiki/Universal_motor ). I need to make measurment and eliminate all motors which produce to many sparking.

I am using DAQ device, to get input signals (Voltage, current). Then i use Spectral analyse to get frequencyes. But there are really not a lot of diferences between good and bad motor.  How can i detect when motor is sparking too much?

Sincerely!

[johnsold]

By sparking do you mean the arcs which occur at the commutator?  How do you define "too much sparking?"

Commutator arcing occurs all the time in such motors so the key issue is what you choose to define as an excessive amount of arcing.  What is your ultimate goal? Is this a quality test at the point of manufacture or a monitoring process to determine when brushes need to be replaced? Or something else?

Lynn

[djpiky]

Yes, i have universal motor which contains commutator. Where brushes are touching commutator there are usally some sparks produced. If there is something wrong with motor (brushes are not OK, commutator is not OK ...), motor produced a lot more sparks  when running. And this is bad motor, which should be eliminated from production.

This appliaction is quality test for new motors. Motors have max power of 900W.

I am using NI 6321 to aqusite voltage and current from motor. My goal is to prepare VI which will detect difference in current/voltage, when there will be bad motor. I am trying with FFT. Here are results from good and bad motor.

In theory if there is a lot of sparking on commutator, there should be some anomalies in spcetral space.  But i dont get it.

[Daikataro]

I think you would find your solutions by asking yourself the following questions:

-What IS sparking? Physical sparks? Too much current consumption? Stuttering?

-What are you measuring? Power (current consumed on a specific voltage)? Torque? Harmonics?

-If there ARE differences between good and bad motors, however few or little they are... What are they? Are they consistent? As in, will the difference always present itself between good and bad motors? How can you measure said differences, and if possible, amplify them?

Let's say a "bad motor" will consume more power than a "good one", so for example you could feed them with a very specific voltage and see if current consumption surpasses a threshold. Then it is something an "in-range" function can answer with a boolean.

EDIT: So you have physical sparks, have you considered just having some visual inspection equipment recording the test, and measure if brightness/luminosity passes a certain peak?

[Henrik_Volkers]

Sparks have a high frequency content. How fast are you sampling?

Do you have a scope ( I would use an analoge one, but they get rare nowadays, still keep a Tek 466 in my office 😉 ) ?  Or monitor the voltage with the highest possible samplerate.

Sparks usually create high voltage amplitudes, so how do you protect your voltage input?

Just by own experience... if I run small mechanical commutated motors that are not well filtered, and I listen AM radio, I hear the difference. ... so I would AC couple my voltage input with ... say 200pF  and look in the 100kHz to 2MHz region.

If you can't sample that high (10MHz??) look what engineers have done in the past to detect RF: They used a diode and a small cap to rectify the RF.  🙂 First link

900W motors.... mhhm,...  I would try a ferrite antenna harvested from an old AM radio and measure without contact to the high voltage... 

And if you want to do that seriosly: EMC test equipment to test it is on the market 😄

If you can see a difference in the time or frequency domain, but fail to detect it: post raw data.