Author Topic: Tracing PWM Waveform from an Inverter using an Oscilloscope  (Read 2202 times)

0 Members and 1 Guest are viewing this topic.

Offline Vigren

  • Newbie
  • Posts: 1
Tracing PWM Waveform from an Inverter using an Oscilloscope
« on: April 24, 2014, 08:52:32 am »
Hi everyone,

I am trying to trace PWM waveform from an Inverter that drives a 3 phase Induction Motor using an oscilloscope. The PWM waveform had been obtained but I am not sure if  the method used is correct.

Here are specs of equipments used :

Motor : 3 phase, 0.75kW Induction Motor (MarelliMotori, MAA 80 MB 40)
Inverter : Emerson SKA 1200075
Oscilloscope : Agilent DSO 3062 (2-channel)
Differential Probe : Picotech TA 041 active differential oscilloscope probes

My goal  :

To obtain the PWM waveform on oscilloscope and subsequently convert the signal to frequency domain using the FFT function on the oscilloscope. I am trying to correlate the noise and vibration of the motor with harmonics content in the PWM.

Here is what I have done:

1.   Connect the differential probe lead to the motor terminals. In this case, they are connected to terminal U and W. (Phase-to-Phase)
2.   The BNC end of the differential probe was connected to the oscilloscope’s Channel 1.

setup2 by vigren_6, on Flickr[/img]
Experimental Setup

uvw2 by vigren_6, on Flickr[/img]
Probe leads connected to motor terminals

3.   Display screen was split into two : Top- PWM waveform & Bottom-FFT result.
PWM3 by vigren_6, on Flickr[/img]
Screenshots of oscilloscope display

4.   The motor was run to 1000 rpm and the result is shown below. The result indicates that the waveform is not a PWM but instead a SPWM (Sinusoidal PWM).  Pulse width which starts thin becomes wide in the middle before returning back to thin in the end.
5.   FFT has indicated harmonics of  3, 6 and 9 kHz was obtained. Cursor A – 3 kHz and Cursor B – 6 kHz.



My question is :
a.   Is this the correct method to trace SPWM going into the motor? (phase-to-phase)
b.   Or should it be referenced to ground?
c.   Can the measurement be done without using the differential probe? By using 2 channels of the oscilloscope & measure phase-to-phase?

Thank you.
« Last Edit: April 25, 2014, 05:44:50 am by Vigren »
 

Offline HalfSpace

  • Regular Contributor
  • *
  • Posts: 111
  • Country: au
Re: Tracing PWM Waveform from an Inverter using an Oscilloscope
« Reply #1 on: April 26, 2014, 01:58:38 am »
Hi Vigren,

Using 2 channels of an oscilloscope referenced to ground in subtract mode will not have the same CMRR as differential probe. If the motor is “delta” wired then phase to phase measurement is correct.
 
HalfSpace  :)
"He who asks a question is a fool for five minutes; he who does not ask a question remains a fool forever."
 

Offline qno

  • Frequent Contributor
  • **
  • Posts: 422
  • Country: nl
Re: Tracing PWM Waveform from an Inverter using an Oscilloscope
« Reply #2 on: April 26, 2014, 06:55:47 am »
Hi Vigren,

Try to analyse the source of the vibration.
It could be imbalance in the motor or the load.
It is usually not the inverter.

To check you inverter:
The vibration usually is generated by the torque and torque is dependant on the current. So maybe measuring current is a better way.
Depending on the algorithm of the inverter (standard pattern or vector control) it will generate a voltage that will result in currents in the windings that will make a rotating magnetic field.
Specially at low frequencies this is difficult because the impedance of the windings at DC is much lower than at 50Hz. You have to PWM the current to prevent blowing up your inverter. This will make your current not perfectly a sinus wave.
Also the inhertia of the load does not help.

Look on the web. Schneider/Emerson have loads of info in the website.
In the industry it is known as the Torque ripple.
 
Why spend money I don't have on things I don't need to impress people I don't like?
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf