Author Topic: Oscilloscope (and Probes) for High(?) Voltage Measurements  (Read 789 times)

0 Members and 1 Guest are viewing this topic.

Offline reborn1212

  • Contributor
  • Posts: 6
  • Country: my
Hey guys, first time posting here. Sorry if any forum rules were not followed.

I need to display the output voltage waveform of an industrial AC drive which operates on 3-phase 415V supply.
The DC bus voltage of the drive is about 600V (415 x 1.414) and the maximum switching frequency is 16kHz.

As for the output waveform shape, I have only a very vague idea that it is a PWM waveform obtained by switching the output terminals ON and OFF to the DC bus.

I had done some research and concluded on obtaining an oscilloscope and a differential probe (links datasheets attached) to measure the phase-to-phase output of the drive:
Oscilloscope: GW Instek GDS-1052-U (http://www.farnell.com/datasheets/3168229.pdf)
Probes: TESTEC SI 9002 (http://www.farnell.com/datasheets/10524.pdf)

Can I get the output waveform with these equipment? Is it safe to do so?
 

Offline rs20

  • Super Contributor
  • ***
  • Posts: 2247
  • Country: au
Re: Oscilloscope (and Probes) for High(?) Voltage Measurements
« Reply #1 on: May 12, 2021, 02:46:48 pm »
FWIW, that looks like a reasonable combination to me.

(Of course, in order to be safe, you have to use the probes correctly!)

Here's the manual for the SI 9002 probes.
 
The following users thanked this post: reborn1212

Offline threephase

  • Regular Contributor
  • *
  • Posts: 186
  • Country: gb
Re: Oscilloscope (and Probes) for High(?) Voltage Measurements
« Reply #2 on: May 13, 2021, 06:37:51 am »
Yes, unless the drive has an output filter on it, you will see a PWM signal as appose to a sine wave with your scope on the output. Your selection of equipment looks fine to me. As per the other comment, the equipment needs to be used in a safe manner to make these measurements.

My setup for measuring this is actually from Micsig using their TO1074 70 MHz tablet scope and their DP10013 HV differentia probe.
 
The following users thanked this post: reborn1212

Offline reborn1212

  • Contributor
  • Posts: 6
  • Country: my
Re: Oscilloscope (and Probes) for High(?) Voltage Measurements
« Reply #3 on: May 13, 2021, 09:00:22 am »
FWIW, that looks like a reasonable combination to me.

(Of course, in order to be safe, you have to use the probes correctly!)

Here's the manual for the SI 9002 probes.

On this matter, was the differential probe with its ratings an overkill? It did cost more than the scope.  :o
I did see lower rated differential probes for cheaper, but most of them were 500Vrms and was below the drive DC bus voltage.

Yes, unless the drive has an output filter on it, you will see a PWM signal as appose to a sine wave with your scope on the output. Your selection of equipment looks fine to me. As per the other comment, the equipment needs to be used in a safe manner to make these measurements.

My setup for measuring this is actually from Micsig using their TO1074 70 MHz tablet scope and their DP10013 HV differentia probe.

With the output being a PWM signal, it would have a fundamental frequency equal to that of the switching frequency, right?
I think the scope's bandwidth is sufficient to capture most of the significant harmonics.

How would we go about choosing an oscilloscope if I were to probe higher frequency signals?
 

Offline tggzzz

  • Super Contributor
  • ***
  • Posts: 13798
  • Country: gb
    • Having fun doing more, with less
Re: Oscilloscope (and Probes) for High(?) Voltage Measurements
« Reply #4 on: May 13, 2021, 09:42:14 am »
With the output being a PWM signal, it would have a fundamental frequency equal to that of the switching frequency, right?

Not necessarily. You have to be careful what you mean by the fundamental frequency; it certainly wouldn't be 50/60/400Hz
 
Quote
I think the scope's bandwidth is sufficient to capture most of the significant harmonics.

How would we go about choosing an oscilloscope if I were to probe higher frequency signals?

Measure the waveform's risetime, tr. The rule of thumb is that the scope+probe should have a bandwidth greater that 0.35/tr.

The probe's bandwidth (including rejection of common mode components) is more likely to be the limiting factor.
There are lies, damned lies, statistics - and ADC/DAC specs.
Glider pilot's aphorism: "there is no substitute for span". Retort: "There is a substitute: skill+imagination. But you can buy span".
Having fun doing more, with less
 

Offline penfold

  • Frequent Contributor
  • **
  • Posts: 488
  • Country: gb
Re: Oscilloscope (and Probes) for High(?) Voltage Measurements
« Reply #5 on: May 13, 2021, 11:14:20 am »
On this matter, was the differential probe with its ratings an overkill? It did cost more than the scope.  :o
I did see lower rated differential probes for cheaper, but most of them were 500Vrms and was below the drive DC bus voltage.

I think you got the rating correct, I would generally use a probe rated (for voltage) at least twice the DC bus voltage. Higher voltage spikes are more likely when there's longer cable runs and faster rise times, but difficult to predict. Rated at 1400V you'll have a more useful probe that you won't risk damaging so much if you then go on to investigate the voltage at the load side of the cable or during short circuit transients etc.
 
The following users thanked this post: reborn1212

Offline reborn1212

  • Contributor
  • Posts: 6
  • Country: my
Re: Oscilloscope (and Probes) for High(?) Voltage Measurements
« Reply #6 on: May 14, 2021, 04:33:06 pm »

Not necessarily. You have to be careful what you mean by the fundamental frequency; it certainly wouldn't be 50/60/400Hz

By fundamental frequency I meant the lowest frequency in the output waveform; maximum of 16kHz (equal to the maximum switching frequency of the drive)


I think you got the rating correct, I would generally use a probe rated (for voltage) at least twice the DC bus voltage. Higher voltage spikes are more likely when there's longer cable runs and faster rise times, but difficult to predict. Rated at 1400V you'll have a more useful probe that you won't risk damaging so much if you then go on to investigate the voltage at the load side of the cable or during short circuit transients etc.

Glad to know, seems like my proposed setup will be fine for the job; I'll update my results here after I get the equipment.
 

Offline threephase

  • Regular Contributor
  • *
  • Posts: 186
  • Country: gb
Re: Oscilloscope (and Probes) for High(?) Voltage Measurements
« Reply #7 on: May 14, 2021, 08:46:32 pm »
In terms of power analysis of the waveform, the fundamental frequency we use will be the line frequency of the output, after all that is what we use to calculate what the motor speed will be for a given output. The PWM signal is designed to mimic that fundamental frequency, that is smoothed by the inductive nature of the motor and of course any output filtration present.

For power analysis of inverter outputs, I have gone up to the 375th harmonic of the line frequency to obtain all the power components of the waveform. That takes you up to around 19 kHz on a 50 Hz line frequency. If you want to look at the harmonics of particular PWM pulse, then you are obviously going to need much more than that, I cannot imagine you going over a 50 MHz bandwidth though. For harmonics you are looking more at a spectrum analyser / FFT analysis of the waveform, so you would need to make sure your scope has that capability.

Unfortunately, differential probes are expensive, but they allow these kind of measurements to be made safely, so in my opinion are worth the investment. The probe I have is rated for 1.3kV differential and 1000V to earth. I personally wouldn't want to go lower than this, so I think your selection is fine.
 

Offline reborn1212

  • Contributor
  • Posts: 6
  • Country: my
Re: Oscilloscope (and Probes) for High(?) Voltage Measurements
« Reply #8 on: May 29, 2021, 03:13:39 pm »
Got the inverter output waveforms; refer attached.

The no-load output waveforms seemed fine but when the inverter was driving a motor there were some weird spikes when switching to negative DC bus voltage.

Note: The oscilloscope probe settings were chosen from a preset of x1, x10 and x100 but the probe was set to attenuation ratio 1:200 so the actual voltage is actually double of what is displayed in these images.
 

Offline penfold

  • Frequent Contributor
  • **
  • Posts: 488
  • Country: gb
Re: Oscilloscope (and Probes) for High(?) Voltage Measurements
« Reply #9 on: May 29, 2021, 04:20:00 pm »
It's not unusual to have some ringing and overshoots. But its worth investigating nonetheless.

Is the motor itself loaded or is it just free-spinning?

Is it consistently repetitive, i.e. happens on every cycle in relatively the same position or does it vary significantly?
Are you able to take readings from the other phases? It is also worth measuring the waveform on the motor end of the cable (or inverter side if already at the motor side).
 

Offline reborn1212

  • Contributor
  • Posts: 6
  • Country: my
Re: Oscilloscope (and Probes) for High(?) Voltage Measurements
« Reply #10 on: May 29, 2021, 05:36:45 pm »
It's not unusual to have some ringing and overshoots. But its worth investigating nonetheless.

Is the motor itself loaded or is it just free-spinning?

Is it consistently repetitive, i.e. happens on every cycle in relatively the same position or does it vary significantly?
Are you able to take readings from the other phases? It is also worth measuring the waveform on the motor end of the cable (or inverter side if already at the motor side).

I forgot to mention; the motor would be spinning normally but vibrate violently (the motor body itself would move) at seeming random intervals when connected to this particular inverter.
The voltage spikes and irregularities in the output waveforms were only present during these intermittent episodes. Otherwise, they look the same as in the first picture (no-load output waveform).

The motor itself is not loaded as the output waveforms were captured on the oscilloscope.
 

Offline penfold

  • Frequent Contributor
  • **
  • Posts: 488
  • Country: gb
Re: Oscilloscope (and Probes) for High(?) Voltage Measurements
« Reply #11 on: May 29, 2021, 05:47:33 pm »
...
The voltage spikes and irregularities in the output waveforms were only present during these intermittent episodes. Otherwise, they look the same as in the first picture (no-load output waveform).

hmmm... is the motor in good health? bearings ok? no rattles or squeeks?
 

Offline reborn1212

  • Contributor
  • Posts: 6
  • Country: my
Re: Oscilloscope (and Probes) for High(?) Voltage Measurements
« Reply #12 on: May 29, 2021, 05:56:53 pm »
hmmm... is the motor in good health? bearings ok? no rattles or squeeks?

I don't think I'm qualified to assess the motor mechanically; but it's the test motor used at the office and the vibrational episodes don't occur with the other inverters.
 

Offline threephase

  • Regular Contributor
  • *
  • Posts: 186
  • Country: gb
Re: Oscilloscope (and Probes) for High(?) Voltage Measurements
« Reply #13 on: June 11, 2021, 11:29:50 am »
Is the motor vibration occurring at certain VSD settings?

Sometimes as a motor is accelerated you will hit a resonance issue of either mechanical components of the motor or physical fixings at a specific shaft speed and you will get momentary vibrations that will drop as the speed increases through the resonance bandwidth. On long power trains you can hit a number of these critical speed points.

A similar effect may also occur if the harmonics from the drive hit a certain level to excite the natural resonance of the motor frame, but again I would expect it to occur at specific speed points.

If it is randomly occurring whilst the speed is constant, then you may have components starting to break down. Is the winding rated for operation on a VSD. Does the motor pass electrical tests, insulation resistance, winding resistance balance and winding inductance balance?
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf