Peaks in sinus wave from measurements made by Rogowski coil to NI 9215

[gibon]

Hi.

I have 3 x iFlex 2000) (Fluke) coils (batery supply) to measurement currents.Bandwidth (-3dB) 1Hz - 20 kHz, up to 20, 200, 2000A. This unit transform current to voltage with level up to 2V and this go to NI 9215. cDaq graunded by resistor 500 MOhm ( but without resistor was the same). So this go to cDaq and from cDaq to Notebook. Unit is very simple.

In fact I have 2 problems:

1.   one when I measured I saw DC component in sinus wave. This sinus wave was curent from frequencer to electric motor. I cannot understand why this DC component appear (this was wery big, for eqample DC component 100 A RMS, and overall RMS 180 A). I change coils (phase R to S, phase S to R

2. I have peaks in reading current, in all phases an in source without frequencer. Peaks are very short, (1,2 points of reading in data flow) and with value ... big, I thing up to 2 times amplitude of signal, in different probe freqency (20kHz, 10kHz, 5 kHz). 

can some one help me with this ?

[BalazsNagy]

Hi gibon!

Sorry for the late reply.

Are You using the BNC or the Screw terminal version of the 9215 module?

Are You using any shielding on the cables between Rogowski coil and the NI 9215?

Spkies could come from EMC distortion if You ar close to a motor, or any other high current device.

Could You post some screenshots from Your code? Block diagram and front panel, may be with the experienced spikes?

Thanks.

BR,

[gibon]

Hello, Thanks for answer.

This module have BNC connection. My rogowski coil look like this on picture, with connection to universal meter. There is adapter from this connection to BNC, what I got with coil. I don't use any other filter (Rogowski coil works with own batery). In datasheet to this coil there is no information about any filter. I use this coil with scopometr and I didn't see any peaks (similar sample rate but only 8 bit analog to digital converter and I think there is a filter.

I was created cable with resistors what I founded somewere to connect "floating source" - I mean resistor about 100k Ohm from hot to ground and 100 kOhm from BNC shield to ground. And was the same. 

cDaq I have connected to ground by resistor 500 kOhm.

[BalazsNagy]

Dear gibon!

What do You mean with "cDaq I have connected to ground by resistor 500 kOhm." ?

Could You make a simple drawing from the connections You made?

Since If You have an adapter for the coil, that makes it's output avaliable for BNC connection, You don't need any extra connections. As You can see on the picture below, the NI 9215 with BNC has a resistor that ensures the input voltage does not drift outside of the common-mode range.

Any way this device has a Successive approximation register (SAR) type of ADC, which can cause, that distorting signals in a freguency range highe than the Nyquist frequency (fsample/2) overlap into our sampled signal. This You can compensate with filtering. Of course this can be done in SW as well. (Normal box Scopes have low pass filters on the input)

Are You using LabVIEW to implement the data acquisition with the 9215? If Yes, You cen use he Filter express VI to implement a lowpass filter on Your acquired signal, so You can eliminate the spikes.

Description for the Filter VI: http://zone.ni.com/reference/en-XX/help/371361J-01/lvexpress/signal_filter/

Manual of 9215: http://www.ni.com/pdf/manuals/373779f.pdf

Please let me know, if my sugestions helped.

BR,

[gibon]

Dear BalazsNagy

Thanks for replay.

You wrote many information, so piece by piece I'm reading and looking for more information.

"cDaq I have connected to ground by resistor 500 kOhm." - on chassis cDaq 9188 there is on right side connection to ground. I grounded this but by 500 k Ohm resistor (my colleague electrionic engineer told me this manner to protect from big voltage on cDaq cards).

"As You can see on the picture below, the NI 9215 with BNC has a resistor that ensures the input voltage does not drift outside of the common-mode range."

Yes, its true, but I found other connection diagram (look in Attachments). There is diagram with 2 resistors - so I try it.

My coil have a wide range of current frequencis: 10Hz to 20 kHz (-1 dB). I thing a problem is becaurse type of ADC is SAR. After You wrote this (about SAR), I looked for more information. 

"Of course this can be done in SW as well. "- I don't know what is "SW" 

If I remember right I tried some time ago a digital filter and it not help me. 

I think the solution is : create elektronic filter.

When will by not busy time in work, I will try this.

In Attachments I add piece of current measured on electic motor source from frequencer. Current spikes are very big.

Thanks for Your information.