voltage acquisition problems

[jackie_brown]

Hello, I am having some problems when it comes to read a voltage input from my power supply.

I am using a NI 9215 with BNC connector. My power supply receive an input (usually a sine wave) that can vary from 0 up to 5 volts.

When the input is 5 volts the power supply is able to deliver 10 000 V, then I am using a Fluke 80K-40 H.V. probe that scales the voltage

as 1:1000. Consequently, using an input of 5 volt I obtain 10 000V and I am able to read 10V on my multimeter.

The same happens also using Labview and the DAQ for the signal acquisition but only if the voltage frequency is extrememly LOW like 0,5-1 Hz.

If I increase the frequency, even if it is 10 Hz, the signal that I read has a different amplitude like 9 V and the amplitude decreases as I increase the frequency.

Since this happens even at very low freequency I dont thik is a problem of bandwith of the NI 9215 card.

Moreover the output is supposed to be symmetric: I mean if the peak voltage is 7 000 V, I will have a sine wave travelling from + to - 7 000V

so I should read after the HV probe +/- 7 volt, the fact is that increasing the frequency doesn't only affect the amplitude but causes also the loss of symmetry

(for example I read a sine wave between +7 and - 4 V).

I hope someone can help me.

Thanks in advance.

[natasftw]

What's your sample rate?

[johnsold]

The probe probably has a very high resistance. That and the input capacitance of the DAQ device (plus the cable capacitance) forms a low pass filter. A resistance of 100 megohms and a capacitance of 100 picofarads could produce frequency effects similar to what you are reporting.

Lynn

[Henrik_Volkers]

Well have a look at the Spec:

The probe is build to match with the (complex) input impedance of a multimeter with 10M Ohm (and a not specified capacity) . 

Your 9215 with BNC has a input impedance of 200k Ohm (according to spec)  and no specified input capacity .....   so RTFM help a little 😉

The offset migth be due to the (max) 10nA input bias current... small currents on high resistors make some volt, 😉

Hunt for a TEK 6015A 1000:1  probe and a device with a scope compatible input  (1M || ~40pF ) and you can go up to 25 MHz 😉  (well, it will load your source with 100M instead of 1G)

or grab a OP and build a buffer with 10M input impedance and a low output impedance. (Nice as a protection for your DAQ too, just in case 😉 )

[jackie_brown]

Hello, thank you for your advices.

Natasftw, I used different sample rate: if the frequency was 60 Hz I used 1200Hz up to 10KHz but the result doesnt change.

Regarding the probe: I dont think this is the problem because as I said the probe is connected to a multimeter and on this multimeter I read

the correct value, the problem is that my DAQ seems to acquire a different signal.

Henrik_Volkers, your explanation seems quite detailed but I'm not really into this yet, so I would like to rephrase your words:

as I understood you said the problem can be a current, but why it happen at frequency beyond very low??; also you said the problem is probably the probe...

how the new probe will be different?? and also what kind of device are you referring to when you said 

"Hunt for a TEK 6015A 1000:1  probe and a DEVICE with a scope compatible input  (1M || ~40pF )" ???

Thank you.

[Henrik_Volkers]

The effects of source impedance and input impedance and how they affect measurements is a quite basic one. Should be a tutorial here ... or in the web... or in a book on EE ...  see it all as (complex) voltage divider 😉

The 9215 Spec isn't that clear about the input impedance... wether it is 1GOhm or 200 Ohm .....   maybe a 10MOhm resistor between AI+ and AI- can help.

Input bias current and common voltage range, ground loops and .... all made to make 'simple' measurements fail.

Some basic tests you can do: 

Measure  +7 kV , -7kV and 0V (GND)   all DC with your multimeter and with your DAQ (same samplerate and setup as for AC, shouldn't matter , but ... )

Your multimeter and your DAQ will behave differently since they have different input impedances. If you add some more cable between probe and DAQ it will even be worse for the AC measurement.

How are all your ground connections wired?    Have you wired any other 9215 inputs? Have a look at the manual /spec about the internal circuit :

All inputs must remain within 10.2 V of the average AI–inputs.