Help with reducing RF noise for small signals that pass through PXIe-6529 + NI TB-2635

[JKSH]

Hello,

I wasn't sure if I should post here or in the Signal Conditioning subforum.

Anyway, I need to pass small, high-frequency signals (about 100 nA - 1 μA, 1 MHz) through a switch, and I'm currently trying to use a PXIe-6529 + NI TB-2635. The signal would then be amplified using a transimpedance amplifier, at about 10⁶ - 10⁷ V/A. I'm using coaxial cables to carry my signals.

Unfortunately, my coaxial cables seem to be picking up quite a bit of RF noise through the PXIe-6529 + NI TB-2635. The noise is very small in itself, but becomes large after passing through my amplifier. I've observed (through an oscilloscope at the amplifier output) that the noise comes on when I power up the PXI chassis, but disappears when I turn of the PXI chassis. Also, if I enable the amplifier's builti-in 1 MHz low-pass filter, the noise amplitude is reduced by about 1/4. However, I can't use this filter as it would attenuate my signal too.

I don't know much about with best practices when it comes to measuring small signals at high frequencies. I'd apprecieate some tips on how to started on shielding or noise suppression. Also, are there any PXI matrix switches that are more resistant against RF interference?

Thanks!

[Henrik_Volkers]

Some general thougths:

Cables to the TIA as short as possible.

Currents need to flow in a loop, so RF picked up in the ground return end up in your signal. A good ground topology avoiding loops is essential.

A ferrite core that forces equal currents in your shielded cable can help.

Shielding and finding the source of the RF....

A good source of information to measure really tiny values is the 'low level measurements handbook' from keithley...

Sorry no link to it.... these %$§'%  search engines don't provide direct links anymore 

nA at 1MHz ...  I would give a good shielded teflon isolated twisted paire a try.   A network cable (100 Ohm impedance, not 50!)  as a first shot???

I have seen delicate signal cables shielded in solid iron tubes 😄

[jozeywhales]

Hello,

I have a few suggestions.  You can try amplifying the signal before sending it through the coaxial line to the PXIe-6529.  Another option is to configure the digital filter on the PXIe-6529.  There are detailed notes on how to configure the digital filter in the PXIe-6529 users manual.  A third option is to purchase or construct a bandpass filter that will allow your digital signal through and reject all other noise.

USER GUIDE AND SPECIFICATIONS NI 6528/6529

http://www.ni.com/pdf/manuals/372124d.pdf

I hope this helps.

[JKSH]

Aargh, what a typo... I meant PXIe-2529 (matrix switch), not 6529!  I'm measuring analogue signals.

Is it possible to modify my post title?

[JKSH]

Thanks for your suggestions, Henrik and Joe. I'll give some of them a try.

Keithley book: I found a link, if you're interested 🙂 http://www.tek.com/sites/tek.com/files/media/document/resources/Keithley-LowLevelHandbook_7Ed.pdf