60 Hz noise on PCI-4461 AO

[jmoses]

Every time our PCI-4461 powers up, it puts out 60 Hz noise on AO.  The noise goes away when my software starts an AO task.  I guess the output powers up in a floating state.  I wish it would power up grounded, but it is what it is.  I can deal with that; but yesterday I found a more troublesome problem.  Occasionally, rarely, I don't know how to duplicate it but we've seen it twice, the card puts out 60 Hz noise even with an AO task running.  I can program it to send out a 100 Hz sine wave, and the signal I get looks like 60 Hz and 100 Hz superimposed!  To make it go away, I have to shut down the computer and restart.  Unfortunately, I don't know how to cause this problem to appear, so we can study its root cause.  I did notice that connecting the AO signal ground to the PC chassis with an aligator cable eliminates both problems.  I am thinking about soldering a jumper wire on the PCI-4461 to make this connection.  Would there be a downside to this modification, that I should be aware of?

Josh Moses

The Modal Shop

[Jennifer O]

Hi John,

There are two output configurations with the PCI-4461, differential and pseudo-differential.  In pseudo-differential mode, the output is grounded to the chassis via 50 Ohm termination (see the DSA help for the output connection diagram). Is it possible that you are seeing noise in one mode but not another?  Make sure you are measuring the signal at the 4461 to make sure the noise is internal and that it is not being induced in your cabling.

Tying the chassis to to signal ground would bypass the 50 Ohm resistor.  This would essentially create a single ended signal, however if your chassis has any possibility of being at a different ground potential from your external connection, this could cause a ground loop.  Also, impedance matching can be important to prevent signal reflection.  The 50 Ohm impedance is one of the standard BNC impedance levels.  Bypassing the 50 Ohm resistor by tying your channel to chassis ground could result in mismatched impedances and signal distortion due to reflection.  

To address the power-up noise, you are correct that there is not a configurable power-up state.  You could however use relays to disconnect the outputs until your task begins and a voltage is being driven.

Regards,

Jennifer O.

[jmoses]

Thanks for your thoughtful ideas, Jennifer.

Josh