NI 9227 and NI 9229 Suggest Wiring Practices

[Williamx]

I'm going to be using these two ADC's for current and voltage measurements respectively. The operations manual doesn't discuss using either twisted pair wiring or shielded pair wiring. Should this be obvious or is input impedance sufficiently high to not worry about noise being picked up on the inputs or crosstalk between channels due to the wiring?

[AnalogKid2DigitalMan]

Well, it depends on the noise nature of your environment, if there are any AC mains wiring ran near your signal cabling, frequency content and risetimes of your signals, length of your signal cabling, and other factors unique to your situation.

I typically use shielded cable with single or multiple twisted pairs contained within and earth-ground the cable shield only at one end.

-AK2DM

[Williamx]

Thanks AnalogKid2Digi… for your reply. That was certainly my assertion to other people questioning my concern over adding noise to the signals being measured. To add clarity to my question which you inspired, it's not so much a concern over the AC that will be in close proximity (although that is also an issue). I'm afraid of picking up digital noise because the wiring is routed near open chassis computer motherboards and the measurement points go directly to motherboard connectors. The fact that the measurement lead lengths will be 2 to 3 feet each (equal lengths to match lines losses) I consider this to be an environment rich with noise and the intended measurement wiring a pretty good antenna. Do you agree that this is asking for questionable accuracy of the signals being acquired? I can attempt to compensate for the noise by taking static measurements without the signal the interest and attempt to measure the noise exclusively. Then take on the arduous task of subtracting the noise from the signal of interest (i.e. the signal plus its noise), but that seems like way too much trouble and inaccurate when I could just minimize the noise from the start. Your thoughts?

[GovBob]

National Instruments has a great document concerning mitigating noise problems.

Hope this helps

[Williamx]

GovBob, thanks for your reply.

I have reviewed the article you referenced and although it answered some questions, it also raised some. My signal source is floating (but has a grounded shield) and my Analog Inputs are isolated from ground. My measurements are to be Differential, but I have no Analog ground to tie bias resistors to which means they can't be used and I hope means that they are not needed. Steve Harrington of Flometrics mentions inserting a 0.1uf cap in between the Analog Inputs and Analog Ground on the measurement system. This too, seems not possible. The only thing that seems left for me is shielded measurement and signal cables. Any thoughts?