Noise problems with differential measurement

[SpeerResi]

Hey everyone.

I'm using a NI USB 6211, LABView 2011 and Win7.

I am trying to measure a voltage across a resistor using differential mode. But unfortunately im getting a lot of noise (about +-25%).

The voltage source I'm using is floating and can go up to 600V. With this as my power source I'm basically trying to heat a metal plate.

A voltage divider in parallel is used to reduce the voltage by a hundredth (1 MOhm and 10 kOhm).

Two leads attached to the lower resistance then go straight into the DAQ's analog input 0+8.

So if I'm feeding 600V to the plate, the DAQ should get about 6V ... and this is what I measure with my voltmeter attached to the pins of the DAQ's Input.

I also tried those BIAS resistors and connected the + and - leads to the analog ground, as a resistance i used 10k, 100k and 1 MOhm and 2  MOhm and

still receive a pretty bad signal.

Two sketches of the wireing as well as a signal (100V, supposed to measure 1V, no BIAS) are attached to this post.

Regards

EDIT: I forgot to write that I even tried another NI DAQ and still get this noise problem.

Also, i measured the signal of the voltage source by using an oscilloscope and I see this noise. But isn't the differential mode supposed to 

reduce those noises to a minimum?

[johnsold]

My suspicion is that the source or load is not completely floating.  Even if the coupling to ground is capacitive or inductive, substantial current could  flow through the USB ground back to the computer ground and through the computer power supply to a real ground.  The voltage drop across those ground conductors may well be the source of your noise.

Try using your oscilloscope and set the triggering to Line. Set the horizontal scale to 2 or 5 ms/division. If you see a "stable" periodic signal, It i slikely power line related and, techically interference, not noise.  Your chart image does not identify the time scale, but the fact that you are using a chart on a thermal system leads me to guess that you are using a slow sample rate.  This would result in any power line frequency interference being aliased and possibly appearing as noise.

The sketch 2 wiring should be the appropriate method.

Lynn

[SpeerResi]

Hey Linn,

thank you for your response.

Im going to try this and reply later on.

Regards

[SpeerResi · Accepted solution]

Hey everyone,

just figured out that the resistors of the voltage divider are way to high and that they were the main reason for the noise

problem.

Before I used 1 MOhm and 10 kOhm, now I'm using 100 kOhm and 100 Ohm and with a median filter on it it works just fine!

But still, if I use those BIAS resistor they don't change anything.

Regards