NI-9252 on NI-9147. Sampling issue.

[MaxJoseph]

Hi all,

I have a 9147 chassis (Ethernet RIO) with a 9252 (8 channel, 24 bit, 50 kS/s/ch) attached. I wrote a very simple FPGA program to readout from the 9252 and put the data into a FIFO which I then read into a host process, see below. 

When I do this readout on the host I get some odd results that I cannot make sense of and Google is not helping me resolve. I connect channel 0 to a 2.5 V source and I get 2.5 V reading on the 9252 ch0 for 50% of the time and what looks like data from another channel the other 50% of the time. I have changed sample rate and hardware filtering on the 9252 with no affect beyond changing samples and smoothing the transition between the two levels. The fact the filtering changes the output makes me suspect that this is a hardware configuration issue.

Here is the readout code

I was hoping to use the Scan Engine with this product, but I understand that this feature is not available to the 9252 on the 9147. So I have been using the FPGA target scope instead. Are there other considerations I should be aware of when using these hardware in conjunction?

Thanks,

Max

[ZYOng]

Is your input signal floating or grounded?

Are you connecting using RSE or DIFF mode?

See Table 1 of Field Wiring and Noise Considerations for Analog Signals for the recommended wiring.

[MaxJoseph]

Hi,

Great questions!

The signal is floating. I am using DIFF mode across AI0+ and AI0-, pins 1 and 20 on the NI-9252. Does this sound unreasonable.

I will take a look at the linked document and see if anything jumps out.

Thanks!

[MaxJoseph]

Hi again,

I have found that the problem was caused by the 9252 not making a differential measurement. I am trying to make a differential measurement of two wires by sampling AI0, but this gives me a lot of noise. When I connect one of the wires to ground I get a great measurement, but this is a single ended measurement, not differential.

I should not have to ground one of my wires as the differential measurement is supposed to cancel out common mode noise in the two wires.

Everything I read about the 9252 says that it should be giving a differential measurement, but this appears not to be the case. I cannot see anywhere in the configuration of the 9252 in the project that refers to differential measurement. I have not tried the other AI channels (assuming AI0 is defective) as this would require a recompile and I didn't have time for that yet.

Thanks,

Max