amplifier output voltage changes during high speed acquisition

[BigEv]

I have a PCI-6030E DAQ board where the AI signals are connected to the output of 16 individual instrumentation amplifiers (AD620).  The incoming signal to the amp is always -6.87 volts.  The output when the 6030 is not acquiring data is -6.87 volts.  When the 6030E is acquiring data across all channels (16) at 2,000 hz the measured voltage is -6.87 volts.  When the sampling rate is increased to 5,000 hz across all channels, the measured voltage is -6.26 volts.  The 6030E is configured in RSE, with the input range set to -10 volts minimum, and 10 volts maximum.  The AI ground signal is connected to the amplifier output reference pin, so essentially I have a differential input to single ended output converter if everything is working properly...  Attached is a schematic of the interface board.  Question is, why would the measured voltage change at higher sampling frequencies?  Note: at 5,000 hz, the AI clock rate is at 80,000 hz, this is below the maximum AI clock rate of 100,000 hz...

[Dustin_D]

Hi BigEv,

I wanted to clarifiy a couple things with your post.

1. Are you connecting only one signal at a time while performing your measurment?

2. Do you exhibit the same behavior if you input -6.87 V to each instrumentation amp?

3. Do other channels exhibit unexpected behavior if you are providing input to just one instrumentation amp at a time?

Thanks,

[BigEv]

Hi Dustin,

Thanks for the reply.

At the moment I have 5 signals connected.  These 5 channels also exhibit some amount of change at high ACQ rates, but significantly less than the channel connected to the -6.87 V signal (probably because the signals are less than ±1 volt to begin with).  Although, the other channels have some amount of change, I haven't had the chance to quantify the amount yet.  My big concern has been the -6.87 V channel, since I am implementing a control system about the -6.87 V transducer signal.

It is possible that the inputs for the unconnected channels are saturated.  I will try putting a 10k resistor across the input to the ad620 op-amps that are not being used.  Do you think that a saturated channel on the input to the 6030 board could be causing the odd behavior at the higher sampling rates?

Kind Regards,

Ev Hall

[Dustin_D]

Hi BigEv,

I am slightly concerned you may be suffering from some sort of ghosting effect.  If that is the case you can try a few things: 

• Make sure all of your signals have some sort of defined reference, whether it be ground or another voltage.
• Check to see if this phenomena only takes place while reading the transducer or is it any voltage source?
• Try reading just a single channel without scanning through the rest of the channels(this will help determine if it is ghosting or something from the IA).
• Try removing your circuit and reading some voltages to see if you can replicate without your circuit in between.
• Characterize your circuit without the DAQ attached to make sure the gain is what you expect.
• Shorting differential pairs.

I know you may have done some of these things already I just wanted to add them for the sake of being complete.  Let me know the results!

Regards,