Input signal conditioning - voltage dividing or biasing

[tmoritz]

I have a NI USB-6218 BNC DAQ unit.  Input voltage range is +/-10V.  I want to collect data from six Bently-Nevada proximity probes with a linear output range of -1 to -17V.  What is the best way to condition this signal?  Bias the common ground or set up a voltage divider?  Are there better ways?

Thanks in advance.

[santo_13]

Biasing the common ground works if the proximity sensor has a floating output. Voltage divider would be the easiest option provided the output impedance of the probes are good and your choice of resistances for the divider doesn't load the signal.

[tmoritz]

I'm going to try using some in-line attenuators.  I have some 20dB and 6dB.  Hopefully the 6dB will get me in range.  If not I go to the 20dB.  We'll see how this goes.

[santo_13]

Do note that these attenuators load the signal from the sensor, ensure that this loading is not significant or you will get inaccurate measurements.

[Henrik_Volkers]

insert a 9V batterie?

OK, while this would work, the voltage stability of such a source migth not be the best 😉  adding a 7-8V  low drop voltage stabilizer would do the job.

So one stable voltage source and a OP amp per channel to add the voltage ( and another for an input buffer) 

[tmoritz]

Mr. Volkers - Thank you for your response.  I am now planning on building a signal conditioning box using op amps to amplify (or attenuate) and bias these vibration signals.  

[tmoritz]

Santo_13 - thanks for the heads up.  Turned out that attenuator's low pass frequency was above the synchronous speed of my equipment.  I'm back to building an aliasing circuit using some op amps to minimize any extra burden on the transducer circuit.