Noise Feedback, how to measure?

[qu1nn]

Is there a standard procedure for measuring noise feedback of a sensor?

First off, I am assuming that noise feedback would be the noise from the sensor that gets injected back into the power supply.

If I had to guess you would want to place something inbetween the power supply and the sensor (R/L network for isolation ??) then attach an oscilloscope between the R/L network and the sensor.

What values of R/L (if I guessed correctly ..for R/L network) would I pick?
Suggetions?

Thanks for your time
Qu1nn

 

[BrianG]

Hi, more details are necessary:
1)what kind of sensor? and
2)what kind of power supply?
3)general application?

 

[qu1nn]

Pressure transducer 0-5Vdc output range, utilizing a std wheatstone bridge configuration . Typically the specification states that the transducer can be used on an unregulated powersupply between 20 and 40Vdc, typically 30Vdc is used.

 

[IRstuff]

I'm confused. I dno't see that your sensor would generate much noise at all compared with the rest of the circuitry. If anything, I'd be concerned about noise feedthrough from the unregulated supply getting into the bridge. TTFN

 

[qu1nn]

The specification really means that one could run the power supply over a wide range of excitation voltages. Regulation is performed by an isolated dc-dc converter, which will oscillate, and regulate as long as it falls within min/max excitation voltages and can provide a minimum current.

Typically the unit is ran at a known voltage. For example 30Vdc, and lets also say that it is regulated (+/- 10mV noise)

My question is; how can you isolate the power supply from the powered unit so that I can measure noise generated back into the power source.


Thanks
Qu1nn

 

[melone]

THE SENSOR DOESN'T GENERATE ANY REAL NOISE!!!!!! If you have noise on your power, it's from something else! Why are you asking sure a strange question? Did someone ask you to measure something?

 

[Lewish]

Your DC-DC converter is going to dominate your noise.

Pay attention to what Melone is saying.

 

[qu1nn]

It is part of a performance specification that a customer is asking about.
Qu1nn

 

[IRstuff]

Are we talking MIL-STD-461 conducted emission testing or something more benign?

If the latter, you can only realistically do this by measuring the noise of the power supply and sensor with the sensor turned on and turned off. The extra noise, if any, would be attributable to the sensor. Any other approach short of building an exact impedance surrogate for the power supply in question will give erroneous results. The reason is that any noise injected back to the power supply must manifest itself across the output series impedance of the power supply itself. TTFN