how to filter ripple and line pickup from a DC signal?

[sukhiray]

I have to filter a DC signal coming from a photo detector. It contains ripples and line pickup.


Message Edited by sukhiray on 08-04-2008 12:22 AM

[Hillman]

Hi,

I found this knowledge base article which it titled: How to Eliminate AC Noise When Measuring DC Signals. Which would suggest in labview your best approach would be to use the signal processing (lowpass filter) and/or maths pallet (integration) to remove your AC noise.

http://digital.ni.com/public.nsf/allkb/552D26DA22F436368625729200409E6E?OpenDocument

I used photovalics recently, which also have issues with unwanted ac characteristics and we found the use of an AC-DC converter, which used a microprocessor to alter the signal conditioning circuit to maximize its performance was the best solution.Depending what hardware your using some NI hardware have built in filtering functions.

To help others answer/talk about this post, it would help if you could give us more details on your set up and maybe ideas you've tried. Thanks.

Message Edited by Hillman on 08-04-2008 04:10 AM

Message Edited by Hillman on 08-04-2008 04:11 AM

[johnsold]

It is almost always better to prevent the signal from getting to the software than to try to eliminate the interference afterwards. Do you know how it is getting into your signal? Is it optical, from the power supply to the photodetector, or induced into the cables?

Fluorescent lights are horrible generators of power line frequency and harmonics optical signals. If you can prevent stray light from getting to your detector, that may help. Fluorescent lights generate very little energy at the red and infrared end of the spectrum, so a long wave optical filter can sometimes help.

Synchronizing the sampling with the power line is also helpful. If you sampled at exactly the zero-crossing of the power line, the power line interference on each sample is exactly zero.

As Hillman said, tell us more about what you are doing.

Lynn

[Pie56694]

Hi Sukhiray,

I agree with Lynn on this one.  Here are some basic things to try.

Use differential measurement mode with a DC bias resistor on the negative input terminal.
Use the shortest wires possible.
Use twisted pair wiring with a grounded shield from the signal source to the DAQ.
Amplify the signal at the source to improve Signal to Noise Ratio (SNR) at the DAQ.
Use input signal conditioning like an SCC-LP01 module.
Avoid high-noise sources to power the circuit (don't use bench-top power supplies wherever possible, use batteries).
Avoid high-noise environments, or move the measurement station away from high-current equipment.
After the measurement is in software, you can filter, average, and statistically reject the value to work-around the noise.

Field Wiring and Noise Considerations for Analog Signals

I hope this helps.