LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Signal Amplification

I am using two accelerometers in my experiment to measure the time delay. The signal which is farther away from the source gets decayed due to damping. How can I amplify my later signal using LABVIEW? Thank you.

 

0 Kudos
Message 1 of 6
(1,323 Views)

Hi,

What is basically happening is that your signal that is away from the source is losing in Signal to Noise ratio. If you try to amplify your signal, you will also amplify the noise which will not be helpful for you. However, what you can actually do is gather more signal to noise ratio by averaging multiple samples of your accelerometer data. This way, you will be able to reduce the noise level and then increase your SNR. However , the faster the delay rate is, the less samples you can use to average your signal .

Best regards,

Mourad FAKHFAKH

Certified LabVIEW Architect

Message 2 of 6
(1,303 Views)

Accelerometers measure acceleration, not time, so I assume you are just using them to e.g. measure the propagation of a shockwave or similar. In this case, the amplitude is relatively unimportant. LabVIEW is not an "amplifier", so the multiply function might work, depending on the datatype of your signal. Maybe you could just normalize the signals. Maybe you need to filter.

 

We really need a little bit more information about the experiment. What are the distances involved? What are typical time delays (seconds? nanoseconds?) What kind of signals do you get? How do you process the signals? What kind of hardware is involved? Maybe a solution would be in hardware (amplify before transmission, using better cables, using fiber optic instead of wires, etc. Process locally and transmit via wifi)

Message 3 of 6
(1,292 Views)

@ Mourad FAKHFAKH

 Thank you so much for suggesting this. I also found this in signal processing notes. One thing is that my input signal is not fixed in the time domain. So averaging in the time domain will not work, it will just mix up everything. Perhaps can I stack in the frequency domain and then do inverse Fourier transform. Will that work? 
Thank you for your reply.

0 Kudos
Message 4 of 6
(1,190 Views)

@altenbach

 

Thank you for your reply. Yes, the amplitude will not matter as when I am measuring only the time delay. The distance between sensors is about a meter. And usual time delay is in milliseconds to microseconds depending on the type of medium. The thing is that my first signal is near the source and later is around a meter away and hence the later signal is too weak as compared to the first. I can not put my source away from the first sensor. So I thought what shall I do to improve the second signal.

I will need the amplitude as well while obtaining a velocity signal from the accelerometer signal by integration. I do not have much experience with filters . I think it distorts the phase values of the signal so I am not currently using it. But yeah I have to use it while performing integration as otherwise, it's difficult to integrate. Kindly help with what type of filters shall I use to avoid distortion of phase values. While integrating I am using the 6th Order Butterworth filter (Bandpass) depending on my frequency content to remove noise. 

 

Thank you so much. 

0 Kudos
Message 5 of 6
(1,186 Views)

In order to average your signal in the frequency domain, you will have to group a burst of samples from your signal and assume that there is a periodicity in your signal. If your signal is variant in time as you mentioned (no sign of periodicity), this strategy won't resolve your issue. What I suggested in my first post is more of a signal smoothing (A moving average or even a Savitzky-Golay smoothing) if you can assume that the information you are looking for in your signal is invariant over a number of samples that only you can define knowing your setup and specs. As altenbach suggested, either you will have to increase your amplitude using a preamplifier, or try to lose as less as possible from your signal power by using wires that dissipate less signal power. 

If you provide additional details about your setup, you may have more precise answers to your problem.

Best regards, 

Mourad FAKHFAKH

Certified LabVIEW Architect

 

Message 6 of 6
(1,145 Views)