# High speed sampling

Discussion in 'General Electronics Discussion' started by saurabh17g, May 4, 2010.

1. ### saurabh17g

72
0
Mar 8, 2010
Hello,
I have to develop an application wherein I have to capture a waveform(sinewave) of frequency from 0Hz to 1 MHz and display it on PC. What are the possible ways to do so? I feel that , its not possible on a real time basis. So, signal should be stored and slowly transmitted to PC. One cycle(period) of waveform is enough to be shown.

2. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,451
2,809
Jan 21, 2010
Haven't we been here before?

If it's really a sine wave, then all you need to capture is it's period and amplitude, the display will always look exactly the same, only the axies will vary.

If that is the case, then you should consider a circuit that determines the frequency (which may already be known), or simply measures the amplitude (which is the only other variable -- since phase seems not to be an issue).

Now, if the waveform gets distorted, and that is significant, then you really do need to sample the waveform at numerous points to build up a picture of it.

So, exactly what is it that you want to use this to measure?

3. ### saurabh17g

72
0
Mar 8, 2010
we are determining the properties of rocks by passing waveforms(preferably sine) and studying the input and output waveform shapes. The frequency will vary as shown in first post.So, the intention is to sample the waveform and plot them after getting the points. Problem is getting the high speed counters , ICs , RAM etc.
is there any better and smarter way to implement all this.

4. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,451
2,809
Jan 21, 2010
Purchasing a storage oscilloscope would be my first option.

edit: you see, there's a problem with being specific. When you say sine wave, people may believe you mean what you say. In this case you didn't. Better to say a "distorted sine wave", if that's what you really mean.

Depending on the amount of distortion, you will need more data points and higher resolution to measure the effects.

You can get cheap (Chinese) storage scopes from as low as \$US300. They may not be the best devices in the world, but would likely be better than something you cobble together yourself.

Last edited: May 5, 2010
5. ### saurabh17g

72
0
Mar 8, 2010
But that is not sufficient.

The entire circuit of frequency generator and receiver will go in cetrifuge where there is no space to keep DSO .So a transmitted and received signal need to be captured and digitized and sent to PC in a wireless format. Also, there is no need to capture entire signal. Just one period or wave is sufficient to be captured and stored.
Person outside will have a remote who will set a frequency count and that will be transmitted wirelessly to frequency generator circuit.

Last edited: May 5, 2010
6. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,451
2,809
Jan 21, 2010
Why not inductively couple your signal to a receiver outside the centrifuge (you may need different techniques for various frequency ranges though). Like this.

If the signal was always used to modulate a (say) 10 MHz carrier, you could display the resulting signal on your scope. The carrier would effectively determine the sampling rate, and would probably have to be constant given that you would not want to change coils for each change in frequency. One or 2 turns may be all you need on either side.

And do you really need a centrifuge? Could you not place the sample under compression?