Is it possible to measure the pulse height of 20 ns pulses at 100 Khz with Fpga?

[itai]

I would like to measure the pulse height of 20 ns width pulses at 100 Khz. The pulses arrive with a gating signal. I guess this would require a sampling speed of at least 500 MHz.

Itai

[tbob]

20nS pulse translates to 50MHz frequency.  Nyquist theory says your sampling rate needs to be at least twice the frequency you are wanting to measure.  So 100MHz sampling should be able to detect 20nS pulse.  I like to sample faster than twice the frequency.  I would use 200MHz at a minimum, but that is just personal preference.  500MHz should be plenty fast enough.

[Ben]

I agree with tbob if you are just trying to detect the pulse. If you want to measure the rise time ripple etc, then the 10 samples per pulse may not be fast enough. Nyquist talks about the highest freq component of your signal. For a 50MHz pulse the lowest freq component is 50MHz and the others are above (Check CRC handbook for fourier transforms).

Discalimer: Anything and evrything that I say that is math related should be taken with grain of salt. I once tried to take the log of a negative number!

Sanity Check Q:

Can a FPGA sample faster than it can loop? I thought the top end loop rate fro a FPGA was 40MHz

Oh well.... just trying to help!

Ben

Message Edited by Ben on 01-02-2008 01:03 PM

[BruceAmmons]

An FPGA can't sample analog data nearly that fast, so you won't be able to do that.  With digital inputs, it is possible to get the FPGA up to 100 MHz, possibly even 200 MHz by using clock multipliers.  You might be lucky enough to detect the pulses, but you wouldn't be able to measure anything about them.  However, your pulses may be too short to register in the TTL circuitry.

Bruce