Real Time Sampling Rate

[Suraj143]

In my oscilloscope it has "Real-time Sampling Rate = 1GS/s".
Bandwidth = 50MHz.

In the 50Hz AC waveform one periods time will be 20mS.

My question is from my oscilloscope how many samples will give within this period?

Thanks

 

[cowboybob]

.02s X 1GS/s = 20MS for one cycle.

 

[Suraj143]

Ok thanks bob.

You are correct.That means I get 20 Mega Samples per cycle.So I get an extremely smooth output waveform from my scope!!

So here comes the secret doubt!!

Does this sample rate & Bandwidth has a relationship?

In other words my Osc bandwidth is 50Mhz.If I feed a 100MHz pulse will this sample rate reduced?

 

[Nigel Goodwin]

Ok thanks bob.

You are correct.That means I get 20 Mega Samples per cycle.So I get an extremely smooth output waveform from my scope!!

I'm pretty sceptical about that!

There's no point whatsoever in using so many samples in such a short time for a 50Hz signal, much more sensible to sample much slower and for longer, so you can scroll along the sample buffer.

I would expect the highest sample rate to only be used on the highest timebase settings, or at least to be adjustable manually.

So here comes the secret doubt!!

Does this sample rate & Bandwidth has a relationship?

In other words my Osc bandwidth is 50Mhz.If I feed a 100MHz pulse will this sample rate reduced?

You can't feed a 100MHz pulse in, the scopes bandwidth is only 50MHz, and that's essentially for a sinewave, and 3dB down at that point.

 

[MrAl]

Ok thanks bob.

You are correct.That means I get 20 Mega Samples per cycle.So I get an extremely smooth output waveform from my scope!!

So here comes the secret doubt!!

Does this sample rate & Bandwidth has a relationship?

In other words my Osc bandwidth is 50Mhz.If I feed a 100MHz pulse will this sample rate reduced?

Hi there,

Sampling systems often used bandwidth limiting filters on their front end. This helps reduce some problems that come up with sampling a real time signal.

The bandwidth is usually specified in terms of the -3db points in the frequency response characteristic. This means that the signal at the bandwidth spec frequency is 3db down from DC, or about 0.71 of the amplitude it would be for DC (or say a 1Hz AC signal).

So if your scope is 50MHz, then if you measure the output of a signal generator with a 1Hz signal that is 2v peak to peak and it shows up as 2v peak to peak (say 2 divisions peak to peak) and then increase the frequency to 50MHz without changing the output amplitude the new signal, even though the generator puts out the same amplitude, will now measure only 1.42v peak to peak (1 division). The reason for the decrease on the scope is because the input filter reduces the input amplitude as you approach the 50MHz point.

At 100MHz (we are talking sine waves here) the signal that shows up on the scope would be even less than that, down to 1 division now because it would be attenuated even more at this higher frequency. If the filter has a constant -6db rolloff, then the new signal on the scope would only measure 1 division.

A pulse is a different story however. That's a rectangular wave and it is made up of many sine wave harmonics that for a 50MHz signal start at 50MHz and increase higher and higher to infinity. Since the scope input filter attenuates more and more with frequency, the harmonics will be affected differently so the wave that shows up on the scope will have a different shape, notably with some sloping to it rather than a clean rectangular wave.

 

[crutschow]

..........................
The bandwidth is usually specified in terms of the -3db points in the frequency response characteristic. This means that the signal at the bandwidth spec frequency is 3db down from DC, or about 1/2 of the amplitude it would be for DC (or say a 1Hz AC signal).
.............................
At 100MHz (we are talking sine waves here) the signal that shows up on the scope would be even less than 1 division now because it would be attenuated even more at this higher frequency. If the filter has a constant -3db rolloff, then the new signal on the scope would only measure 1/2 of one division.
.........................

dB is always in power. -3dB is .707 of the voltage. It is 1/2 of the power.

The rolloff for a simple, single-pole filter is 6dB/octave so, if the scope has that, it would be around -7dB down at 100MHz or 0.477 the low frequency amplitude.

 

[MrAl]

Hi,

Yes, my mistake, i'll have to correct my previous post...Fixed.

 

[Mr RB]

...
In other words my Osc bandwidth is 50Mhz.If I feed a 100MHz pulse will this sample rate reduced?

Most 50MHz scopes with 1G/S will still display a 100MHz signal, but the display is not guaranteed to perfectly reproduce the 100MHz signal as it is above the bandwidth so the display will probably be smaller, more smoothed, etc. But a 50MHz sine should display properly being within the bandwidth of the 50MHz scope.

 

[Suraj143]

Ok guys now I understood.That 3db story comes in sine waves or also in square waves?

Or this 3db story is acting when exceeding my bandwidth (>50MHz) !!

 

[crutschow]

The -3dB is for sine waves. Square waves have higher frequency components so even a square wave with a lower fundamental frequency, of say 10MHz, will show some rounding of the square wave edges due to the 50MHz rolloff. A 50MHz square wave will likely look much like a sine wave for the same reason.