Oscilloscopes bandwith and sampling rate (digital)

[@xi@g@me]

Hello all,
I'm trying to get minimal knowledge around oscilloscopes so I understand the technical spec and I can buy an oscilloscope with features good enough for personnal use.
Recently I tried to get information about bandwidth and sampling rate (for digital scopes of course), and found this video made by Keysight :

The key elements I understood from this video :

- The given bandwidth is the frequency at which with have a signal loss ratio of 3 db (~70% is the signal passes through);
- The best practice to have a good measurement would be to use a scope / probe combo with a bandwidth at least 3 times higher than the main frequencey of the signal I want to probe;
- The scope AND the probe bandwidth have a global influence, if both of them have a bandwidth of X the actual bandwidth would be less;
- For non sinusoidal signals (like clock pulses), the rising edge timing shall be used to compute the main frequency

The info I got there is for a single source however, so I would like to read your thoughts about that, especially if you are a professional with great experience in using oscilloscopes, so I can have a better idea if the video says right and if I got the info correctly.

(As for the rest, I guess that sample rate is the frequency at which the signal is sampled by the digital oscilloscope, and "kdots" or "Mdots" value is the number of samples the tool can store in memory)

Thanks for sharing your knowledge with me !

 

[Nanren888]

The info I got there is for a single source

You mean that the information was from a single source, Keysight?
Or something about combinations of signals?
.
The video is about probe selection. It's all good.

the rising edge timing shall be used to compute the main frequency

The rise time is used as an indication of the bandwidth required in your measurement system to adequately reflect the rising signal. Higher frequency signals rise faster. Fast rise times need higher frequency capability.

sample rate is the frequency at which the signal is sampled by the digital oscilloscope

Yes. Check out the scopes modes in this respect. For periodic signals, many scopes allow you to catch a few samples each time the waveform repeats, overlaying them as if they were captured at a higher sample rate than it is capable of. This can be great to look at periodic signals but can also lie very convincingly. To do this well, the scope sampler needs a really good sample and hold, aperture time, compared to the actual sample rate. Such modes will be in the specs about sampling and triggering.

8-bit wide memory? Scopes sample to digital. Have not bought one in a while, so not familiar with the width of the samples, that is how many bits per samples are offered.

Long memory can be great if you need to trigger on something infrequent or complicated and then scroll back to something before.
If you are looking Keysight, then I guess you've got a reasonable budget. If that's an issue, they offer some more economic models specifically aimed at universities.

 

[bertus]

Hello,

Have a look at this PDF:
www.tek.com/dl/03W_8605_6.pdf

More scope subjects over here:
https://www.tek.com/learning/oscilloscope-tutorial

Bertus

 

[@xi@g@me]

You mean that the information was from a single source, Keysight?
Or something about combinations of signals?

I meant my source of information, yes. To consider something in a domain I am not an expert at, I'd rather check multiple sources.

or periodic signals, many scopes allow you to catch a few samples each time the waveform repeats, overlaying them as if they were captured at a higher sample rate than it is capable of. This can be great to look at periodic signals but can also lie very convincingly. To do this well, the scope sampler needs a really good sample and hold, aperture time, compared to the actual sample rate. Such modes will be in the specs about sampling and triggering.

Didn't know about that, good to know

If you are looking Keysight, then I guess you've got a reasonable budget

I'm not specially aiming at keysight, it's just that I know they make oscilloscopes and it's the first video I've found that would explain what I was looking for.

Hello,

Have a look at this PDF:
www.tek.com/dl/03W_8605_6.pdf

More scope subjects over here:
https://www.tek.com/learning/oscilloscope-tutorial

Bertus

I'll have a look, thanks for the info

 

[(*steve*)]

I meant my source of information, yes. To consider something in a domain I am not an expert at, I'd rather check multiple sources.

Keysight are a reliable source for technical information like this. Alternatives could include Tektronix and Rohde & Schwarz. Any manufacturer that produces scopes with bandwidths in the 10's of GHz or higher (not sample rate) is very likely to know their stuff.