is it true, oscilloscope must reach at least 4x observed freq?

[oxy]

Hi,

some people say, if I wanna observe 100MHz, I must use an oscilloscope that reaches at least 400MHz.

Is it true?

[bdunham7]

The first step in understanding this question is to accurately explain what you mean by "100MHz".

[coppice]

Hi,

some people say, if I wanna observe 100MHz, I must use an oscilloscope that reaches at least 400MHz.

Is it true?

Are you talking about bandwidth (100MHz) versus the sampling rate needed to achieve that in a digitising oscilloscope? By the Shannon/Nyquist/Whittaker/Kotelnikov sampling theorem you will need a sampling rate more than 2 times the bandwidth. So, by theory you need a sampling rate of 200Msps. In practice you need something a bit higher than that. Many scopes use something like 2.5x, while others use something considerably higher, like 4x or 5x.

[Fungus]

some people say, if I wanna observe 100MHz, I must use an oscilloscope that reaches at least 400MHz.
Is it true?

No. The minimum needed would be 250Mhz to observe a 100Mhz sine wave.

That's why you see (eg.) 100Mhz, 4-channel 'scopes with 1GHz sample rate.

[bdunham7]

No. The minimum needed would be 250Mhz to observe a 100Mhz sine wave.

That's why you see (eg.) 100Mhz, 4-channel 'scopes with 1GHz sample rate.

MSa/s and MHz are two different things and the way you put it is very confusing at best.  The OP didn't actually specify a DSO and from the nature of his question, I don't think he's referring to sample rates.

[m k]

Hi,

some people say, if I wanna observe 100MHz, I must use an oscilloscope that reaches at least 400MHz.

Is it true?

What shape?

[jonpaul]

See Claude Shannon 1948  BSTJ paper and book " The mathematical Theory of Communication"

He was indeed the "Father of Information Theory.

jon

[nctnico]

The first step in understanding this question is to accurately explain what you mean by "100MHz".

Agreed. 100MHz bandwidth means you can get a 100MHz sine wave on screen with 3dB amplitude error at most.

[oxy]

Lets take the example of the picoScope Series 6000E:
It has several models with bandwidths ranging between 500MHz and 1 GHz, yet all of them with sampling rate of  2.5Gs/s.

As I understand the sampling rate sets over Nyquist the max. frequency that I can observe.  Thus what hardware specification differentiates those oscilloscopes on the bandwidth?

Thanks for ur very nice inputs!   

[coppice]

Lets take the example of the picoScope Series 6000E:
It has several models with bandwidths ranging between 500MHz and 1 GHz, yet all of them with sampling rate of  2.5Gs/s.

As I understand the sampling rate sets over Nyquist the max. frequency that I can observe.  Thus what hardware specification differentiates those oscilloscopes on the bandwidth?

Thanks for ur very nice inputs!   

The analogue front end bandwidth.

[nctnico]

Lets take the example of the picoScope Series 6000E:
It has several models with bandwidths ranging between 500MHz and 1 GHz, yet all of them with sampling rate of  2.5Gs/s.

As I understand the sampling rate sets over Nyquist the max. frequency that I can observe.  Thus what hardware specification differentiates those oscilloscopes on the bandwidth?

Likely they all have the same hardware but there is a software controlled filter in the input circuitry which limits the actual bandwidth. In some cases the lower bandwidth models have some extra filtering components in the input circuitry but the basic hardware design is the same.

[bdunham7]

Lets take the example of the picoScope Series 6000E:
It has several models with bandwidths ranging between 500MHz and 1 GHz, yet all of them with sampling rate of  2.5Gs/s.

As I understand the sampling rate sets over Nyquist the max. frequency that I can observe.  Thus what hardware specification differentiates those oscilloscopes on the bandwidth?

That is entirely unrelated to your initial question.  As already mentioned, the bandwidth rating is typically designated as the frequency where the amplitude response is down -3dB or to about 70%.  The high frequency roll off can be due to the bandwidth limitations of the analog inputs or it can be the result of something further up the line (software filter, etc).

[tautech]

If as it seem sampling rate is what the thread is really about however 5x sampling vs BW is more the industry standard today.
This with a shared ADC design provides 2.5x BW sampling which comfortably meets Nyquist.

[tggzzz]

Lets take the example of the picoScope Series 6000E:
It has several models with bandwidths ranging between 500MHz and 1 GHz, yet all of them with sampling rate of  2.5Gs/s.

As I understand the sampling rate sets over Nyquist the max. frequency that I can observe.  Thus what hardware specification differentiates those oscilloscopes on the bandwidth?

Thanks for ur very nice inputs!   

For digital signals the period is irrelevant; only rise time is important. For a little theory and some practical measurements, see  https://entertaininghacks.wordpress.com/2018/05/08/digital-signal-integrity-and-bandwidth-signals-risetime-is-important-period-is-irrelevant/

If you have a repetitive signal, the sampling frequency is separate to the signal frequency. I have a scope that measures 50ps rise times with ~40kSa/s. No, that does not violate Shannon/Nyquist! Various manufacturers have different names for the techniques, e.g. equivalent time sampling.

[tggzzz]

If as it seem sampling rate is what the thread is really about however 5x sampling vs BW is more the industry standard today.
This with a shared ADC design provides 2.5x BW sampling which comfortably meets Nyquist.

So you think that to usefully observe a 1kHz digital signal, you only need a 5kHz scope? That's nonsense, as you well know.

The only thing that matters is the rise time. Digital circuits don't "care" when the next transition might occur, if ever.

[coppice]

Lets take the example of the picoScope Series 6000E:
It has several models with bandwidths ranging between 500MHz and 1 GHz, yet all of them with sampling rate of  2.5Gs/s.

As I understand the sampling rate sets over Nyquist the max. frequency that I can observe.  Thus what hardware specification differentiates those oscilloscopes on the bandwidth?

Thanks for ur very nice inputs!   

For digital signals the period is irrelevant; only rise time is important. For a little theory and some practical measurements, see  https://entertaininghacks.wordpress.com/2018/05/08/digital-signal-integrity-and-bandwidth-signals-risetime-is-important-period-is-irrelevant/

If you have a repetitive signal, the sampling frequency is separate to the signal frequency. I have a scope that measures 50ps rise times with ~40kSa/s. No, that does not violate Shannon/Nyquist! Various manufacturers have different names for the techniques, e.g. equivalent time sampling.

I think you mean "for repetitive digital signals".

[switchabl]

This with a shared ADC design provides 2.5x BW sampling which comfortably meets Nyquist.

Yes, but the sampling theorem refers to the total bandwidth of a (band-limited) signal, not the 3dB bandwidth of the front-end. So realistically, there is nothing comfortable about it. At 2.5x you will either have to live with some non-negligible aliasing or with some non-negligible overshoot (steep filter).

[tautech]

This with a shared ADC design provides 2.5x BW sampling which comfortably meets Nyquist.

Yes, but the sampling theorem refers to the total bandwidth of a (band-limited) signal, not the 3dB bandwidth of the front-end. So realistically, there is nothing comfortable about it. At 2.5x you will either have to live with some non-negligible aliasing or with some non-negligible overshoot (steep filter).

Sure, however those that understand their instruments and could push them near their limits will use a single channel on each ADC so to be back near 5x sampling to minimize the chance of aliasing and also provide for greater accuracy measurements.

[tggzzz]

Lets take the example of the picoScope Series 6000E:
It has several models with bandwidths ranging between 500MHz and 1 GHz, yet all of them with sampling rate of  2.5Gs/s.

As I understand the sampling rate sets over Nyquist the max. frequency that I can observe.  Thus what hardware specification differentiates those oscilloscopes on the bandwidth?

Thanks for ur very nice inputs!   

For digital signals the period is irrelevant; only rise time is important. For a little theory and some practical measurements, see  https://entertaininghacks.wordpress.com/2018/05/08/digital-signal-integrity-and-bandwidth-signals-risetime-is-important-period-is-irrelevant/

If you have a repetitive signal, the sampling frequency is separate to the signal frequency. I have a scope that measures 50ps rise times with ~40kSa/s. No, that does not violate Shannon/Nyquist! Various manufacturers have different names for the techniques, e.g. equivalent time sampling.

I think you mean "for repetitive digital signals".

No, for repetitive signals. To put it anthropmorphically, the scope neither knows nor cares how you interpret the voltage waveform.

[coppice]

Lets take the example of the picoScope Series 6000E:
It has several models with bandwidths ranging between 500MHz and 1 GHz, yet all of them with sampling rate of  2.5Gs/s.

As I understand the sampling rate sets over Nyquist the max. frequency that I can observe.  Thus what hardware specification differentiates those oscilloscopes on the bandwidth?

Thanks for ur very nice inputs!   

For digital signals the period is irrelevant; only rise time is important. For a little theory and some practical measurements, see  https://entertaininghacks.wordpress.com/2018/05/08/digital-signal-integrity-and-bandwidth-signals-risetime-is-important-period-is-irrelevant/

If you have a repetitive signal, the sampling frequency is separate to the signal frequency. I have a scope that measures 50ps rise times with ~40kSa/s. No, that does not violate Shannon/Nyquist! Various manufacturers have different names for the techniques, e.g. equivalent time sampling.

I think you mean "for repetitive digital signals".

No, for repetitive signals. To put it anthropmorphically, the scope neither knows nor cares how you interpret the voltage waveform.

OK. I have read the article now. What you said is confusing. You replied to something talking about sampling rates, and gave a reply including reference to a slow sample interval. I assumed by "period" you were referring to the sampling interval. From the article you are referring to the period of a waveform to be sampled. Most digital waveforms don't really even have a period. They don't repeat, so they aren't periodic.

[tggzzz]

Lets take the example of the picoScope Series 6000E:
It has several models with bandwidths ranging between 500MHz and 1 GHz, yet all of them with sampling rate of  2.5Gs/s.

As I understand the sampling rate sets over Nyquist the max. frequency that I can observe.  Thus what hardware specification differentiates those oscilloscopes on the bandwidth?

Thanks for ur very nice inputs!   

For digital signals the period is irrelevant; only rise time is important. For a little theory and some practical measurements, see  https://entertaininghacks.wordpress.com/2018/05/08/digital-signal-integrity-and-bandwidth-signals-risetime-is-important-period-is-irrelevant/

If you have a repetitive signal, the sampling frequency is separate to the signal frequency. I have a scope that measures 50ps rise times with ~40kSa/s. No, that does not violate Shannon/Nyquist! Various manufacturers have different names for the techniques, e.g. equivalent time sampling.

I think you mean "for repetitive digital signals".

No, for repetitive signals. To put it anthropmorphically, the scope neither knows nor cares how you interpret the voltage waveform.

OK. I have read the article now. What you said is confusing. You replied to something talking about sampling rates, and gave a reply including reference to a slow sample interval. I assumed by "period" you were referring to the sampling interval. From the article you are referring to the period of a waveform to be sampled. Most digital waveforms don't really even have a period. They don't repeat, so they aren't periodic.

When considering sampling signals, it is necessary to be clear what the signal is. If you have an 10kHz audio waveform modulated onto a 10MHz carrier, then to recover the audio waveform you don't have to sample at >20MSa/s; >20MSa/s is sufficient. You do still need a >10MHz front end and a <50ns sampling pulse width, though.

If, as in a TDR, the waveform you are observing doesn't change over time, then you can sample arbitrarily infrequently. The slope I referred to has >10 times slower sampling mode that matches the speed of a thermal pen recorder

[coppice]

When considering sampling signals, it is necessary to be clear what the signal is. If you have an 10kHz audio waveform modulated onto a 10MHz carrier, then to recover the audio waveform you don't have to sample at >20MSa/s; >20MSa/s is sufficient. You do still need a >10MHz front end and a <50ns sampling pulse width, though.

If, as in a TDR, the waveform you are observing doesn't change over time, then you can sample arbitrarily infrequently. The slope I referred to has >10 times slower sampling mode that matches the speed of a thermal pen recorder

You only need to sample at twice the bandwidth of the signal (or at one times the bandwidth for analytic sampling). However, for that to work well you have to be sure the signal has no content outside the band of interest. Also, the jitter in the sampling aperture needs to be just as low for a 2Msps ADC being used to sample a 1MHz band around 2GHz, as it would for a 5Gsps ADC sampling the entire band up to 2.0005GHz.

Scopes don't have any way to clean up a signal, and their sampling jitter is not usually any better than it needs to be for their basic needs. So, it is rarely practical to use them to sample a limited band well above DC. The jitteriness of the sampling aperture can work in your favour in some situations. It makes ETS work a lot better, for example.

[David Hess]

Bandwidth and sampling rate are completely independent.  The sampling rate must meet the Nyquist criteria to accurately reconstruct the waveform, but this has nothing to do with the bandwidth as defined by the -3dB amplitude response, which is why equivalent time sampling works.

As pointed out by tggzzz, a bandwidth limited signal can be reconstructed with a sampling rate greater than twice the bandwidth no matter where in the frequency spectrum it is, within the bandwidth of the sampler.  The sampling function itself is equivalent to RF mixing, and the circuits can be identical.  RF mixers make great microwave samplers when driven with a suitable pulse through their local oscillator port.  The sampling part of an analog-to-digital converter can be modeled as a down-conversion mixer.

some people say, if I wanna observe 100MHz, I must use an oscilloscope that reaches at least 400MHz.

Is it true?

If you want to see anything other than the fundamental sine wave component of the 100 MHz signal, then the oscilloscope bandwidth needs to be much higher to cover the harmonics.  Otherwise a 100 MHz oscilloscope viewing a 100 MHz signal will only display the 100 MHz fundamental sine wave component, and with a 3dB amplitude error.

A more useful oscilloscope specification is rise time, which in the general case in nanoseconds is 350 divided by the bandwidth in MHz, so 100 MHz yields 3.5 nanoseconds.  Oscilloscope rise time needs to be several times faster than the rise time of the signal for an accurate display.

[tautech]

If you want to see anything other than the fundamental sine wave component of the 100 MHz signal, then the oscilloscope bandwidth needs to be much higher to cover the harmonics.  Otherwise a 100 MHz oscilloscope viewing a 100 MHz signal will only display the 100 MHz fundamental sine wave component, and possibly with a 3dB amplitude error.

FTFY.

It need be noted label BW can be substantially different from actual -3dB BW as is the case with the 100 MHz rated SDS2104X Plus which the 1st we received I tested with 3 yes 3 signal sources to convince myself the ~185 MHz result was actually real !
Supporting that result is they are also supplied with 200 MHz probes.

[Fungus]

You only need to sample at twice the bandwidth of the signal (or at one times the bandwidth for analytic sampling).

Completely false.

Imagine a sine wave that you're sampling at exactly 2x frequency.

a) You might sample the signal exactly on the peaks/troughs in which case you'll be fine.

b) OTOH you might sample it exactly on the zero-crossing points, in which case you'll see nothing at all.

You can also get every possible value in between (a) and (b), it's just dumb luck.

If you sample at 99.99999% of Nyquist you'll drift slowly between (a) and (b) and see the amplitude varying on screen ("AM effect").

2.5x Nyquist is the minimum to avoid this AM effect.