Why is scope bandwidth by thumb rule approx equal to 0.35/t(rise time) and not 0.35/t(fall time) ? In some logic families fall time is faster than rise time

 

The assumption is that riestime and falltime are roughly the same, so it is only necessary to use one in the calculation.

 

Why is scope bandwidth by thumb rule approx equal to 0.35/t(rise time) and not 0.35/t(fall time) ?

Because from the perspective of the scope, there is no difference between the two: a scope's "risetime" is the apparent time, as viewed on the scope trace, that a signal takes to transition from one voltage level to another, even though that transition is actually instantaneous.

In some logic families fall time is faster than rise time

That refers to the rise time and fall time of the logic's output, and has nothing to do with calculating the scope bandwidth.

 

Thank you Dick Cappels and OBW0549 for your response. OBW0549, for logics with a faster fall time , assuming I have a scope with bandwidth Just about covering the rise time of such a logic , if I did want to measure the HIGH TO LOW transition time of signals in this logic family through observation on such a scope , the measurement would be in error because due to the the higher harmonics at the transition being beyond scope bandwidth? Would I end up measuring apparently a slower rise time than actual ?

 

OBW0549, for logics with a faster fall time , assuming I have a scope with bandwidth Just about covering the rise time of such a logic , if I did want to measure the HIGH TO LOW transition time of signals in this logic family through observation on such a scope , the measurement would be in error because due to the the higher harmonics at the transition being beyond scope bandwidth? Would I end up measuring apparently a slower rise time than actual ?

That's exactly what would happen; if the scope risetime is of the same magnitude as that of the signal you're trying to evaluate, the rise/fall time of the signal will appear to be slower than it actually is.

Solution: faster scope.

 

That's exactly what would happen; if the scope risetime is of the same magnitude as that of the signal you're trying to evaluate, the rise/fall time of the signal will appear to be slower than it actually is.

Solution: faster scope.

Thank you

 

This might be useful -

https://www.tek.com/document/online/primer/xyzs-scopes/ch3/evaluating-oscilloscopes


Regards, Dana.

 

Why is scope bandwidth by thumb rule approx equal to 0.35/t(rise time) and not 0.35/t(fall time) ? In some logic families fall time is faster than rise time

Hi,

I think it is based on BOTH rise AND fall, but the rise and fall are assumed to be nearly the same.
I say this because that calculation is roughly one half what a particular rise time would be able to produce if it was a sine wave and the device was an op amp for example rather than a scope.
If you compare with logic families or op amps, you'll find the spec is roughly 1/2 that of the scope because there they are talking about just a rise or a fall and not both simultaneously.

Does it make sense? Not really, until you figure in the criterion for reproducing a sine wave with a given sample time and then you need 2 samples as a minimum.

We could look into this more too.

 

Why is scope bandwidth by thumb rule approx equal to 0.35/t(rise time) and not 0.35/t(fall time) ? In some logic families fall time is faster than rise time

Because it's the same, fall time is just a negative rise time. What matters is that there is a semi-instantaneous change in voltage level, the direction is irrelevant.

Besides, it should be noted that the risetime factor of 0.35 is only valid for scopes that have a gaussian response at the upper BW limit (i.e. slow roll-off), which in general is true only for digital scopes with a BW of less than 1GHz. Digital scopes with a bandwidth of 1GHz and more usually have a brickwall filter (sharp cut-off), which translates to a risetime factor of 0.4 or more.