If we apply a signal to a device to measure its rise time ( for example in oscilloscope) then rise time calculated from output shown by that measurement device is the combined rise time of the device and the signal because the device itself has some rise time. No one device has zero rise time. hence we can't measure exact rise time of a signal .
This involves the question: what means "exact"?
I think, theoretically, no measurement really is "exact".
However, there are methods which allow to approximately recalculate the rise of the DUT (device under test) - that means to eliminate (up to a certain degree of exactness) the unwanted contribution of the oscilloscope.
This approximation is based on the assumption that the scope behaves as a first order lowpass circuit.
In this case the squared rise time of the DUT is calculated to be the squared measured rise time minus the squared inherent rise time of the scope, which of course depends on the bandwidth.
It can be shown that this formula results in a maximum error of app. 8% for t(DUT)=t(osc).
please have a look to my slidind DFT paper for time of arrival measurement.
It explain new high resolution technique.
Said
Lutz von Wangenheim:
If i want less than 1 % error then...what i have to do....?
Simple advice: Choose a suitable (broadband) oscilloscope) which ensures that the inherent rise time is much smaller than the rise time of the DUT.
Knowing the calculation method mentioned in my former post (geometrical sum of both contributors) you can calculate "how much" rise time difference between both contributors is necessary for an error of 1%.
the technique I referred to in my paper gives you less than 1% errors with a minimum of 5dB SNR.
Please considered this comment only as an increase of the objective pursued by another answers, due that you must include no only the rise time constants of the oscilloscope tosc itself but you must consider the test probe input rise time even.
Sincerely Hugo Colombo
In my paper,'New one cycle microprocessor based differential relay". I found i (the instaneteneious value of current)and di/dt (slop or rise of signal) by taking samples, the same method can be used, to find the rise time of signal, instead of C.R.O..
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering
- Signal Processing