Using timers instead of ADC to measure tx signal length

[Tepco]

This question of ADC vs. timers is actually more interesting and can end whit unusual conclusions. Unfortunately, most internal ADCs, even on "high end" MCUs are quite crappy, slow and low resolution ones. Comparator-timer combination can be even worst, due to it's analog part (comparator), noise and reference related problems etc, hardly can match internal ADC resolution, also no S&H, slower conversion speed and all...Newer do in software something that is already done in silicone. Real deal is probably some dedicated high performance ADC, say 16 bit and 1MSPS (TI recently introduced something "reasonably" priced). Later you don't need any serious processing power (there is plenty of time between pulses for calculation), point is to collect as much information from target, and fast, say, 64 points 1uS apart, and then process it. This will eliminate major drawback of analog detectors, using integration as a first signal processing step and losing all time domain data from target.

[pebe]

I assumed that these two lines in mario's original posting

I have been able to use timers at a rising edge to measure the frequency of square waves with a resolution of 5 uS.

Why do most designs insist on using ADC when a simple 16bit timer will suffice.

both refered to frequency measurement. Perhaps they were dealing with different subjects?

[mario]

I assumed that these two lines in mario's original posting



both refered to frequency measurement. Perhaps they were dealing with different subjects?

Not at all pebe, those two statements were dealing with the same subject since a shift in frequency would reveal information about a target. Depending on the amount shifted, it would indicate response to a metal, earth magnetic field, mineralzation, etc.

I believe this is what Tepco had in mind when he mentioned " 64 points 1uS apart" in his last comment. The more information sent across the coil, the finer the resolution obtained from any given target.

It's an extreme approach, compared to what has already been done in analog, like he also mentioned.