Quote:
Originally Posted by mark_in_manc
I think I was concerned as not only the pulse rate, but the mark-space ratio will be changing, since the 'mark' length is fixed and the 'space' will just last until the next trigger event, which depends on the period of the signal coming out of the transducer. That doesn't sound like a recipe for a linear relationship between frequency and average output voltage to me. Well, we'll see when the bits arrive - maybe I'll have to go for really short 'marks' and then amplify the output to make it more linear.
|
The relevant point here is that the pulse height and width are fixed, so the average voltage of a train of them is proportional to the rate at which they happen.
If pulse height is 5V and width is 1ms and frequency is 100pps, average voltage will be 0.5V. Double the pulse rate and the average also doubles to 1V. Double it again gets 2V.
This is a linear relationship of 0.5V/100Hz. Obviously, for the 1ms pulse width, the relationship hits the endstop at 1kHz and 5V when the pulses run into each other and the output is a continuous 5V.
Pick your pulse height and width appropriately and theoretically you can have any range you want, though in practice, circuit non ideality will impose limits to what can be achieved.
This retriggerable monostable driving a moving coil meter technique was used for diy rev counters historically and no doubt other applications too.