Thanks Bob. Capturing both edges on a single timer works like a charm (I use a global var as a counter, and use it to populate a small global array from within the interrupt). By capturing three values, I can do some simple math to figure out which edge is which, and hence grab the pulse width. -Pete. --- In lpc2000@yahoogroups.com, "lpc2100_fan" <lpc2100_fan@y...> wrote: > > Peter, > > use one Timer and capture on both edges. Depending on your system > setup this limits the minimum pulse width somewhat but assuming it is > more than 1 usec it should be doeable. You have to make sure that you > did already read the first value before the second got captured. > Another option is to use two capture inputs (not two timers) and do it > like you proposed the difference of the two timer snapshots. > > hth Bob > > > --- In lpc2000@yahoogroups.com, "Peter Gray" <peter_f_gray@y...> wrote: > > > > On other microcontrollers I've used, it was possible to start a single > > timer on a rising edge, and capture-and-reset on a falling edge (to > > capture the pulse time). > > > > It doesn't look like this method is available on the (for example) > > LPC2131. What would be a good mechanism for these microcontrollers? > > Start two timers, feed the PWM to two pins, have one capture-and- reset > > on rising edge, the other on the falling edge? (the difference between > > the two capture registers would be the pulse time) or is there a > > better way? > > > > Thanks, > > -Pete.
Message
Re: PWM demodulation
2005-04-26 by Peter Gray
Attachments
- No local attachments were found for this message.