Hello,
I'd like to use a PIC or MCU for an application. I can find many parts that fit most of my needs, but the hardest problem is that I have an
80MHz PWM signal I need to measure. Could anyone suggest an external timer chip that would run that fast that I could use to time the pulse width (up to 24 bits)?Then I could take the time from that chip and use it on a slower PIC to gather a couple of other pieces of data to send to a host?
Thanks,
Les Elkins
So you want to measure the width of a 80MHz pulse with 24-bit resolution? This would require a 1.3 PHz clock, about the frequency of UV light.
Even a fast FPGA (say counter running at 1GHz) can only give you
1G/80M=12.5 counts, a resolution of 2.5 bits.Can you tell us more about this application and what you want to do? Because just measuring what you ask seems out of the question. Or wait another 40 years, with clocks doubling every two years, we should have arrived in the PHz range by then.
-- Stef (remove caps, dashes and .invalid from e-mail address to reply by mail)
A CPLD would be ideal for this application, it shouldn't be difficult to design the logic required.
Leon
Of course, as Stef has said, 24 bits at 80 MHz is impossible to achieve.
Leon
Create your own in a PLD/CPLD/FPGA and read it from the micro of your choice as 8/16/24/32 bit width data depending on micro and CPLD/...
--
Paul Carpenter | paul@pcserviceselectronics.co.uk
PC Services
GNU H8 & mailing list info
For those web sites you hate
Hello,
Thanks for the replies.
Sorry if I mis-spoke. I meant I need to measure a pulse width clocked at 80MHz, and the widest pulse should fit in 24 bits. (Actually about
16 bits, since the pulses will come along at a max of about 5KHz, 2^16 is a little bigger than 80,000,000/5,000).It sounds like I may need to build a CPLD/FPGA to do the counting. If this is the case I'd be best off to put the whole thing on one chip, I suppose. I'll look into that. My preferred solution was an external timer chip, and ingest the results into a microcontroller, but hey, that's just because I'm really just a software guy, and that would be less work for me....
Thanks,
-Les
If you want full duty cycle range, then yes, doing it all in a CPLD makes sense. You can set up a >=14 bit counter, and a >=14 bit latch, and have the uC read the latch very cycle, whilst the counter spins for the next reading.
- or if this is stable, you can capture one, and read during the next, which is less macrocells.
Another device you could look at, is the 74HC590, for ideas, or for part of the chain. Struggles at 80MHz tho.
-jg
Try
Regards Anton Erasmus
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.