Post by thread:16c54,RC osc.. how slow can I go?

I've coded a 16c54 to drive a serial device that requires CS, CLK and Data. I currently only have LED's connected to the output ports but the 'bit twiddling' is too fast for me to check if I have all the timing correct. I am using an RC osc (100K and 330pf cap). What I want to know is how slow can I make the osc using RC parts? If I increase the cap too large am I stressing the PIC's shunt transistor? (I am assuming the resistor charges the cap and the PIC shorts the cap for the osc freq). If the resistor gets too high in value what problems might I have (any powerup reset problems?) Thanks for any information. The data sheet on the PIC doesn't give a wide enough range on the RC osc chart. I know I could code in a delay loop, but since the end configuration doesn't need it, I don't want to add extra unneeded code. Ideally I would like to optionally dip shunt in say a 0.1 uf cap across the 330pf. The code wouldn't have to change. David Schmidt at home!

David G. Schmidt <spam_OUTdschmidtTakeThisOuTRAIN.ORG> wrote: >I currently only have LED's connected to the output ports but the >'bit twiddling' is too fast for me to check if I have all the timing correct. >I am using an RC osc (100K and 330pf cap). What I want to know is how >slow can I make the osc using RC parts? David: How about just changing your PIC's oscillator-type from "RC" to "XT", then clocking it with an external frequency-generator (which could even be a suitably-debounced pushbutton switch)? -Andy -- Andrew Warren - .....fastfwdKILLspam@spam@ix.netcom.com Fast Forward Engineering, Vista, California

Yeah, I've thought about clocking it externally, but I don't want to reprogram the chip. I was looking for a way to debug a finished circuit if 'bit timing' was in question. My solution (unless this will prematurely destroy the PIC) is to have a three position header that will either parallel a 0.01uf cap across the 330pf with a 10Meg resistor, or when moved on the other two pins, remove the 0.01uf from the circuit and parallel the 10Meg with a 100K. Now it is slow enough for me to watch the LED's without having to reprogram the chip (which would be a problem later with OTP Pics). I'm just wondering if the 0.01uf will exceed the Pics internal transistor's current rating during switching. Dave

Hi Folks, yes you can get them right down to zero. I've used a TTL o/p signal generator as a clock source. It can be VERY useful for spotting bugs. I have used this several times, ot was especially usefull when debugging an interface to an LCD module. Cheers, David dj-eastlKILLspamuwe.ac.uk