I have three chips which could run off a 13Mhz clock.

At least two of the three can be driven by a reference frequency. Meaning they don't have to drive a crystal they can be hooked up to an oscillation directly.


I could use one chip to drive the crystal and then i was thinking i could buffer the signal through an inverter or basic emitter follower bjt circuit to the other two chips.

Questions:
1) I know this will work but is it an okay design practice. One problem you will run into is that the crystal will be further away from the chips. I have a radio chip, an external demodulator chip, and a microcontroller. I was thinking if the crystal was close to the radio chip the PIC would probably wouldn't be affected by the crystal traces being a little longer.


2) Anything else that would be affected? Perhaps it'd be more susceptible to temperature?

Thanks,
~blake

 

It should be fine, it's quite often done in commercial equipment.

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk

 

You could also use a clock oscillator module for this application. They often look like a little metal box, a little larger than an IC chip (they make plastic ones too) with the output freq printed on the top. You just hook it up to your +5VDC and ground on the appropriate pins and it puts out the square wave on its's output pin (they also make sine wave ones). They're usually pretty darn accurate with respect to frequency error with no tuning required. I've used them many times and I've had real good results with them.

 

13 mHz is not high enuf frequency that you should be concerned about a few inches of trace length.

__________________
see my website: www.geocities.com/russlk