View Full Version : maximum oscillator trace length

- 5th April 2005, 00:41
I am working on a project using picbasic pro and have a question about the maximum trace length going to the oscillator. I am using a 18f252 that will be on a board with a display above it to save space. It is running at 4mhz currently. Everywhere I have looked I see "oscillator should be as close to the pic as possible". I may have to move mine as it will not fit under the display properly. Is there a more defininte answer to its maximum distance from the pic? Help is greatly appreciated, thanks.

- 5th April 2005, 00:45
never tested the maximum distance but 1-2 inch can be safe. can you solder your crystal or ceramic resonator under your PCB directly on your PIC pins???

- 5th April 2005, 00:51
Steve, you must be on here all the time for me to get an answer that fast!!!

I can put it on the bottom, but I have no components on the bottom now. I think I have room around the display to move it, and still be within 2 inches. I was just curious to what other people may have gotten away with. Thanks again for your quick response!


- 5th April 2005, 03:00
You're welcome.

Steve, you must be on here all the time for me to get an answer that fast!!!

I was in the past before i answer to too many redundant questions and lazy user that don't want to do some forum search and/or home work.


- 5th April 2005, 13:18

it also depends on what type of clock source you are using,
Resonator, Crystal, Oscillator (w. TTL Output)

Oscillators are less critical than crystals with external Caps.

If traces get too long you might have problems for the OSC to start up.

I personally would never move a crystal 2 inch away from the MCU.

If it is just the "height" of the crystal causeing the problem:
a Crystal in HC49U-S package is less in height than a PIC in a DIP package,
so you might be able to fit one of these.
SMD crystals are even smaller.

- 5th April 2005, 15:47
Hello Ralph,

Thanks for the response. I am now about 1 inch away so I think all is good. I appreciate the response.


- 5th April 2005, 17:25
Don't forget also that long leads on oscillators also produce heaps of radiated signals... your product might fail CE certification if it's a commercial product.

- 5th April 2005, 23:11
Thanks Melanie for your input. I will not be going through any approvals so I think I am OK. Many thanks to all giving their opinion.


Ron Marcus
- 20th May 2005, 03:16
For really small crystals, the Epson mini barrel crystals can't be beat. If you can use a resonator, they already have built in caps in some tiny SMT packages.