Andrew:

Why not do a firmware DPLL on the incoming data to recover the clock.
Generically, it goes like this:

1. Have a "time" reference at the same rate as the data; that is known.
2. Interrupt on the "edge" of the incoming data.
3. Note the difference between your edge (in time) and the edge of your
known internal time reference.
4. "Correct" time by adding or decreasing NOPs in the service loop. The #
of NOPS you add or take away is a sort of "filter."

This will establish a timing chain, sampling point, if you will.

Sorry I don't have acutal code, but have done this before and that is my
recall.

Phil
Alden McDuffie Consulting.
philip_n_anderson@hotmail.com









On Wednesday, September 13, 2000 4:53 AM, Andrew Seddon
[SMTP:seddona@HOTMAIL.COM] wrote:
> Hi..
>
> I am currently using a radio module in my latest PIC project.
(RADIOMETRIX) BIM if anybody is interested.. However I have a natural
aversion to paying for things I think I can build myself, i.e their Radio
Packet Controller. As this is basically a PIC I want to see if I can make
something similar myself.
>
> If anybody is interested the datasheet for this part is at address
http://www.radiometrix.co.uk/products/rpcsheet.htm
> if you are doing anything to do with radio and you need some concepts
about clock recovery etc explaining simply it`s pretty good, however
obviously there are no technical details as they want to make money selling
it!
>
> Anyway down to my question. Basically I am transmitting as Manchester
encoded data. What I need to do at the receiving end is recover the clock
so that I know where to sample from. This is exactly what the radiometrix
module does in software. However that looks a little bit beyond my skills
at present so I am looking for some suggestions. I have been trying to put
a PLL circuit together with a 4046 chip but I am having trouble selecting
the correct resistors/capacitors. Help greatly appreciated in that area, I
need to lock to a 2khz edge so I figured with 16x oversamping I need a
centre point of 32khz and a fairly narrow loop filter bandwidth.
>
> Alternatively if somebody could point me in the direction of some
existing code that is similar I could attempt to do it in software.
>
>
> This is common stuff and I am sure that other people will need to do it
so if we can get a solution together I will gladly put up a webpage,
circuit diagrams, layouts etc.
>
> Thanks in advance for any help given.
>
> --
> http://www.piclist.com hint: To leave the PICList
> mailto:piclist-unsubscribe-request@mitvma.mit.edu

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu