[mc1322x] suggestions for code errata

[Mariano Alvira]

On Fri, Oct 09, 2009 at 01:17:46PM -0400, Peter Lawrence wrote:
> 
> These trim values seem to be hardware dependent, and the reference
> manual does not establish a procedure for determining the correct
> values.  (Presumable, it involves going in a continuous transmit test
> mode and measuring the frequency of the RF waveform with calibrated
> equipment.)
> 
> However, the RM suggests 0x15 coarse and 0x10 fine as starting values.
>  Using those, I experience much better reception of packets.
> 
> So, if you experience the same problems, it may be worth trying to
> write XTAL_CNTL with, say, the value of 0x28E0F52 to see if reception
> improves.

Hello all, 

I have come up with a test procedure that has allowed me to determine
the proper values for XTAL_CNTL. To make this measurement I pigtailed
a piece of coax (RG174 in this case) in series with 1M Ohm to the
24MHz xtal_in, pin 49, on the MC13224v. This allows me to measure the
frequency of the main reference oscillator using an HP 8753C VNA to
within a few Hz. I verified that the cable capacitance did not
substantially alter the operating frequency (by repeating the
measurement with an additional 3pF on the cable).

I've updated rftest-tx.c and rftest-rx.c with the new values.

For the Redbee module XTAL_CNTL should be:

   reg(XTAL_CNTL) = 
   (0x1 << 25) | (0x3 << 21) | ( 0x6 << 16) | (0x1F << 8) | 0x52;

For the CEL modules, their docs say coarse tune should be 4 and fine
tune should be 14.

-Mar.