-----Original Message-----
From: Johan Bodin <[email protected]>
To: [email protected] <[email protected]>
Date: 19 January 2002 00:34
Subject: Re: LF: G0MRF/SM6LKM DDS VFO and Jason

I think it's time for you to replace the batteries in your calculator.

That's neat! But I make it 0.0083333 Hz resolution,

I make it 0.00083333 Hz = 10.0000 mHz / 12.
Both are submultiples of 1Hz...

OOPS - Can't read exponential notation on a calculator properly.

The 4.194304MHz GPS - frequency locked loop seems to be going now.  Once
locked, it 'blips' within 1Hz at fundamental and in the long term stays spot
on.  Dividing down, at 262144Hz the mean phase stays constant but blipping
every second over abot 15 degrees.  So if the fundamental is used to drive a
DDS at 137kHz looks like better than 10 deg short term stability and GPS
standard in the long term.  Ideal for 30s BPSK but not much much good for
higher frequency work, 1Hz sidebands ought to be noticable as well.   Chip
count has gone up to five now unfortunately, One for the oscillator , two for
the binary counter, an 8 bit latch and PIC.  But all cheap, the unit does lock
up within a few minutes of turn on and does not need a high stability VCXO.

Just to see how it worked, I also knocked up a conventional PLL using divide
by 2^22 and 74HC4046 PLL chip.  Did manage to get it to lock to GPS 1pps
pulses eventually, but only after forcing the two signals into sync manually
then precharging the filter capacitor to give nearly the correct frequency.
That way it could be persuaded to lock up in 20 minutes.   If started from
cold and asked to lock itself, I estimate would take something like 5 - 8
hours !

Andy


:-)

Thanks for the source code. It explains everything.

It's late, CU tomorrow.

73
Johan SM6LKM