An accurate and stable reference frequency (part 2)

After the disappointing results with the 153 kHz in the new town, is was time for some investigation.

It was obvious that the 153 kHz should left as an external reference. The first attempt was not to start over again but use as much as possible from the existing design. With this restriction it was clear that an other station on the VLF had to be used. BBC4/Droitwich has a very strong signal at 198 kHz. This station is many dB's louder as the one at 153 kHz.

A computer program was written to figure out the possible frequency combinations for LO and IF to make a fast test possible. It showed that only the IF of 3 kHz had to be changed to 48 kHz and afterwards divide the IF signal by 16 to get 3 kHz for the phase comparator. With retuning the receiver this was a small operation.

The first results were hopeful. But after a longer observation time some big phase jumps in the 3 kHz signal were observed. These jumps are as one could be expect related to man made noise. A new attempt was made by changing the phase comparator from 3 kHz to 48 kHz. This leaves out the digitized phase error due to the 16 times divider. When a phase error occurred due to an noise peak the phase error stayed at that value and could only adjusted by changing the TCXO phase/frequency.

The results of the the phase comparator running at 48 kHz and retuning the 48 kHz filter were satisfying. But some optimalisation work has to be done.

First a list of optimalisations was made. In the old design some solutions were not optimal.

Smaller IF filter
No limiting before the IF filter

A smaller IF filter could only realized with the use of a crystal. Unfortianal such a crystal was not in my junk box at that time. So a new program was written to calculate all the useful IF frequency's. This resulted in a list of useful frequency's.
At that time a small OEM GPS receiver was available for a small price. With the idea of having a more accurate reference frequency compared to a VLF station the whole idea of locking to a VLF station was left.

The results of the GPS controlled reference can be found in part 3.

Back to PE1CQQ Page.

(c) PE1CQQ Last update 22-2-2007