FREQUENCY STANDARD LOCKED TO NORMAL TV SYNC
At the Norbreck Hotel Blackpool NARSA 2003 Rally I was asked by a visitor to our NEW FYLDE ATV stand could I make a 10MHz Frequency Standard that locks onto normal TV sync that in turn is atomically locked. This has been checked out and is the case..
It would seem that the normal TV syncs are very accurately locked onto a stable source such as Rubidium vapour cell and also the GPS. Because of the high accuracy even two TV transmitters on the same frequency don't beat with each other. If you are in an area that gets both.
Locking onto RUGBY MSF 60KHz has it's problems of phase shift early morning and late evening. Most old timers like me will remember listening to RADIO LUXEMBURG and hearing the phasing in and out. It does even happen on lower frequencies as well.
Well I sat down and started to map out what was required and then came up with the problem of the frame equalizing pulses how to sort them out. These would upset the Phase Comparator
After a few hours of designing it was put to the MACH chip to do all the logic and filters and phase lock loop drives..
So after doing a bit of simulations it seemed to in theory work..
I did a PCB layout and built the unit up and had to do a little bit of design to make the oscillator work correctly with CMOS gates that are in the MACH device but soon got it stable and the XTAL 10MHz not being overdriven.
G3RFL OFF-AIR PROTOTYPE PCB MK1
The next step was to add test points at various stages of the design to check that everything was functioning correctly...just a few tweaks to the MACH logic chip and we had both Reference and Delay counters counting. Also the delay counter was triggering on the main line sync pulses from the LM1881 sync separator. A simple diode seems to move the XTAL over a range of 20Hz so it's just a simple job of selecting stable components to get the XTAL near to 10MHz at first it was running high by about 60Hz but that is the side I want it on running a little high.
The Phase Comparator is working at 15,625Hz (line frequency).
After a lot of adjusting and checking I was not happy with the 10MHz output signal not being a clean square wave but had lots of ringing this was traced to the fact that the main 10MHz oscillator is working in an analogue mode and seemed to reflect on all outputs also a little pulling of the oscillator was noticed when terminating the external 10MHz signal.
It was decided to develop a FET oscillator and FET buffer external to the MACH and screen it so a small can will be made to buy a VCO would have cost £25 trade then you have to rely on a design that needs special parts and wonder are they always available. Also this will keep the power, and thus the stability, to the XTAL low. Also putting the OSC in a can will stop sudden thermal drift problems with air movement.
G3RFL PROTOTYPE 10MHz VOLTAGE TUNED VXO
FET OSCILLATOR PCB
The New FET oscillator worked first time and gave enough output to drive the MACH clock line input but it needs about 9v supply so it will need some extra bits on the main PCB. I ran it all night and was within 1Hz free running from going to bed when the room was hotter and checking this morning before the central heating had got up to temperature. Also I noticed good isolation on the output I used two BF244B FET's. G3RFL
G3RFL PROTOTYPE 10MHz FREQUENCY STANDARD PCB
Will all be on one PCB soon!
Draws Just 95m/A @12Volt