This unit produces (in the writer's case) a 10 milliwatt signal at 96MHz exactly. It is a crystal oscillator source, phase locked to the MSF standard frequency signal radiated on 60 kHz. This signal is receivable over the UK and no doubt in many parts of Western Europe. The version shown here is the Mark 1 but since the writer constructed this unit, Andy Talbot, G4JNT, designed and published a Mark 2 version which has even greater long and short term frequency stability.
The internal construction
(by G3PHO)
The oscillator uses a 12MHz crystal which is multiplied, in discrete transistor stages, to 96MHz . The 12MHz is also divided down to 60kHz and locked to the MSF transmission (received via a simple bipolar head amplifier with a ferrite rod antenna). The phased locked output at 96MHz then drives a choice of diode multipliers to give outputs in the various microwave bands of interest. A multiplication factor of 108 produces a weak but usable signal on 10.368GHz for example, using a simple varactordiode multiplier in waveguide. Even without an external multiplier, the unit produces strong harmonics well into the UHF bands.
The unit can easily be adjusted for output at 108MHz instead of 96MHz. Multiplying by 96 produces a signal at 10.368GHz again.
The unit shown above works well and is a useful tool. The writer has the Mark 2 version on his intended construction list though, as a number of shortcomings have become apparent with the Mark 1:
The time loop constants are too short to ensure a really stable signal when the output is multiplied up to 10GHz since the signal wanders over a few hundred Hertzs, over a few seconds, particularly at night when the 60Hz MSF signal has QSB on it!
A 10MHz output would be useful so as to drive the reference for a counter or synthesiser. The Mark 2 unit is therefore based around a 10MHz oscillator.
These problems have been addressed in the Mark 2 version and the Mark 1 can be upgraded very easily.
Full constructional details of the Mark 1 unit can be found in the RSGB RadCom Magazine April 1994 (with errata fixes in August 1994) as well as additional constructional notes in the May 1996 and October 1994 issues of the RSGB Microwave Newsletter.
The Mark 2 version is fully described in the April 1996 RSGB Microwave Newsletter.
Other methods of obtaining a highly accurate frequency source are possible of course and there have been several articles on using TV sync pulses to phase-lock a local crystal oscillator. In the UK, the BBC TV signals are locked to a rubidium standard I believe (WARNING!...dont use ITV!) and many overseas countries have similar standards. Look for a copy of Electronics Australia, October 1993, for a most useful constructional article on this theme.
RETURN TO SIMPLE TEST GEAR INTRODUCTION