|The Nutty Professor is at it again ! ! !|
As many who experiment with home construction would no doubt relate, one thing quite often leads to another, which in turn leads to another, and so on (or to misquote a line from the "Commitments" movie - they go "spiralling off").
It's a wonderful thing on the whole, but one of the frustrations with it is that one never seems to get anything quite finished!
Readers of earlier blog entries will have spotted references to crystal controlled transmitters and receivers for use with the QRSS aspect of the amateur radio hobby.
Successful (one hesitates to say "the best" as this implies some sort of competition, and I don't think QRSS folk are particularly competitive) QRSS stations, whether simple or complex have to be well engineered.
Sloppy engineering practice soon becomes evident, manifesting itself in unreliability or in poor performance, and the poor performance I am thinking of in particular is that of frequency stability, or rather the lack of it.
My own home made efforts have exhibited acceptable, if not outstanding stability, and one quickly looks to ways of improving performance.
For some time now I have been pondering this issue, and a chance visit to a rather interesting web site produced a nicely written article explaining how to make a simple crystal heater
Thermostatic control of the temperature of the crystal used to produce the QRSS signal, or local oscillator in the case of a receiver seems to be a fairly obvious thing to do, and I had already looked at a commercial offering which offers high stability for a few pounds, and is a microcontroller based solution.
The heater to which I am referring though is a much simpler affair using a darlington transistor to modulate the flow of current through some low value resistors (the heating elements), with a thermistor to provide regulation and feedback.
This struck me as a brilliantly simple idea, so I resolved to build something similar if not quite identical.
With the exception of the thermistor, all parts came out of my junk box, in fact the darlington transistor was concocted from a small power transistor rescued from some long-forgotten project (the leads had been cut and had traces of old solder on them) and a low power Japanese 2SC something-or-other freshly removed from a scrap board which was probably a transistor radio or portable TV 20-odd years ago. Total cost of parts, probably quite a bit less than a pound.
The crystal being thermally controlled (heated) is physically but not electrically connected to a heated piece of thin tinplate (formerly part of a screening can also rescued from a scrap board).
When connected up to a power supply, my little heater worked first time (it is so simple it could hardly fail), and after some experimentation with the settings I had it heated up to just under 40 deg C, and was able to adjust the crystal FREQUENCY (the crystal was electrically connected to a completely separate test oscillator) by tweaking the temperature of the heater. The heater once it reached equilibrium maintained its set temperature perfectly.
My test set up involved monitoring the current drawn by the circuit, and the collector voltage of the transistor.
|Pretty it ain't - interesting it certainly is!|
One quickly realises that this little heater is really trying in its own small way to regulate the temperature of the
whole planet, and for it to be as economical as possible, it needs to be thermally insulated as much as possible from the outside world, and a little more thought reveals that in a way this little circuit is a miniature representation of, say, a domestic dwelling where the occupants try to maintain a comfortable temperature inside when the weather may be up to all sorts of tricks on the outside.
|The heater schematic|
Also, as I have an interest in the application of microcontrollers, this application is also, as stated above, one where a microcontroller could be used, but this is where one needs to think about cost. Even buying new components, the unit cost of my little device would be no more than a pound, and the cheapest microcontrollers cost in the region of 75 pence per unit, so by the time you've bolted any additional components required, the simple version probably wins out, but only just. Die hard microcontroller enthusiasts would probably dispute that assertion!
Either way the mere consideration of one approach versus the other illustrates perfectly how experimentation can lead one through all sorts of interesting thought processes, which is really the whole point of me writing this note.