I am working on a novel design for a 70MHz transverter. It uses a diode ring mixer as a bilateral parametric up-downconverter. This unusual method has some advantages. For example I think that I can build a complete transverter with just three transistors. These are configured as a charge pump (aka high power crystal controlled oscillator). The Pump drives the diode ring in such a way that it makes the diodes work as varactors. This allows the mixer to work as a parametric amplifier. This reduces loss and improves noise figure. This is critical as the mixer is used for both up and down conversion. Another plus of this odd system is that no antenna change over is required due to the bilaterallity of the mixer. I have only seen this done once before. So far I have the charge pump working. This uses a cheap 20MHz crystal which will allow me to transvert from 50MHz.
In reply to G3CWI:
Takes me back to my University days - my final year project was a parametric amplifier for 23cms pumped with a 10GHz oscillator. Price of varactors took your breath away (thank you Mullard for the samples!). Hopefully you don’t get all the problems I had trying to get it to amplify, I think you may find the pump frequency is too low - time will tell
All understood. I agree about the pump frequency but the only write up of this technique that I have found, is vague to say the least about how it worked. I some ways it is more interesting for what it does not say than for what it does! I suspect that the designer was not really sure how it worked…
In the (many) years since the original article, much has changed so it will be interesting to see what can be done with more modern techniques.
Had they finished teaching phlogiston theory by then?
In reply to G3CWI:
Richard, all sounds very interesting. Whilst you’re beavering away can you work on knocking up a time machine. 48hrs in day would help me no end. As would being able to travel back 30 years and start hillwalking then rather than just a few years back.
Some more progress. Soak tested the charge pump for a couple of hours and it didn’t blow up. Built the mixer section using broadband trifilar transformers instead of the narrowband tuned system in the article. Connected everything together and heard a weak Ssb signal on 70.200. Not so bad with dipole across shack. I have transmitted on it and a signal emerges. Not measured much yet but it looks hopeful. The resistor that sets the parametric gain does not seem to do very much. So more experiments required.
I am beginning to think that this design is flawed in some way. While it works (sort of) I had been unable to get anything like the output that the originator acheived. The best I have done is about 15mW - and that is with 1 Watt of oscillator input and 500mW of drive. I am about to concede failure and revert to a standard transverter.
One of the things I seem to remember was that it was essential to use a circulator to isolate the pump from the wildly varying reactance of the varactor and to keep the amplified signal from being swamped by the pump circuitry. How are you keeping the different signals apart?
Mind you the pump was a Klystron.
Andy - you mean the phlogiston theory is dead? Next thing you will tell me is that radio signals don’t need the aether to propagate. Methinks you need to take more water in it!
I now have my doubts about the design. I have built three variants and none works - nor do they show any signs of working. I have corresponded with the writer who said he had been disappointed by the lack of take-up - maybe no-one else can get them going either.
In reply to G8ADD:
We should activate a few mountains together and watch those snails zoom by. Happens to me all the time! [mutter bloody Moel y Golfa under breath]
David M0YDH
Moel y Golfa GW/MW-026 - yes! Hard work whichever way you go up. It seems to rise up almost out of nothing - then keep going up, and up, and up! I well recall finding it surprisingly difficult on Halloween 2003, when Jimmy was 11 and Liam was 6!