question about pixie transceiver
The transceiver shown on this site: http://homepage.ntlworld.com/lapthorn/pixie.htm
supposedly works OK. So I am wondering why I have been having so much difficulty with the audio circuit and with frequency pulling, with the one I am building of my own design, but which started out before modifications to be about the same as the pixie. Can anybody shed some light on that, please?
Mine is shown on: http://colpittsoscmxr.net63.net/
George
The circuit you copied has a crystal for selecting the frequency, you are using an LC circuit. It is difficult to get a crystal to oscillate at anything other than its resonant frequency - they are pretty difficult to "pull" very far. An LC resonant circuit is very easy to shift the frequency. A small change in the capacitance due to transistor loading or changing of its biasing or supply voltage will cause it to wander all over the place. It is tricky to design and build VCOs which are stable for RF transmitters. You could try decreasing the inductance and therefore increasing the capacitance. Reduce the capacitor to the base of the transistor. Use a regulated power supply. Add a buffer (the original circuit has one - you don't seem to).
Keith.
Thanks Keith, I added a buffer earlier this week. If I have any more problems with th frequency I'll change the capacitor on the base (I presume you mean on the buffer input?). The buffer seemed to fix the frequency shift problem. I was just wondering why I had to use so much buffering, but I guess you cleared that up. I should have known thats what it was. I also had to add input and output amps. I was wondering about that also.
Thank you,
George
The reason for needing the buffer is that any load changes will get reflected back through the collector/base and collector/emitter capacitance. Also, as I said, the loading changes can alter the biasing of the transistor which alters the transistor capacitances.
Some oscillator designs are a bit more immune to load pull but in your case the tuned circuit is directly connected to the output.
Keith