You ask why did you turn on the iron when ham radio is not about building things. It is all about contests, operating and watching You Tube Videos of YL's making POTA contacts.
Just because is my answer as I want to explore circuits that may be obscure or not very often used. It is like a technical challenge thing for me. Maybe the challenge for most hams today is staying awake during contests and operating.
But for me right now.... I want to revisit using PNP Transistors in a CW transmitter. Yes, Sony has identified that NPN transistors are better for their products. But unlike the ARRL and their proposed CSI grading system, PNP transistors can work in RF circuits without fear of getting flat feet, jock itch, dandruff, toe fungus or an STD. The key to this project is no cost and use what is in the junk box. Brent Crude today > $113 a barrel.
Firstly, I should mention I have a box of 100 Russian Germanium PNP transistors that I bought years ago at a cost of 20 cents each. These are good in the UHF range and found it can produce 0.8 Volts PTP at 14 MHz, in a crystal oscillator circuit. That my friend that is 1.6 milliwatts. So, it will need a big boost to put out at least 100 milliwatts (about 20dBm) or 18 dB gain from 1.6 milliwatts. We are using what we have in the junk box.
My first hurdle was a Google literature search for PNP Oscillator circuits and that turned up lots of sample circuits... all NPN which of course was not useful. The same type of results with Copilot. Then I thought about my library shelf. Forget EMRFD but struck Gold with the Transistor Radio Handbook by Stoner and Earnshaw.
I went back to Copilot and said did you consider the Stoner/Earnshaw publication. Copilot's reply was Pete you struck Gold. It then spit out a circuit using either a 2N2907 or 2N3906. Of course, I want to use the Russian Germanium Transistor, so I have a starting place.
Back of Envelope Schematic... It Bombed!
Here is an interesting story about the Russian Germanium Transistors. I had the scope on the output of a test circuit and pulled the crystal and found out I still had a scope trace. The built-in counter read 234 MHz. [kB=1] It was either self-oscillating or a parasitic oscillation. So indeed, in the UHF range.
I had previously built a PNP Driver Stage that now looks like this.
I am still working on the oscillator circuit itself as I really want about 10 milliwatts from the oscillator stage as that would get me about 300 Milliwatts out of the Driver. Now we are cooking!
Looking through the junk box, I found an NPN Oscillator board and with a bit of jury rigging have it working with the Russian PNP Transistor and an earlier version PNP Driver using a 20M QRP Crystal. The Output was only 50 milliwatts. Thus, more work at boosting the output of the oscillator.
Ultimately, I want it to work to develop several hundred milliwatts using PNP devices. We now have the Copilot suggested Colpitts topology and so we will begin building that circuit.
Well, that Colpitts circuit did not do as well as the junk box NPN adapted to PNP. I swapped out some bias resistors and with a bit of good luck now get 8.4 Volts PTP out of the Driver at 14.06 MHz and that translates to 176 Milliwatts. If I could get 9 Volts PTP that is 200 Milliwatts [9X9X2.5 = 202.5]. There is also a change in that now there are 2N2905's in all three sockets: Oscillator, Driver and Keying Transistor.
At the Oscillator, with some resistance changes I now get 1.24 Volts PTP with the 2N2905. That is 3.84 Milliwatts. Thus, our Driver stage is producing 16.6 dB of gain. [10*log (176/3.84) = 16.6] LT Spice calculates the Driver stage gain at > 18dB.
I also earlier found a P MOSFET equivalent to the IRF510 and so a possible pair of "boots" for the basic PNP circuit. OTs know a pair of boots is an outboard amplifier for boosting your signal.
For me this is more fun than yelling CQ POTA, 5X9, Park 12345. But for the 99% majority of hams the 5X9, Park 12345 is the soup du jour.
Them that know, survive. Stay Tuned!
73's
Pete N6QW
