Today's Blog posting is a result of an email I received commenting on yesterday's post on introspection.
I am at this point uncertain whether the input was a tongue in cheek comment (maybe even cheeky) or perhaps a serious input. Essentially it said are we done looking at tubes, transistors, loop antennas and perhaps wean myself from 12AU7's and 2N3904's and focus on where the rigs are headed. You know FPGA's, digital signal processing, pre-distortion, SDR and external audio panels using Sennheiser Microphones.
Shazam, I thought I get it, the input is suggesting I come up with a SSB radio that uses nothing but 2N3904's. Yes, someone has done an all 2N2222 CW transceiver (Jim Kortge, K8IQY) and the PSSST came close with 5 out of 7 transistors using a 2N2222A, but how about a transceiver using nothing but 2N3904's.
First and foremost, the 2N3904 is a very common device and really cheap. As a reference, the Jameco 100-piece price is $0.06 apiece. Stay away from the Amazon BoJack variety as likely they are duds... I found that out with the IRF510's I bought from Bozos, (BoJack + Bezos).
The 2N3904 could fit the architecture of the P3ST but the real challenge is the Driver Stage and Final Amplifier. These circuits would have to be developed using only a 2N3904. Keep in mind the Bitx 20 used the 2N3904 in the lower-level stages but stopped short in making it all 2N3904's in the rig. Why even the die-hard TIA aficionados could get their 2N3904 day in court. Luckily mostlydiyrf.com sells the TIA as a kit.
The idea of using a single device in all circuits is something I saw in a CQ article where during WWII, the Germans used the same tube everywhere in a simple field radio. Neatly packed in the case was one spare tube... usable in any socket.
The 2N3904 is a general-purpose NPN transistor with the following specifications:
The above specs for the 2N3904 define why it would be a good choice for the proposed rig.
For those with an itch in your shorts, the audio amp would not use the LM386 or LM380N-8. The BFO and VFO would be fabricated using the 2N3904. I likely would use a crystal switched Super VXO rather than fiddle with a drifty VFO. The VXO and BFO would use but four or five 2N3904's. After all we bought 100 pieces of the 2N3904.
A Sanity Check on expectations. At best the rig could do maybe 1 to 2 watts out and that would come from something other than a single 2N3904. But let's start first with driver stage as that would need to deliver a minimum of 250 milliwatts and an interesting approach might be a push pull driver stage. This would need development.
I remember seeing in the G-QRP SPRAT where a whole bunch of 2N3904's were paralleled to develop greater than 1 watt output. That again would need development. But that sort of effort is what makes this a worthwhile project. I have worked DX running two watts on SSB.
The only thing left is to distinguish between tongue in cheek and cheeky.
Tongue in cheek we know is to say something knowing full well it is likely untrue. Mary Jo is svelte. Cheeky is British slang which is close in meaning but pointed with an added twist of being disrespectful or impolite.
There is no question in my mind that it was tongue in cheek. I only mention this because I was once called out for being cheeky for calling a rig a piece of crap. What the person didn't realize is I use the term CRAP as shorthand for Cool Radio And Project.
Them that know can make things go. (That is sort of cheeky)
73's
Pete N6QW