January 4th, 2024. SSB Transceiver Design Notes
As posted over the last month, I have been spending a good deal of time in the local ER with the XYL. What a place to see a spectrum of misery. There are long stretches of time where it appears nothing is happening and you just sit there thinking of 10,000 things you would rather be doing. Raking leaves and cleaning the garage sounds pretty good right about now!
I cannot have an idle mind so it is during these visits that I have been pondering ideas for new SSB transceiver projects. That seems to be a bit of an answer to the time warp, but of course the best answer is to not be there at all.
Some things are pretty evident from my past designs that any new Transceiver must be Digital for the LO Generation. But now we have some choices beyond an Arduino Uno (Think of a Model T Ford) and I have coming some Seeed RP2040 boards (Think of a Ferrari). The RP2040 form factor is really small and will lend itself to compact designs (Think Portable Rigs).
But it is not just the small size (This may be the exception to the size matters urban legend) but what is under the hood is a 133 MHz dual core ARM Micro-controller Unit. Think fast program execution and enough clout to do the exotic math functions like Hilbert Transforms and FFT. The Program storage is also significantly larger.
The price is also amazing -- under $6 from Digi-Key. Now if this was a unit offered from AliKazam in China -- don't buy it, but we are talking a US Distributor.
The architecture floats strongly to the Module Approach. (Just like Ten Tec did in many of their radios.) The advantages are many fold such as troubleshooting / maintenance and significantly experimentation with new features and functionalities. Just swap out a Module and you are there.
The Steered Module Approach has proven to work like a charm and we see that further applied in the kit offered by Mostly DIY RF.com. I sure hope you took advantage of Todd's (K7TFC), introductory sale price which I believe ended yesterday.
Above is an example of a Steered Module that in Receive the Module is the Receiver RF amplifier that gets inserted between the rig Low Pass Filter and the rig Band Pass Filter. So, for the purists you are not amplifying the world but a smaller part of the spectrum. In the Transmit mode, the stage is inserted following the Band Pass Filter and ahead of the Transmitter RF Amplification Stages and finally through the Low Pass Filter.
The next item in the list is the device used in the block diagram above and that is the BFR106, literally a DC to 5 GHZ device costing less than 50 cents. Forget the 2N3904 or 2N2222A when you can have this jewel. I have a working transceiver that uses two BFR106's in the IF Module and one as shown above.
For those Cool Guys who belong to the G-QRP Club in the latest SPRAT #197, you will see a superb article penned by N6QW on how to apply the BFR106 in a new RF Amp stage added to an existing transceiver. The piece also includes the LT Spice simulation. At 12VDC think 17 dB gain out to 6 Meters with only a handful of resistors and capacitors.
For those who build nothing and nitpick everything the drop off at 50MHz is only 0.2 dB.
I chose the BFR106 for this piece as it also demonstrates another facet of new designs and that is Surface Mounted parts. Just like those who only use Analog VFO's through hole parts are giving way to surface mount. So instead of rationalizing why you can't do surface mount, man up and move on to SMD.
You are in luck though, as Todd K7TFC sells the BFR106 mounted on a circuit board that you can use just like a through hole part.
Quit winding toroids for Double Balanced Mixers (DBM) and just move on to the SMD ADE-1. Unless you exercise great care in how you wind the toroids and how you build the DBM with the layout and the use of quality parts your homebrew DBM will suck! Again you can purchase mounted ADE-1's from Todd.
BTW, I did use a BFR106 as a Microphone amp stage and that was not a great idea -- It worked really well BUT too much gain at 12 VDC. You hardly cracked the gain pot and you were over-driving the radio. I succumbed to slipping a SMD 2N3904 in its place and now you have to have the gain pot full open before you flattop.
The audio amp stage is something often is an afterthought -- slap in an LM386 and you have audio. Lately I have been using a 2N3904 driving an LM380N-8 which gives more and better quality audio.
I have also recently tried an LM383 where if you also want to play you American Made Fender Stratocaster through your rig, you are good to go with 5 watts. Using the higher wattage amps at low levels appears to my somewhat diminished hearing to just sound better.
Circling back to the Digital LO with the Seeed RP2040 the minimum device pin count limits you to LCD's (4 Pins and shared with the Si5351). But that is not all bad as you could drive a 20X4 LCD and have plenty of room to even include an S Meter and other pertinent info.
Now there are those who will shout Port Expander and that certainly would give more pins but the LO Physical size might grow a bit. Were I to expand the ports then I would go for this display which is less than 1 Inch Square.
Lots of time and lots to think about. The Final RF Amp stage where so often the IRF510 is used, needs a lot more thought. The Mitsubishi RD06 and RD16 RF devices are End Of Life and now obsolete. Mitsubishi has some suggested replacements but so far the greater ham community seems slightly unimpressed with the offerings.
Looks like a bit more Hospital time and then back to the B&C for the OL. Thus I have a bit more time to think about that final RF amp stage.
Just in passing, for the IF frequency 4.9152 MHz seems like a good choice (as used in the Elecraft K2) as then 17M is not an issue as would be with a 9 MHz IF.
73's
Pete N6QW
