As a follow on to yesterday's SSB transceiver resurrection here are some details on a couple of circuits from that era of my radio fabrication.
Firstly, I always build in Modules which has several really desirable features. The big plus is troubleshooting, as you can easily isolate a module that has a hiccup. The second, almost as big as the first, is experimentation and upgrading.
The first circuit I will soon describe is straight from the LBS (Lets Build Something) project which I co-authored with Ben, AI6YR. The whole thrust of that project was to build a working SSB transceiver that could be undertaken by those new to the hobby. Using readily common parts and a with simplified design, it was aimed at encouraging homebrewing.
The LBS which appeared in QRP Quarterly (10 years ago) was undertaken by a radio club in Canada and one of their radios won a prize at FDIM (Four Days In May at Dayton).
There is a subtlety to this design as the output is always terminated in 50 ohms using a 100 Ohm resistor and a 100 Ohm trim pot in parallel. The output is taken from the wiper on the pot. This provides a means of optimizing the stage gain to the next stage while keeping 50 Ohms as a termination for the amplifier stage. A pair of these are used as the Rx RF Amp stage and Transmit Pre-Driver.
Relays controlled by the TR selects the correct amp for receiving or transmitting.
I am not a fan of homebrew Double Balanced Mixers (DBM) and the reasons clearly come from anything that gets published by me will likely get undertaken by those new to homebrewing. Often the new builder neglects some key elements that are needed to have a successful DBM. Then the trouble starts.
The first issue is the diodes and the need that they be matched. There are published videos by some notable hams that conclude you do not need to match diodes. I disagree and you do need matched diodes! The name itself Double Balanced Mixer connotes balance and that starts with closely matched parts.
Taking a similar example if you are homebrewing a crystal filter you start by finding a number of crystals that are extremely close in frequency - no successful filter is built by having all of the crystals hundreds of hertz apart in frequency.
Still another parallel argument for matched diodes is the audio phase shift networks using op-amps. The resistors must be precisely matched to the specified values. Even 0.1% tolerance resistors will not give a precise 90 Degree shift. Using 5% resistors puts you out in left field!
The next piece is the physical construction where often parts are tack soldered on glue down Manhattan pads that have no symmetry or balance to the layout. Crossovers and signal leakage are the issues. Double Balanced Mixers result when all aspects are in Balance.
Finally, the construction of the Ferrite cores themselves requires a triple trifilar winding that should be evenly spaced around the core. To assure a coupling coefficient of 1, I twist the three wires with about 8 twists per inch. Multi-color wire keeps the wiring straight.
For a little icing on the cake, I include a balance pot and a means (39 Ohm Resistor) to unbalance the DBM for tune up purposes.
.jpg)
Above is an Alternate Layout to achieve symmetry and minimize cross over wiring. If you look closely at the photo of the DBM build you will see the Red and Green Wires. This alternative arrangement will facilitate shorter connections to the matched diodes.
Again, I dislike the homebrew DBM's not as a component in itself but the fact that new homebrewer's have difficulty building a high-quality unit and then I get emails about your circuit sucks.
If after homebrewing a DBM you find the carrier suppression sucks then look at matched diodes, physical layout and evenly spaced wiring on your cores using twisted wires.
There is a low pass filter so that only audio frequencies pass into or out from the 2nd Ferrite Core. On the side that says AF and Audio In/Out simply connect the output from the Microphone Amp and the Input to the Audio Amp at this point. Only one of these circuits are "ON" at any time thus they can be paralleled.
The audio filter serves a very important function on transmit to limit the audio frequencies that ultimately gets impressed on your crystal filter. The 2.7kHz filter means just that -- anything above that falls outside the filter pass band.
Next is the IF Module which uses the Plessey Amp circuit. Matching transformers are a must and for those that insist on building their own crystal filters it is important you know the Z in/out of your filter.
If you follow the directions of the various filter software packages you will derive that number and then you need to build matching transformers to match to 50 Ohms for the Plessey amp. The Matching transformers also impact band pass ripple of the homebrew filter -- don't leave them out.
Undoubtedly this is where the excitement builds to try out your $30 Nano VNA. That same matching transformer will be needed for your Nano VNA measurements. For a simple homebrew ladder filter typically, the impedance is usually not less than 100 Ohms nor more than 300 Ohms.
The transformers in the schematic below are 10 to 1 as it so happens the Filter is a Commercial Filter with a Z in/out of 500 Ohms. 19^2 =361 and 6^2 = 36. Thus 361/36 = 10:1. You can have two separate windings of 19 and 6 turns or a solenoid type winding of 19 turns tapped at the 6th turn.
Have fun and enjoy.
73's
Pete N6QW