Since I use a TCU-30 on my station, I can get some of the values of the discrete components, but the toroids are the issue. Those, both assumed to be ferrite type would have to be measured.
The info might be around, someone on one of the groups was trying to recreate a version of the coupler for their own station and had gotten some of the parts.
The LPB is the 'Cadillac' of coupling systems really, but you can also build one out of separate parts, like any antenna tuner has.
The theory isn't too complex. The main factors that a power line coupler handles are inductance correction, using series capacitance to correct the high inductive reactance a power line usually has, and impedance transformation, taking the 50 ohm transmitter output and feeding the usually lower impedance of a power line, down to a few ohms or less.
For theory, take a look at Part15 Engineer's document dump here and look for the one marked Carrier Current Techniques. In a later section of the book, Ernie Wilson talks about coupling, and has a couple of schematics for it.
Now these are for the ATU-30, used in LPB Traveler's Info Service installs, but it's likely to be the same circuit as the TCU, just re purposed.
Started sourcing parts. Sorted through the junk box for fuse holders, barrier strips, a real Amphenol 083-1R SO-239 for the input, a microamp meter. Have a solid lead on 5 deck, 4 four position switches.
Converting the drawing to DXF as well so there is a clean copy
It seems pretty complex with the selector switches they use, and the matrix they use for the capacitors, so that the 4 capacitors add up over 9 switch positions.
I thought of an idea to make it easier to find parts and cheaper, why not convert the capacitor selection to binary, using simple on-off slide switches? It would be 8 switches total, 4 for Coarse, 4 Fine, allowing you to switch any combination of capacitors into the circuit for the inductance correction.