This little connector is sort of a strange anomaly. It was included in the GF and SCR-**-183/283 transmitter sets back in the 1930s and early 1940s as a way of connecting a test meter for maintenance and adjustment, but also persisted through the entire command set series (ATA, early SCR-274N, and AN/ARC-5 (both HF and VHF) (primarily but not limited to a way of provisioning a remote antenna current meter (see I-71 remote meter.)) Unlike the more robust early connectors which used the standard size banana pin approach, it has a relatively cheap "table radio" quality feel to it, almost as if it was an afterthought, borrowed from a low cost commercial set of some sort. These connectors had at least four alternate numbers but are basically identical otherwise except as noted below - ARC 3146, Navy 77, Signal Corps PL-77, and PL-157 are all I could find with a cursory run through the manuals, but there may be more.
The connector is composed of only five pieces...the knurled retention nut, a wave washer, the connector body, two pins, and a phenolic insulator. There is another connector made for the very early SCR-183 models (a PL-76) that is identical except that it only has a single pin. That is the one pictured below.
The original connector body is made from brass, and has a wall thickness of about .016". It reminds me of the low cost RCA audio plug design. Since I prefer to work in something more substantial, I decided to make this particular sample from stainless steel. It has the same wall thickness on the tail and front portions, but a bit thicker through the center. See drawing and finished part above. Note that the diameter of the tail differs depending on the particular connector being replicated. The Navy 77 connector wire entrance tails are .370" in outside diameter while the Signal Corps 3186 is .250" OD.
I usually modify designs where they need some extra insurance, and that's what I did to the insulated insert. It is made from some teflon rod I had here in the junk box, and fits ino a cavity in the connector body with a lot more support than the original phenolic disk. I also put an insulating "skirt" portion on it to prevent a stray wire strand from touching the side of the connector body...a design flaw that had bothered me from the start.
The two pins are made from brass rod and are the only parts that need to be riveted. An alternative, as I did here, is to thread them and the insulated insert, and just screw them into the tapped teflon with a pin vise. They do need to be drilled for the wire and solder to wick down through them. Electroless nickel or tin plating would be useful but isn't essential unless your radio shack is subjected to the salt spray of a carrier flight deck...
Installation of the Teflon insert in the connector body is simply a matter of sliding it into the cavity and burnishing the edge of the connector body with a polished burnishing tool as it rotates in the lathe to form the edge over the insert. Some lubricant is useful. The end gets turned over about .050" of the overall length. As you can see in the first photo, the original also has a stake in the periphery to keep it from rotating, but the purpose of that step isn't clear to me. There are no indexing devices on the connector like a groove or boss that require a particular angle of insertion other than aligning the two pins with their female counterparts, and it works fine without the staking step.
The final piece to consider is the knurled nut. It is relatively easy to make from scratch with the right tools, but it is also (mostly) the same size as are used on the large diameter Bendix tuning cable ends detailed here. The only difference is in the size of the "through" hole, and needs to be smaller - .375" on this connector, or .255" for a 3146 replica. If the tuning cable nut is used, a thin .375" or .255" ID washer can be inserted before flaring the end of the connector body to retain the nut. Flaring isn't essential. It is done to provide some end radius to minimize cutting cable insulation, and to keep parts from getting lost before the cable is installed.
Cable installation is one of those endlessly irritating procedures that requires you to thread two small wires into the connector pins while essentially flying blind in the process because the insulation on the cord blocks all view (this is especially fascinating if you are using a phone cord with its ultra-flexible cloth/copper wires, a choice I do not recommend very highly at all). I'll leave it as an interesting exercise in patience for the reader. Using normal two conductor instrument power cable of the correct OD would be my recommendation.