Finishing the build of the substitute controller card to hook a 1627 plotter to an IBM 1130

BATTLING TO INSTALL THE GOLD CONTACTS THAT WILL FORM THE SLT SOCKETS

This printed circuit card fits into the IBM 1130 backplane in card compartment A-C1, where IBM would have plugged in a double width Solid Logic Technology (SLT) card 5806223 to provide the controller logic to interface the plotter to the IBM system. The backplane has two slots vertically that accept the card, each having twin rows of 12 gold pins that fit into the sockets on the SLT card. 

My card has springy contacts that will make good contact with the gold pins on the SLT backplane. I used gold plated RF shield contacts to build the card, each of which being soldered onto a pad on the card edge. These are very challenging to install, since they are small and hard to hold while soldering them down. I used solder paste, my heat table and hot air rework gun to solder them, having placed them onto the solder paste by hand. 

These move around a bit, not anchoring perfectly onto my PCB pads. Eventually I got once side set up with contacts, although not as neat and orderly as I would like. 

A few of the pads have no connections to circuitry on the card, so I left the contacts off in those spots. Once I checked for shorts and dealt with any excessively objectionable positioning, I moved on to the other side of the PCB since each SLT socket is two sides of 12 pins each. 

My first try ended in the card sliding off the heated table and scattering contacts whose solder was still molten. After a suitable break, I went back and completed the construction of the other side. 

I had 3D printed covers that fit over the contacts and form the shape of an SLT socket on a card. These slide over guide plates as the card is inserted and position the card socket into the backplane over the pins that form the complementary sockets. 

IBM cards have a complex shape to the gold spring connectors that line up with a curved shaped barb on the end of the gold pins from the backplane, forming a mechanical lock to hold the card in place unless it is tugged out. I can't manufacture parts to do the job, so I need a different means to hold the SLT card down in the backplane when it is inserted. The spring contacts on the card will cause it to push back out of the backplane socket otherwise. 

The solution to holding my card down centers on the guide plates that are on the SLT backplane. The card sockets have a notch on them that slides over one side of a guide plate, with the next card's socket fitting on the other side of the plate. 

A guide plate is circled

Pointing at one notch

It is hard to see the relationship based on the way that IBM drew the card socket and the existing pictures from IBM manuals. 

The drawing above shows how the two adjacent cards slide over the guide plate, as seen from the top looking down into the backplane and seeing the edges of the cards. My card is a double width one, thus it would have a guide plate in between the two sockets on the card, much like above, as well as notches on the outsides of the two sockets where other guide plates would fit. 

Adding friction or clamping on the guide plate from my card would seem the easiest solution. I have to make detailed measurements of the guide places and SLT card sockets in order to pick the point where the card should be locked in place, then work out the mechanism to add to my card that will hold it there, ideally by friction rather than a complex clamp.