Sunday, September 6, 2015

Interrupt enhancement to SAC box wired and ready for testing, attempting 1053 pivot screw replacement


I carefully measured the shoulder bolt that I needed to replace for the pivot arm. It has a 1/4" diameter shoulder, 3/64" deep, with the remainder of it a 1/4" long 6-32 thread. The cap is a hex 3/8" diameter, 1/8" deep. I have some leeway in the size/shape of the cap and of the length of the threaded portion, as long as the shoulder is 1/4" and at least 3/64". I can shim a deeper shoulder with washers.

Sloppy, quick sketch of the shoulder bolt that must be replaced
I found I had a nylon spacer that was 1/4" wide, but a bit too narrow for a 6-32 bolt. By the time I drilled it out, cut it down to about 3/64" wide and put it on the bolt, it had expanded just enough to make a friction bit with the pivot. While I was able to install it this way, the pivot arm needs to be pulled in one direction by a spring, which is way too light a pull to handle the friction as the pivot sits.

Back to the hunt for a good pivot surface. Perhaps I can use 1/4" OD metal tubing, which I ordered and should receive by Wednesday night. Its inner diameter is wider than ideal for 6-32. I need about .14" inner diameter but the tubing will be .20". I will use some soft rubber inside to help center the tube over the screw, which should be good enough for the purposes of the pivot mechanism. It will be challenging to cut this in a short segment - thus I likely will have to use several washers to reduce the open width to 3/64".

I believe I can replace the attachment hook onto the end of the nylon tape that shifts ribbon colors, if I loop the tape through the attachment piece and then glue the tape to itself. I am going to try E6000 glue which appears to make a good strong but not brittle joint between nylon pieces.


I am pondering how to make the button connections inside the 1131. The wires are soldered to the switches under the buttons, wrapped in a bundle and exit somewhere where they are fastened to a resistor block. That would be the ideal place to intercept the connections, rather than cutting wires in the bundle, but I had to first locate the illusive resistor block.

My FE Maintenance Manual has a locator chapter and with that, I found the resistor panel is down inside the chassis on the underside of the keyboard well, behind the logic gates. I have a great place to snake the wires up to the relay, a large round hole in the base plate of the keyboard well just where I intend to mount my boards.

Resistor panel where button wiring can be modified
Getting access to the panel to do the wiring will be a challenge, given the limited room behind the 1131 in my workshop. To make room, I relocated the P390 system I was checking out to my datacenter shed, which gives me space and a lighter table to swing away when I need to get in behind the logic gates (e.g. where the resistor panel lurks.

Access problem to swing open logic gates
Logic gates A and B swung out so I can access resistor panel deep inside
Once I looked at the physical panel, I had a plan for a minimally invasive wiring of my relay board into this spot. I needed three pins that would fit into the panel, which I had available from when I needed these to put into the 029 keypunch modification. I would gently remove a bit of insulation and then put on a vampire tap wire to each of three wires, leaving their pins intact but shielded by heat shrink tubing that would cover both tap and pin. These tapped wires would come up to the relay board, as would four new wires with pins attached.

Pins can be pulled out and new pins added to each vertical group of three
Since my relay board is a parallel switch for one button, a series switch for a second button, and a reverser circuit for the third button, it needed those seven wires coming up from the panel plus a wire bringing the +12V that drives these switches.

I deferred the work wiring in the switches, as I will wait until I have tested the rest and see the emulation engine clicking through its relays on command from the PC. This reduces the amount of change I made to the 1130, making any diagnosis easier for now.

It was time to find the sourcing for both +12V and +6V, and ground, that would be used by my added boards. It was most convenient to take it off of the B gate (front most of the two swing out logic gates) where I had ground, +6 and +12 at screw terminals at the bottom. I made up some terminal ends to put on the three wires and routed and dressed them.

The DC-DC converter is mounted to the right side of the usage meter and wired into the +6V supply. Next was wiring the daughter, emulation engine and relay boards to the +5V output of the converter.
Too, I hooked up the +12V to the appropriate relay board contact, as this is needed to implement the Imm Stop button.

I fired up the system to see whether I could watch (and trigger) the two interrupt levels I added, but the machine came with IL4 hot - probably another of the pesky connections to my output board which are so good at working loose and causing problems. I have to get ready now for a dinner with family, so debugging will have to wait for tomorrow. 

No comments:

Post a Comment