Drama and defeat in the battle of the tab cord attachment

SURPRISINGLY HARD TO GET THE CORD OVER THE SPRING LOADED PULLEY

When I had attached the tab cord to the drum and the carrier, it only needed to be pulled out around the spring loaded pulley that will maintain tension. The strength of the springs on that pulley don't allow simply pushing the pulley inwards while manipulating the cord. 

I ended up removing the pulley assembly, fitting the cord on it and then twisting it up into place in order to reinsert the screw attaching the assembly to the typewriter. There it was - a set of cords attached and at proper tension. The carrier was about 1/5th of the way to the right, which was the position that allowed me to attach the tab cord, maneuver it to get a second loop around the drum, and attach it all prior to the spring loaded pulley dance. 

THOUGHT I HAD IT, DID CARRIER RETURN TO WIND UP THE MAINSPRING

I latched in a carrier return to pull the carrier to the left edge of the machine and to wind some tension into the mainspring. Everything moved well until I reached the edge and tried to release the carrier return clutch. The machine did not unlatch, it strained pulling the carrier against column 1 and kept turning the tab cord drum so that the end of the cord was barely in the notch with no spare cord. 

POP GOES THE TAB CORD

As I wrestled with the machine trying to get it release the carrier return clutch and stop turning everything the wrong way, I heard a pop and the end of the tab cord came off the drum. Time to start over and try this again. There must be some secret to doing this, but none of the IBM manuals nor the YouTube videos address it, instead presenting the idealized process where it all just fits together and works. 

TAKING A BREAK FROM THE FRUSTRATION BEFORE I DIVE BACK IN

I did some work building my new vacuum tube curve tracer, a project that I originally cobbled together some years ago based on Ronald Dekker's uTracer design. Ronald works at Phillips Research and definitely knows vacuum tubes. This device connects to software on a PC that gives all the curves and results with a very nicely built GUI. 

The uTracer had a weakness in how it drove the filament voltage which was corrected by a clever design from Stephan Lafferty that adds a very accurate filament/heater power supply. I recently completed the heater part of the project and decided to improve the build and esthetics from the quick and dirty version I made years ago.

I cut a new front panel and made a decal labeling the various controls and sockets, unsoldered everything from the old panel and began construction. I lost one of the knobs so I will be replacing all ten of the bottom knobs with a new type of knob, as soon as the order arrives. I installed all the controls and sockets today and began soldering everything up. 

Each of the nine possible pins of a tube socket as well as the cap electrode can be switched to one of six roles - cathode, control grid, screen grid, anode, or either end of the filament supply. That accounts for the ten knobs at the bottom. The rotary knob at the top sets the filament voltage to 5, 6.3 or 12.6V but allows for an external supply if a tube has an unusual heater voltage requirement. 

There are four sockets, for 7 pin and 9 pin miniature tubes, octal tubes and loctal tubes. Pin straighteners are on the panel to ensure that the pins don't bend or snap while being handled. I can make adapters that plug into the octal socket to accommodate rarer tube types including nuvistor as well as the older tubes like a 4 pin radio tube. 

I also did some organizing of the workshop, collecting items into containers and labeling them. This makes the shop neater and more open but also helps me find what I need much faster than the chaos of before.