Point actuators
Yes, the whole session was spent trying to sort out the point actuator design. Geoff had manufactured all of the remaining units which saved some time. However, he reported that on testing, they all operated in slightly different ways, which did not bode well for setting them all up to be consistent.
We started to investigate separating the servo unit from the tie-bar mechanism. Stephen had used the method on one of his layouts and had linked the two units with wire, which had a connector for adjustment and an omega loop. We found an old actuator unit of sliding plastic tubes. We remember making it, but not what it was for (might even be on this blog somewhere!). Laurence drilled a hole in the end of length of 8BA studding and Geoff soldered-in a short piece of spring wire. The assembly was secured into the plastic tube unit, and the unit screwed over the tie-bar. The servo unit was fixed a short way from the actuator and spring wire, with a simple loop (not Omega this time), connected between them.
The usual DC supply was employed, but there was a problem with the servo arm bouncing off the micro-switches. We also found that the linkages generated too much slack and that the servo would occasionally stall, requiring a finger poke. It was all very dispiriting, and clunking solenoid motors seemed like a good option.
Around this point, we also discussed how we were going to adapt the existing control panel & wiring to the new units. We could use DPDT switches to deliver a fwd/rev DC supply as had been made the previous week. Geoff had also proposed using an AC supply through a SPDT switch with diodes on each output to recreate the fwd/rev supplies. To cheer us up, we decided to try this out and a switch unit was assembled. We tried different voltages to overcome the drop the diodes created and soon found we had a very smoothly working servo - just what we'd been looking for.
However, (lots of 'howevers' this week!) we still had to sort out all of the problems we'd encountered - a tall order. Right at the end of the session, as we were about to pack up, Geoff described how 'mainline' point actuators travelled far more than our model. In fact instead of a couple of degrees, they swung through approximately 160 degrees. This had the effect of varying the speed of the operating arm, which slowed at the ends. Could something like this solve our bounce-problem? It might be worth looking into next time.
