Monthly Archives: October 2018

Further Pontiac restoration

Driving the car down the street showed that something was a little amiss with the gearbox. It would hold first gear for a long time and only change once I let off the gas pedal. At that point the change was harsh and jumped from first to (what feels like) third.

The gearbox is a moderately conventional hydraulic design; two epicyclic gears, each with a clutch and engagement band. The engagement band operates one ratio, the clutch locks up and engages it as a gear in itself offering a second ratio, and then that all disengages and the second set does the same thing. 

It has two hydraulic pumps. the primary pump is on the input shaft from the engine. The secondary, a smaller pump, is on the tailshaft and is only used to bring hydraulic pressure up in the gearbox if the car is to be bump-started (drag car to 15-20 MPH, engage Dr (drive) and that will spin the engine over in an attempt to get it to run).

The faster the input shaft from the engine turns, the greater the pressure it creates in the hydraulic system. Once Dr is engaged, this pressure is directed to the valve block. Once the input spins up to a certain speed, the pressure is meant to be a certain amount and the next valve opens, engaging second gear, then on and on through all four gears. The only other external influence it has is the connection to the throttle pedal, and that operates a relief to the pressure the more it is opened.

So, light throttle the gearbox should cycle through and engage 4th by 21 MPH. The more the throttle is opened, the more pressure is bled off so the longer each gear holds before it changes.

Adjusting the throttle linkage.

I adjusted the linkage, thinking that it was set too far open and was bleeding off too much pressure too early. This did not help, and further research indicated that bringing the revs up then snapping the throttle shut causes enough of a pressure spike to operate the valve block and change gears. In short, the valve block is needing re-sealing as it’s losing too much pressure past the seals to operate correctly.

Broken envelope and burned filament

I turned my attention for a while to the electrics- I had wanted to see if the non-functioning parts of the dash illumination were down to faulty wires or faulty bulbs. In most cases the original 6 Volt bulbs were still in place and had failed. Two here had failed in different ways. The one on the left had burned through the glass envelope, rupturing it and allowing ingress of air. The filament then turned to titanium trioxide (the yellow-white powder) and ceased operation that way. The other one the filament overheated and vaporized, plating the inside of the glass.

License plate light

The license plate light is sadly not original, instead an auto-parts-store generic, and the flitch panel it sits in is a little rusty. It’s been mounted in a steel plate, badly. However, all that was wrong with it was the connections were rusty. Cleaned up, it began to work.


Overall, an improvement. It since stopped working then I accidentally broke the bulb pulling the holder out. Go figure.

Dashboard light

I finished up inside by fitting working bulbs to everything. It needs a clean but overall looks very nice all lit up.