Transmission / Clutch
06.3 / Engineering Notebook
Short records of the actual work, decisions, checks, and results from rebuilding the transmission.
Project methodology
- 01Understand
- 02Model
- 03Inspect
- 04Measure
- 05Analyze
- 06Improve
- 07Test
- 08Reflect
Transmission improvement log
What I did, what I found, and what changed.
I started with loose parts from three incomplete gearboxes and no reliable way to tell which components belonged together. These entries follow how I turned them into one assembled four-synchro, three-rail gearbox.
Laid out the geartrain
I arranged the shafts, gears, synchronizer pieces, washers, and retaining hardware in the order shown by the exploded diagrams. I built a loose mock-up outside the casing and traced the power path from the input shaft through the laygear and mainshaft. This also let me compare parts that looked nearly identical without repeatedly installing and removing them.
Result: I understood how the loose components formed one mechanism, and missing, duplicated, or incompatible pieces were much easier to spot.
Additional research needed
I could identify most of the components, but I did not yet know which wear marks were normal and which made a part unsuitable for reuse. I met with an engineer at Sportscar Craftsmen in Denver who specializes in classic British gearboxes. We went through the shafts, gears, bearings, synchronizer parts, and casings together. He showed me which contact surfaces mattered most, where these transmissions commonly wear, and how to distinguish light polishing from scoring or damage that could cause a failure.
Result: I now had actual criteria for deciding what to reuse, replace, or reject instead of simply choosing the cleanest-looking parts.
Added a third gearbox for parts
The two original collections did not contain a usable version of every part I needed, so we obtained a third incomplete gearbox. It gave me another version of several important components and meant I did not have to accept a damaged part simply because it was the only one available.
Result: I could reject poor parts instead of using them only because they were the sole option.
Final selection of parts
I compared each casing, input shaft, mainshaft, laygear, layshaft, forward gear, and reverse idler beside its alternatives. I checked the gear teeth, splines, bearing surfaces, bush locations, and the tracks left by the needle rollers. One mainshaft was deeply scratched at the bushing surfaces, one layshaft had visible needle-bearing marks, and several gears had damaged or bent tooth edges, so I set them aside.
Result: I ended up with one compatible group of the least-worn hard parts available from the three incomplete gearboxes.
Pressed in the replacement bearings
I used a hydraulic press to install the tail-housing bearing, main-housing bearing, and input-shaft assembly. We made drivers from sections of metal pipe, then sanded and cleaned them so burrs or loose fragments could not contact the new parts.
Result: The bearings seated without visible damage or contamination from the tooling.
Loaded the laygear needle rollers
I placed 25 new needle rollers at each end of the laygear. Tacky blue assembly grease held the rollers in position long enough for me to slide the layshaft through the center.
Result: All 50 rollers stayed aligned while the layshaft was inserted.
Checked the internal clearances
Before closing the casing, I checked mainshaft end float, the clearance between the gears and bushes, and the laygear clearance against the housing and thrust washers. These gaps had to leave enough room for an oil film and thermal expansion without allowing the gears to move far enough to lose alignment. I rotated the assemblies throughout the checks because a clearance that looked acceptable in one position could still reveal a tight spot elsewhere.
Result: The selected parts moved through a full rotation without an obvious bind or excessive movement. I still need to add the original numerical readings to complete the record.
Assembled the geartrain inside the casing
I installed the laygear, mainshaft, gears, bushes, thrust washers, synchronizer parts, and retaining hardware in sequence. The hardest part was holding several pieces in alignment while adding the next one without disturbing everything already in place.
Result: The complete geartrain fit inside the selected casing and rotated without an obvious mechanical bind.
Rebuilt the selector mechanism
I installed new detent balls and springs in the three-rail selector assembly, positioned the selector forks, and checked that each rail moved into its intended position. I then closed the casing with new gaskets.
Result: The detents gave each selector position a clear, positive feel instead of allowing loose movement between gears.
Bench-checked the completed gearbox
I moved the shifter through all four forward gears and reverse after closing the gearbox. Every position was reachable, and the selector mechanism did not show an obvious stationary bind.
Result: The gearbox passed its manual shift check. This confirmed the assembly sequence, but not its behavior under engine torque.
Installed the gearbox in the car
I installed the rebuilt gearbox and connected it to the driveshaft. The engine has not been started yet, and the separate rear axle U-joint has enough play that I do not consider the car safe for a road test.
Result: The transmission is installed, but powered testing will wait until the driveline is safe and the engine can run.
