94
« Last post by RDfella on 25.03. 2025 11:17 »
Following Swarfy's input, was thinking about how to cut out the guesswork and time spent fiddling around. Even if ultimately the gears work (gasket or not) one still doesn't know whether the pawl engagement is satisfactory or whether it's all operating on luck.
I appreciate these gearboxes are getting rare and they're not being rebuilt every day, but after an afternoon's frustration and still not sure if all is as it should be, I thought about how to ensure proper engagement. Given the engagement relies on so many machining operations being simultaneously accurate to a couple of thou using elderly machines and practices, I reckon that to avoid getting a bad reputation BSA must have assembled these gearboxes from bins of parts selected for fit. We have neither the jigs nor measurements to do that, so if ever I face that issue again, here's what I'd do:
1. assemble box case and cover using increasingly thick shims (say 3 x feeler gauges) until gear selection is compromised.
2. If shims are in the region of 2mm (.018" gasket + 1/16" engagement) then all is fine. Go ahead and assemble.
If not,
3. measure the depth of gearcase to cam lobe (say in 1st gear).
4. measure selector / rocker protrusion.
Then whichever of 1 or 2 is incorrect needs machining, but we need a few measured examples to determine which one.
There might instead be the option of making an eccentric selector pin, but I don't have the parts in front of me to tell.
Small addition - just measured distance to cam lug using a spare gearcase shell and a new cam I have - it's 1.510" at its highest between gears.