Looks to me that this larger endplay has the advantage that the dogs have the opportunity to 'find' eachother without being obstructed by the limited forkaction.
So maybe minimal endplay is not that critical....
I'm investigating two gearboxes at the moment, each with burnt selector forks. One is a real problem, the other is fake due to the gearbox being assembled from someones rubbish :(
As with the gearboxes I manufacture I think the answer is to build them by calculation.
The random machining quality of the dog clutches (dog rings) enables you the change the throw buy swapping the faces.
Even the fork throw varies in position and offset.
Too little and endfloat will cause problems in the wrong places.
BSA M20 factory worksheet No. 608 states ['All shafts should have the minimum of end play']...The lightweight box that was fitted to the WB30 also operates in essentially the same way as the M20 box...
WB30 information states...['washers can be varied in thickness to adjust for correct end float of the shafts, which should be .005" for the mainshaft and layshaft and .003" for the selector shaft...']..
You can make of that what you will but it seems clear to me and I have always used these figures to build gearboxes, achieving correct operation in that condition...'Lost movement' caused by excessive clearances (either wear or setting) can have, in my experience, a detrimental effect on both the action of the box and the correct setting of the selector dogs on assembly and is therefore best avoided...
Another relevant point to check is that the washers either side of the main bearing, which have 'depressed' centers are fitted facing the right way....Ian
