It occurred to me that if I was able to run an overdrive on the bench, both engaged, disengaged and neutral, at least up to a thousand rpm or so, it may give a reasonable indication of bearing condition and problems with the drive. I made an adaptor that fits on the electric drill and into the clutch drive dog on the input of the overdrive, then ran with the drill in both directions, the most noise heard was from the drill motor.

I realise, before someone points it out, that there is no load so it isn't 100% accurate, however any serious bearing noise, synchro problems, gear meshing, clunks, shift problems etc. should present at standard running.

Is this likely to be a relatively sound method for testing such devices? Feel free to let me know if (and why) I'm way off the mark if this isn't right as I'm by no means knowlegeable when it comes to gearboxes.