That is excellent news, Tony. I don't know exactly why it is like that, but the fix certainly works.
I’m guessing it’s out of balance.
Like a steam loco on a turntable, if it’s not evenly weight distributed over the pivot, it’s impossible to turn the loco without additional disproportionate levels of energy than if it were balanced.
By removing a shaft your not fixing the balance, just removing a source of tension.
I’ve not had the motor out of the chassis yet to confirm, but i’m Suspecting the motor itself isn’t evenly balanced, perhaps the motors weight is bearing down on one end more than the other, causing one shaft to be climbing, and the other falling meaning they are fighting each other. The motor itself looks to be mounted evenly in the chassis, but the motor design is not identical at each end. It could also be lateral movement of the shaft itself (ever wondered why angular mounted x04 motor locos work better one way than the other)... going one way energy is screwing down on to the gear to drive it, the other energy is unwinding upwards away from the gear. I think maybe the 121 is doing both at the same time, winding at one end, unwinding at the other.
When I looked at it, the gears, gear tower, motor and bogies are all snug fit tight with no resistance to moving parts, so I can only speculate it’s within the motor.
I did a similar exercise with a troubled batch Heljan 17, there it was clear.. the gears were too tight in the gear box and hard to turn even by hand.. this isn’t the case...they are free moving, but fit snug.
On my first pass I removed the shaft from the DCC chip /guards van end, but on learning, i found it’s easier to remove the shaft from the seated end for the other two I have, but I found it made no difference which shaft I removed, as long as one was removed the performance jumped.
Edited by adb968008, 18 January 2018 - 00:02 .