This time the ESU sound decoder installation went perfectly. Instead of sending the old one back in right away, John opted to buy a brand new one and have the sound profile loaded into it. I bought a new speaker just to be on the safe side, and when the decoder went in it worked perfectly and sounded great.
Reducing the flanges did make for some problems on my tight radius layout while trying to program it, though. When the loco goes into a curve it tends to derail the idler (un-powered inner) wheel set on the rear truck, and in a left hand curve the wheel touches the truck frame and shorts out. The wheels derail in a right hand curve, too, but they don't short out. This was not a problem with the plastic wheels, of course.
I didn't notice this problem on the large radius NTRAK layout at NRV, but when I tried to run it on Kato Unitrack 348 radius curves when programming the speed it jumped and shorted at speeds greater than about 45 smph. That means it will probably do the same on the 315mm radius T-TRAK curves. I really hope we won't have to reinstall the plastic wheels because the metal ones just look so much better. I may end up having to reduce the truck frames on that side of the loco to prevent the problem, or maybe I can cut a disk of very thin styrene to put on the insides of the wheels.
But that is an issue for another day.