There is a widely held belief among proto:scale modellers that equipment works better if it is compensated or, better yet, sprung, or even better yet, compensated and sprung. It stands to reason that wheels that are able to stay on the rail are less likely to find the ballast. However, in an email, Rob Kirkham speculated that the tiny differences in angles due to compensation might have an impact on smooth running.
I was out of ideas for how to make Percy run as badly as #10 and #622. So, I designed a simple 3-point compensation modification, allowing 1/2 mm vertical play in the rear driver around a central pivot. The results were very interesting.
Essentially what I found was that the fixed axles that Percy came with constrain the mechanism and force it to behave. With one axle floating merrily along so it could accommodate track irregularities, any misalignment in the wheels was magnified. The rods still did not bind, but the engine bucked down the track like a rodeo bull when I ran it backwards. Interestingly, when run forwards, it was about the same as when un-compensated.
Could such bucking act like a tiny hammer, weakening any bonds that held the wheels in quarter? It’s an interesting question – one that I don’t know how I could test.
Percy’s original wheels were the only ones in my collection that didn’t buck. All others, including those that are nominally correct exhibited some bad behaviour.