I've always been frustrated with the poor running of the Skylark I built a year or so back. The Kato chassis just didn't seem to perform as well as other people's, and indeed the one I have under the railcar, it would run smoothly at speed but stall frequently at lower speed. I'd checked the wheels were clean and all appeared to be in order, but running didn't improve. It would be nice to have it running well for the Shepton Mallet show next week, so last week I took the loco along to a Sussex Downs group club-night and with the help of others we set about a diagnosis.
Eventually, as suspected, we traced the problem to the pick-ups. Although they looked fine, they clearly weren't making good contact, there may have only been one working pick-up each side. We found the only way to fix this was to remove the white clips on the top of the chassis (carefully, they can ping!), and ease up the pick-ups where shown with the yellow arrows. We did one end at a time, and as well as cleaning the pick-ups I tweaked their shape to have more of an outward splay, before feeding them back in behind the wheels. While the chassis was apart we applied a tiny drop of oil to each worm too.
The result is a chassis that runs as smoothly as it should, without stalling and controllable down to a slow speed. Since I was rather pleased with the way the body came out, it is great to finally have the loco running as well as it looks!
Spurred on with that success, I thought I'd address the uncoupling magnets. The Neodymium magnets I had tried in two locations really hadn't worked, they are actually too strong - wagon wheels and coupling tails shoot towards them, sometimes resulting in derailments. However in trying to excavate them I managed to damage the track - it didn't help that there are double rail-gaps and cut-about sleepers over the hole in the board for the magnets, but having rails sticking in the air was never going to work! And with a week to go before an exhibition this is not ideal.
Fortunately I managed to bend the rails back flat (it was the fishplates that had bent) and added a new "spare" Peco sleeper. A piece of paper was glued in place under the track (to support ballast), then the original cupboard-catch magnets inserted from below with lots of hot-glue to hold it. The glue-gun was also used to stick the tips of the sleeper to the baseboard. By this point the track looked like it would at least support trains!
I dug out some fine ballast and tapped it into place between the sleepers, not too thick to not limit the magnetism, but enough to cover the paper and glue lumps! Finally a little acrylic paint was mixed up and slopped over the lot to blend it in. The colour may not be an exact match, but I've tried not to be too even anyway so it looks fine.
Phew, disaster averted. Back to cleaning wheels and testing stock...!