"voodoo and mysticism" and gremlins

[gregc]

As a firmware developer, I often use the phrase "voodoo and mysticism" to describe code usually written by someone else that works and which we're reluctant to touch because of a lack of understanding.

The phrase came to mind yesterday while trying to understand why a tender would derail when pushed backward toward a turnout.  While the tender would derail when on the turnout points, it would very often derail before even reaching the turnout.

Part of the frustration was that I couldn't see the wheel on the wall side of the tender that was derailing.  It seemed that the wheel would just climb over the rail.

Eventually I got a mirror and was able to see that the rearmost wheel wasn't sitting flat on the rail.  When I looked at the tender truck by itself on track I could feel it rocking on its wheels.  Some minor bending straightened the brass truck and the problem was resolved.  (And the mirror is now part of my modeling toolkit).

But the tender and other cars were still having problems on the turnout.  I eventually realized I had filed the top of the point rail a little too much when building the turnout, making it somewhat of a ramp for wheels to climb over.  After some unsuccessful attempts to make it taller I eventually just bent down the tip that is soldered to the PC tie raising the rest of the points rail.

While both problems were caused by me and I have plans to rebuild the turnout, it was satisfying to gain the understanding that I did, which restores my faith that there aren't gremlins in my layout trying to sabotage it.  And that unimpeded and careful observation can make problems relatively obvious.

[mlehman]

greg,

There's another useful task for a mirror in cases like this, although you may need one larger than the one you used to look on the other side of the tender. A plate glass mirror or just a piece of plate glass big enough to set the entire loco on can be used to check things like the problem you found with the truck riding high. Set the loco on the glass, then try to nudge a sheet of paper (Post-its are a handy size) under each wheel. If you have no resistance and free space, you've likely found a problem. If the wheel is flat on the surface and you have to nudge the tip of the paper under it, you're good on that wheel.

[rrinker]

 Indeed. Nothing "just happens", there's always some cause. Whether we can find it or not is another story.

 That's a slightly odd one - my first thought on a tender derailing in reverse is that the drawbar would be pushing slightly sideways. But what you douns makes sense as well - when pulling forward the out of alignment truck would just get pulled through any trackwork and go where it's supposed to, but in reverse that wheel is now in charge and will go the way it wants to.

               --Randy

[BATMAN]

I recently bought a bunch of Kadee sprung trucks as I had one truck just disintigrate like so much bad pot metal. I put my boxcars with the new trucks on them in a forty car train and had one boxcar derail almost every time over a curved turnout. After checking the gauge and everything else I could think of, out of frustration I took all the new trucks off and started firing them through the T/O on their own. The bad one always slowed right down, stopped or derailed every time. To make a long story short, it turned out that the wheels were out of round. My future metal recycler has some added inventory of a brand new Kadee wheelset.

I agree that there is always a cause to a problem. On more than one occasion I have used a video camera and watched on the big screen frame by frame before I have found the cause. Never surrender!