Atlas turnout problem question

[FJ and G]

Once again, I'm hoping you guys can help out my buddy who is building a VERY innovative 3-rail waterfront layout. I'll provide his problem, as he described it to me.

"I'm having a couple of "interesting" experiences with two Atlas switches ...
One works fine when loco runs straight through, but when switch is thrown to
left passing siding (the final block I wired), loco shorts out when it hits
section of switch on curve ... But not on track on blocks to either side of
the switch (switch track receives power from one side; track at end of
switch and turnout have insulators on center rail). I have triple checked
polarity of feeder wires to all blocks adjoining switch--red center rail,
block outer rail--and they're all wired properly, plus loco runs fine
through all blocks; it's just the middle section of the switch turnout where
it shorts. Is the internal isolated rail in the switch turnout wired
internally underneath? Any thoughts on troubleshooting are welcomed!

Another switch has a "dead" section on the turnout so that when the loco
rolls over turnout, before frog, it stops. My "test-car" illuminated caboose
also goes dead when it hits this two-inch section of track in the switch
(just beyond movable section of track at frog), but works fine to either
side. Huh?

I'm going to lift each switch out and test it on tabletop to see if I truly
have two defective switches or a wiring problem in my block. Ugh. I'll be
happy when all wiring is completed--it's not my favorite dimension of our
hobby!"

[Big_Boy_4005]

Does the caboose he's using for testing have rollers on both trucks, or just one? What engine is he using, and has he tried others? Does the engine have traction tires?

My guess is that the dead spot and the shorting has to do with the rollers, and the distance lengthwise between them. If the switch is one of the longer ones it may be necessary to add directional power, similar to what is done in 2 rail operations. A DPDT set of contacts would solve the problem. Instead of switching the power off and on on the center rail, it would need to switch segments of the closure rails where the rollers cross over.

I have drawn a diagram, but it looks like my ISP's server is out at the moment. I'll post it when it comes back up.

[FJ and G]

Thanks for the info, Elliot.

[Big_Boy_4005]

This is a pretty crude drawing, but as the switch gets longer, the red and green sections come more into play. Equipment with short roller bases will have the most trouble. I have some Curtis #8's where they need this kind of wiring.

Ask him if there were any instructions with the Atlas switches.

[eZAK]

Dave,

For the shorted switch use a continuity tester or an Ohm meter.
The switches should be internaly wired via a solid wire inbedded in the plastic ties.

For the one with the dead spot there is indeed a problem.
The dead spot is in the center rail. It looses contact at the joint just outside the points.
To further complicate the matter Atlas used an aluminum rail for this piece.
Which means you can not solder this joint!

The two ways I found to fix this is either drill a small hole in the rail, insert a screw, the solder to this Or replace the rail with a nickel-silver one.

[FJ and G]

thx again guys,

Pat,

That's crazy about the aluminum rail. Perhaps he should just return it and get a replacement.

[eZAK]

Dave,

I wish I could return them! I have five total. Three I had to repair.
I was orignally told by Atlas "...the switches work as designed"