Hi Greg,

Both the Marklin and Viessmann reed switches are designed to fit on one side or the other of the centre pukos. The magnets available from Marklin are rectangular and come in various thicknesses and magnetic strength. Since I operate my layout as you describe with trains running in both directions I have seen the problem that you describe that depending on locomotive and location of the magnet, if the magnet is off-centre it may not trigger the reed switch in one direction. For those locos I have either added a second magnet, one on each side, or switched to the more powerful neodym magnet. (Be careful with those if you have small kids around. Swallowing these can be fatal.) I find these magnets have so much strength that they trigger the reed reliably each time. However, they are also so powerful that they can lift the moveable points of the older style switches and cause the loco to stall or derail.

So the key is the magnet positioning. If you use computer control software to run your railway, ideally you would want the magnet to be located at the same position on each loco so that you stop at the same location for signals, switches, etc. Unfortunately this is not possible due to bottom mouldings, slider position, etc. and I find that I have up to a 10" difference in magnet location between a short and a long steam loco.

Hope this helps.

Cheers
Peter

Just being curious: Is there any specific reason for this? IMHO, using C tracks, it could be easier to just cut the outer rail and connect it to a proper module for S88 bus, or similar?

Peter

Hello Peter.

Yes, there is a reason.

When you use a magnet, you got just a single trigger on the S88.

If you use the contact-track, you could have several triggers/hits on the S88.
Since the axles are the electrical contact, you should take care on the distance between the axles, so at least one axle is everytime on that specific track segment to guarantee a single trigger.
(you cannot use a small segment of track that fits betwwen the axles of a long passenger wagon for example)

Cheers,

Walter

Its the tension between wanting a single signal and wanting a precise location signal.

There are multiple ways of addressing the digital signals that result and defining how to react to create a reaction when you want - but I'm not familiar with the way systems using S88 can react to signals and what they can program (can one program in logic and timing?).. to know whether digital processing techniques can be applied in the control software used in model railroads. I'd hope one could.

Certainly more than one way to accomplish a task. I use reeds exclusively, just preference.
- at entrance and exit of hidden yard and passenger station - CS2 memory will execute sequence of commands when magnet activates a reed.
- at signals - to turn prior one green or red.
- in covered sections, to turn current on or off for blocks

Regards
Harvey

I think it all depends on the application/logic/layout/expectation...

Reed contact: just one hit from the train, usually direction dependent (only works in a one-way direction).
Circuit track: several hits from the train, one for each pickup shoe, no direction dependent (two-ways trigger, identifying both directions).
Contact track: could have one or several hits depending of the track length, no direction dependent (two-ways trigger, but just one identification regardless of the direction).

I think IR sensors could have the same application of the contact track; could have one or several hits depending of the position of the sensor (horizontal/vertical) with/without triggering between cars.

Cheers,

Walter

Thanks for all your comments and details of where to buy magnets. I like the idea of track sensing but not of cutting rails. I like the use of magnets - but not sticking them under every loco, I favor IR as neither is required - but there's the cost and wiring :) There's always a pro and a con it seems. I also like the GamesonTrack - wireless positioning - but we're back to sticking transponders on trains and wiring them ?

I had high hopes for 'virtual blocks' on pc software - and it seems that as long as there is 'one' of the above per block TC can manage all the other contact points along the block through measured distancing. I will experiment over time and see what the best combination really is.

Rgds

Greg