Interesting question. I've already started using a third alternative, Uhlenbrock Lissy. It does not send anything on track, but instead uses easy to install IR-senders in loco, and detectors in track; connected to a receiver. The receiver is a programmable device, which may send differnt LocoNet commands, like change speed or function on detected loco, turn on light on a nearby station or level crossing, or more advanced functions as "make a stop at the station", or "stop, and return after 5 minutes".

Lissy is not cheap, but I love it for it's functionality, which certainly encourages modular thinking (the whole LocoNet does). On a module layout it would only react on locos with an installed IR-transmitter, but for those make some fun things. Normally the programming is "for any detected loco, do ..." but it's possible to do some extra actions for special locos ("send F3 (horn)" if the adress detected is no 1234"). Lissy operates in with any track protocol, but Märklin/Motorola and DCC or combination is the typical thing to use.
Lissy also makes the detection at a point, which in my eyes makes more sense than detected a loco just entered a section. No great difference, but still. But most important; to my knowledge the signals from Lenz or Digitrax can't really be used by ready made equipment yet; just presented, and of course used by software in a computer. Lissy shows the whole chain, and it works very well. For my part, I don't consider changing.

And LocoNet I love even more for it's design.

/Lars

The Digitrax solution is based on modulating the current consumption of the decoder ( same principle as mfx feedback). The Lenz system uses the standard NMRA method of sending data during periods of no track voltage ("cut-outs").

From just a theoretical wiew of the two different principles, I would definitely vote for the Digitrax solution. Reasons:

1. I think current sensing is (at least theoretically) more robust than the NMRA method.
2. No 'polluting' of the track signal.
3. You can use existing DCC decoders and just add the transponder. Certainly cheaper and easier than replacing the decoders in all locos.
4. You can use any DCC control unit.
5. The current sensing works even if you have static load (e.g. lighted cars). The NMRA method may not work in that case (this is also just a theoretical conclusion based on the NMRA spec.).

However, I have no practical experience of any of the systems.

I Don't know if the Digitrax system works also with Märklin/Motorola signalling. The method as (described in the underlying patent application) could work with both DCC and Märklin/Motorola, but it seems like Digitrax don't care about the Märklin/Motorola protocol at all.

In short, it still seems to me that Lenz, sorry, NMRA system, is to complicated at least transition from an old system. And I feel the same as perz say; Digitrax system appears more robust also in the long run. Fully agreed that using IR seems a bad idea outdoors. The abiltiy of Lenz to read out decoder values like this doesn't seem very usable to me. I want to locate my running locos, and cause them trigger actions like:
- change speed, direction and functions of the loco, including modify sound etc.
- cause changes of accessories like lamps, signals, maybe turnouts.
- less important: show train ids on a display, and why not, "dispatch" control of the loco to a throttle near yourself.

Most focus so far has been on the detection part, not only here, but also on other places. However, what I like most with Lissy is how the system is built. The *system* is modular I would say, and also uses LocoNet which strongly encourage modular solutions. Most system rely on stupid sending the info to a PC-program, which keep track on everything, and perform actions upon receiveing info. This would need at least on program configuration for club running, and another for running at home. If one module is missing at the club, the program must be modified.

Of course, it's possible to build software modular, so with each layout module included, one also installs a software module desribing that piece of track, what actions to take etc. But to me it seems easier ot handle to have those programmable receivers as hardware units mounted at the layout module they monitor. When you come to the club, you just plug in the LocoNet, and are ready to go. (Well, I think some adress mapping could cause problems, but if each member has an assign series of addresses it would be OK).

At least the digitrax solution could be extended into this direction, I mean, Digitrax could develop a more intelligent current supervisor, and in that case Lissy and Transponding would be rather equal but use different techniques.

As said, I'm biased towards LocoNet, have it and love it. If you speak about wireless together with mixed protocol (Motorola + DCC IMHO is good enough) I think about Uhlenbrock central and Digitrax throttles, however, no experience about Digitrax throttles. In Europe radio units seems to rare, becuase of all authority approvals needed. ESU have one, but ECoS seems to be a very centarlized solution. Just my few cents; I don't have any market overview. /lw