How well does the RFID stuff work? I'm considering implementing it once I've built the base boards (or retro fitting it at a later date) but I've read mixed reviews on the reliability.
In this shunting puzzle, the trains are running 'relatively slowly' [Programmed to about 1/4-1/2 of possible maximum speed if (Roco Multimaus) throttle turned fully] Positioning the tags is critcal, as is positioning the 'aerial' - which in this case is simply the ID-12 IC with its internal antenna; and the Mk2 MERG Reader boards conveniently fit within the rails of the 45mm gauge brass track.
The 'Field Railway' wagons used for shunting have the 35mm glass tag glued lengthwise to their underside, between the axles, and appear to trigger reliably when the ID12 is at rail-level (but a problem when cleaning track!!) or just below .... if lying on the sleepers (Code332 rail) they are getting less reliable, and the 'full size' (Post) wagon in the pictures did not trigger when mounted on its underside 8-(
Similarly, problems were found when placing the Tags on the Locos ... and I SUSPECT but have not verified, that close proximity to the motor, and its magnetic field is killing the range ... also maybe the 'silent drive' pulses ???
We've had 'easier/more reliable results in 00; perhaps the range of distances involved is narrower. The LDT kit uses an external aerial, of Credit Card Size - maybe using the 2 connections to the external aerial on the IC would also have benficial effects.
[We've had 'similar' problems with 'Optical' IR Detection and visibly opaque plastics which seem to be transparent to Infra Red - white paper benath a model seems to improve that situation!]
A friend, in MERG, did some tests, mainly with the 12mm tags and N gauge a while back, and found speed to be a key factor: too fast and no detection. I can reproduce that on the shunting puzzle - but also you have to consider gaps between tags (too close and conflicts) and the inhibition of 're-reading and sending' the same data ... so if testing, always 'clear' the tag away fom the area, and preferably use a different tag next. On my 00 layout I have rather a lot of Tags waiting to be fitted - to 'all' stock - and hope to read the formations of trains as they depart from 'shunting/storage' locations ... that's 8 readers spread over 5m x 8m ... and the cable used for the serial signal back, and local power supply stability has also been found to be important (The MERG Concentrator Mk2 supplies the 5V to all 8 readers per concentrator)
AT PRESENT I do not see it as 'reliable enough' to actually CONTROL trains by position detection ... I am testing Optical on the Skandi Demo (3m long) layout for that .. and again the short distances between triggers can be a problem. [ By the time softrware debaiunce has occurred, you may be past the next sensor] . We have a 'backlog' of the Optical Detectors to fit to my main 00 layout (5m x 8m, 88 optical detectors planned -> 50+built) . I currently anticipate only using the positition detection, and RFid tag composition
info to 'assist' me in running the trains by 'offering' information which I can use .... and I see the RPi as ideal for that - even for spoken station announcements to match (afterall, the BBC Micro could do that 30 years ago !!!!,
and this is just the latest incarnation of that 'Beechingless Barnstaple' layout. So its definately a case of 'watch this space' (and or my website, or the MERG Journal etc)