Michael Hodgson

Ooooh! Reminds of the LT frames I used to see being worked by a ghost signalman! A lever is just a switch made of cast iron, but the problem with modelling lever frames is that you shouldn't be able to move any lever when that operation logically isn't allowed. Mechanical locking stops you from doing so, and to get the right "feel" you have to provide electrical locking instead (logic plus a solenoid lock on every lever). It's easier to use simple switches and simply ignore any change which isn't permitted and probably sound an audible alarm instead, but that just doesn't have the feel of a lever frame. I'd be interested to learn how one overrides a servo linkage to work a lever manually and still have the effect of working locking.

That's very interesting, I hadn't seen that but I have not really spent much time on the iTrain forum. I ought to find time to look at that, and I rather think I ought to join the MERG Automation SIG and spend some time there too. If the itrain developer does produce something, that will save others from re-inventing the wheel. I appreciate that he needs an idea of whether it will be worth the investment of his effort but I'm not in a position to help him evaluate either the demand or the size of the job. The current version of iTrain is quite recent so I expect any major release is a good way off, but the website talks about pricing policy should there be another major release in the future, which I take as a hint that there's more to come. I must say I was wondering what enhancements one might wish to make to the product and was hard put to come up with much more than support for any new command stations etc, further interfaces like Cbus, maybe a few tweaks to make things even more user friendly. Maybe there are big ideas in DCC yet to come from the manufacturers and he will have to support them - I doubt that the founders of DCC would have anticipated what it can now do As a former MERG Treasurer, I would comment that the most active members who design circuits (maybe less than 1% of MERG membership) seem primarily interested in the electronics rather more than their railways, perhaps even more so than what practical use their circuit is to other modellers - they are often meeting some very specific technical requirement of their own, and yes they do put a lot of effort into doing things cost-effectively and this certainly shows in the choice of components used in the kits. The Journal often describes competing circuit designs or ideas from different people (not necessarily resulting in kits) with similar functionality - a lot of train detection devices for example. Whilst this individualist approach may appear to non-members to be a lack of joined up thinking, Cbus is one area where different ideas do get pulled together and made compatible. I suspect the other 99% of the MERG membership just want to use the kits and the ideas of those whose electronic knowledge is way above our own. My gut instinct is that CBus support ought to be worth implementing, ought not to be too difficult to develop BUT I really haven't looked at it, so there could well be some major issue that I haven't thought of especially as others with more knowledge than me have considered it.

No reason why you shouldn't use LEDs and buttons or switches - some panels on the prototype use that approach rather than conventional instruments. It should be easy enough to drive a block instrument using something like one output of a 4018 module for Line Clear, another for Train on Line (both off for Normal) Over the years I have collected a few miniature block instruments that have been home made by other modellers, and of course there is always the (collectable and now expensive) Triang block. Driving a bell is essentially the same as switching on a relay for a short period - itrain can send short pulses. More difficult is recognising what bell code has been input using a tapper as you essentially have to count and measure the relative gaps between pulses. It can be done using Strowger gear, so of course it can also be done electronically. It's easier putting a processor and software somewhere into the design, especially as the response to a given bell code (although it's often simple repetition) should depend on the status of various other variables.

I see no reason in principle why a banner could not be designed to be worked by a signal wire, but the railsigns website seems to suggest that they were electrically worked from the outset. The is a fail safe issue to consider in the design - it is a wrong side failure for a banner to be off whilst the repeated signal is at danger. Whilst it is still possible, I think the risk of this is a lot lower if it's done electrically using a solenoid rather than mechanically, and probably much lower still with the modern LED types.

Think yourself lucky he doesn't bring it in alive and then let it go!! Our late lamented tabby used to stand outside the back door asking to come in. He would immediately use the litter tray then demand the door be opened to let him back out. But then I also prefer to have an inside toilet rather than burying it in next door's garden!

I've used elasticated thread (source unknown) in O gauge, problem is finding something to tie it to (not the buffers!). Ordinary thread always seems to end up loose. I was never in the Scouts - maybe others can tie better knots than me. I like the idea of model ships suppliers as a source though.

Yes, this is important if you have a push-pull formation, or a DMU with a dummy power car at only one end. You don't want to stop with the dummy car past the signal at danger. However Itrain allows you to specify how far from the buffers the vehicle's first pickup is (it's covered in one of the tutorial videos), so this feature could be used as a workaround for precision stopping. In the case of an offset pickup (tender wheels on one side, loco on the other for example), the loco will be detected when an axle of both vehicles has entered on the section.

OK, so if you can arrange that the last vehicle of the rake is one with lights then you're already there !

If you want reliable detection of an unlit vehicle I would fit one resistive axle to each bogie or both axles of a 4 wheeler, but I don't see a need to fit every vehicle in a train

If you've got a lit coach drawing its power from the rails you don't need resistive axles. The current drawn by the lights will do. If you use battery-powered lights - various modules on ebay - those of course aren't detected. Likewise, if you use detectors other than current based (eg infra red beams, reed switches etc) then the resistive axle won't help anyway.

I must admit I'd never heard of Tunnocks before, but they sound incompatible with my Slimming World diet which I started just over a year ago when I was told I had type 2 diabetes. That has worked, I have reached my target weight (only a couple of stones less) in January and the doc has even told me I don't need to take his diabetes pills any more, so I probably shouldn't start eating these! I do like this canny Scots approach to model tarpaulins though. The one thing about tarpaulins though is that they should be roped to the wagon and I haven't found a convincing way of doing that in scales smaller than O, and even that's not easy. The same goes for loads that should be securely chained, or I suppose the modern image is the ratchet strap.

Future layouts? I've got to get the current one wired first!! If you include a helix like Andy's, you'd best include working catch points and sand drags which were much more common in the 1930s! I can see me needing a couple of helices. The modeller's problem with Block working is the same as BR's - manning. You need another operator to send signals to and who will reply appropriately. Somewhere I may still have some simulator software that talks to a full size block instrument and can respond with the correct bell signals and block indications. It was written in Qbasic and ran on an old 286 PC, and later I wrote something similar to run on a PIC (even including permissive block). I think I could probably replicate full block working in iTrain, but that will be some years away. Single needle telegraph instruments as used here on the GNR lines until the 1970s could similarly be simulated. The latest of Bob & Iain's series of iTrain videos points towards certain features I would envisage using - but their series is evidently not yet finished, so maybe they've beaten me to it ! My ideal layout (if I were to live that long) would be one which has several signal boxes and capable of running fully automatically but with the ability as required to switch any or as many of the signal boxes as desired to manual operation with full block working to adjacent boxes which if not so opened and manned would be capable of simulating the correct block responses. Clear an appropriate signal and locos/trains would spontaneously move accordingly without needing an operator to drive them - unlike most modellers, I don't want to be a train driver. As I see it, itrain appears capable of doing this and I would expect that TC can also do it. The biggest practical limitation I foresee is coupling unreliability, difficulties with uncouplers etc - so I don't envisage that ever being feasible for me other than on the basis of all shunting remaining manual.

Best runaway I've ever seen was years ago on a fully signalled 3-rail O gauge layout (track circuits worked by using axles to short out the two running rails, though they only displayed lights on signal box diagram). Operator at the branch terminus was shunting one wagon on the downhill side of loco, contrary to Rules & regulations, and the wagon uncoupled and ran away. So he sent the correct single line emergency bell signal 2-5-5 Train or vehicles running away. Operator at the bottom of the hill (when he had recovered from saying You What?) took the correct decision to divert the runaway into his bay platform to avoid collision with local passenger train waiting in the loop. Unfortunately he was about half a second too slow in acting and he also pulled the lever for the wrong points. Wagon had already passed both points and its sprung buffers collided with the sprung buffers of loco standing in the branch loop, so it energetically bounced back off towards the crossover which was now set to divert it onto the Down Main line - also downhill! So he now sent the emergency signal 4-5-5 Train or Vehicles Running Away in the right direction to a third operator at the next box on the Main LIne where it coasted to a stand in his Down Fast platform! It would be difficult to recreate if we tried!

You are obviously thinking like me in terms of fixed rakes, at least as regards passenger workings.

Reliable couplings are important anyway on an automated layout. Otherwise you spend all your time re-uniting divided trains, including rerailing after a runaway has caused an accident. The layout size limitation on how many trains you can run per operator on a conventional layout depends on how many simultaneous movements the human mind can cope with competently. On a fully automated layout, the limit becomes how much fire-fighting of such problems you have to do.

I will be using brake vans, but essentially it's me too - the same consideration applies to parcels traffic, and of course you should never show a tail lamp in the middle of the train as too many modern layouts now do - the one on the back of the loco being a case in point.

Unfortunately I suspect he is right in thinking it is "vapourware". I'm sure it can be done - but I don't know how difficult it would be.

On the real thing what they call "Track locking" is rather important for safety. Track circuits ensure you can't move a point whilst there's something standing on it. I find it goes against the grain that points are not detected although I know Itrain handles that requirement by inference from what left one block and did/didn't arrive on the next, and we aren't as safety critical as the big railway, which is why modellers don't see it as essential to have feedback to prove a point is set and locked in the correct position. Before the track circuit was invented of course, they relied on mechanical locking bars and the signalman using the Mk 1 human eyeball. If you want to display pointwork occupancy on an illuminated track diagram, it might be a use for something like the MERG Train on Track Indicators without necessarily being reported to your control software. Carriage lighting is OK for coaching stock, but goods wagons don't usually have lights. Do you intend to fit working tail lamps to brake vans? With iTrain if you have function-only chips in vans you could even display the correct lamps depending on whehther you are on a fast or slow line, and I wold assume you also do that in TC.

I might consider developing my own interface, but I have a lot of more pressing work to do before I need worry about that

Thank you for posting this very useful thread James. I am in the process of setting up a OO layout using Z21 and intending to buy iTrain (I am currently using the 2-month trial licence, which I recommend to anybody considering automation). I am really most impressed by this software and although it is complex with a very big learning curve, the videos by BlasterBob & Iain Morrison are extremely helpful in that they explain the functionality in such a way that they tackle first things first and so you can leave the complexities and niceties until you are ready for them. I am pleased to see you are happily using the 408LN/CS combination, and MERG cut-out for power districts which I also intend to use but not necessarily exclusively. I have considered Railcom, but whilst I understand its advantages I also see a large price tag in the form of railcom decoders when iTrain can keep track of what's where - after I've told it once. I haven't decided what decoder to standardise on, but haven't enough experience of DCC yet and have so far not had any trouble with the cheap Lais ones which seem to attract such poor reviews. Other brands cost double even before I start insisting on railcom. And I am certainly not going to fit a big fleet of locos with sound ! I also particularly like the MERG DTC08/DTC02 occupancy modules, although they are not directly compatible with Z21. The DTC08 design is now quite long in the tooth, having originally been developed by a MERG member to be used as part of a protocol called RPC which predated the MERG CBUS, using RS232/485 as used in serial ports. I use mainly solenoid point motors and find the 4018 modules up to the job and cost effective, but I also like the MERG CDUs The basic problem with systems like CBUS (and for that matter RPC, and PTP - yet another protocol designed by a MERG member) is that they are conceived primarily as a means of wiring the gubbins on baseboard(s) to switches/displays on conventional control panel(s). This suits many modellers just fine, but these devices are much less convenient if you want to drive it all by software like TC or iTrain. Perhaps this is partly what is behind a finding in this month's MERG journal that feedback from member renewals indicated that a third of the membership thought the range of kits was too focussed on CBUS. I don't really fancy trying to interface the MERG kits to iTrain via CBUS + JMRI - it's too inelegant an approach, so I am intrigued by your suggestion that iTrain might one day support CBUS and I hope you're right.

I do remember one price that went down as a result of decimalisation. The Severn Bridge toll had been half a crown but went down to 12p because new halfpence were too much bother to handle on every transaction and increasing the tolls needed an Act of Parliament!

Some of us till do ... and convert today's prices back to real money. I must have stamped thousands of savings account passbooks "converted to decimal 15th Feb 1971" . I can manage without the farthing - 960 of them to £1, so worth marginally more than a tenth of a new penny, and I remember the old French Franc (which became the new French Centime 60 years ago) of similar trivial value.was made of a very light metal (aluminium/zinc?)

Isn't it nice and peaceful in here? By the way, has anybody seen my hearing aid? I must have put it down somewhere.

You may well be right, but I really don't want them to do away with the £1 fee offers - I only list items over a tenner when they limit their fees to £1. I didn't really expect them to continue those offers as they make a lot more with their standard fees and they dont have any serious competition for sales of specialist things like model railways where the market size is only a tiny fraction of the population. If ebay do drop those offers it's going to be Quorn swapmeets for all of my decluttering of bulky railwayana.

Beer doesn't have a "use by" date, only a "best before" date, so it's safe to drink old beer...it doesn't taste as good if it's gone flat though. I believe kegs began to replace barrels just before WW2