Thursday, October 19, 2006

An Update

Wow, it seems like only a matter of weeks since I last updated this website; yet it's been nine months! Fear not, work hasn't been abandoned; though the particularly nice summer weather did reduce the time spent in the stuffy dark loft somewhat, along with my partner moving in.

Following on from the previous post, the framework for the baseboards has now been completed. I haven't yet attached the boards themselves to the top yet though, instead I've simply laid them on top and taken the opportunity to experiment with different pieces of equipment and ways of working.


My experiments with DCC have gone very well, I tried a couple of different products and have settled on those from Lenz. Of those I tried, they had the most complete range and seemed the most flexible. The expandability of the system impressed me, and the fact that they sell an off-the-shelf USB dongle to connect a computer directly to the XpressNet bus used by the other controllers won the day.

To that end, a simple loop was set up powered by a Lenz LZV100 command/power station and LH100 throttle. I've since expanded this adding a second LH90 throttle and the aforementioned LI-USB computer interface. The sheer elegance of being able to work multiple trains on the same track seems a far cry from the days of isolating sections.

My next experiments were with point control, I added a couple of Peco Electrofrog points to the layout (the extra wiring being a happy price to pay for better reliability) and connected them up to point motors from the same vendor. I initially obtained a Lenz LS150 accessory decoder to control these motors, but after a little investigation, replaced it with an LS110 and added a couple of Peco microswitches to the motors. The extra investment pays off with having accurate feedback of the state of the point, so when the hand of god strikes, the command station knows what you've done.

Thus being my first foray into the feedback bus of the Lenz equipment, I carried on and played with the block detection units; an LR101 feedback module and LB101 block detection modules were added and wired into a couple of sections of the track. Now I could tell from a throttle or the computer when a block was occupied or not, though sadly at this point there's no way to tell which train occupies the block. The LRC120 railcom display module only displays the number on its LED display, it doesn't report it back to the command station. I'm hoping that this will be a future expansion.

Block Management

I also obtained a couple of the different asymmetric DCC block management modules. The basic BM-1 was easy to wire in, and the locomotive fitted with the Lenz Gold decoder stopped in that section accordingly. However this unit would require additional work to be able to enable and disable at appropriate moments. The BM-2 has all the necessary extra circuitry wired in, and comes with a block detection circuit as well, so there's no need to stack multiple components on top of each other. I used the spare LS150 accessory decoder and wired the inputs of the BM-2 to one of its outputs, so the state of the signal can be easily controlled from the handset. I even wired a signal kit in to give a visual feedback.

I'm not completely happy with the signalling or block control at this point. Obviously I'm hoping that it'll be possible to not only detect the presence of a train, but also be able to read its address so that train location is more reliable. The block detection also does not feed directly back into the signalling/block control; it may be up to the computer to ensure that signals are only green if the block ahead is not occupied, and ensure that they fail to red otherwise.

Finally I'm a little worried that the automatic stopping that asymmetric DCC gives us is actually taking away some of the pleasure of the layout. If I'm controlling a train, I want to stop it at the red signal myself, not have it stop without me doing anything.

To this end, I think that the ideal solution will be to have the signals and the block control modules separate; the signals will be tied into the block occupancy detection, possibly even electronically rather than via computer. The block control module will be activated by the computer if it believes the block is occupied by one of its trains, and believes that the signal being approached is at danger.

The reason for using the block control module to do this, rather than just have the computer stop the train itself, is simply reliability. The computer lacks the eyes and finesse to know how to stop the train directly in front of the signal.


As I've experimented with the different possibilities, I've begun to change my mind about what I want to get out of the layout. I ended up drawing up a list of what interested me:

  • Having a busy sub-urban station on a main-line junction with numerous trains coming and going.

  • Wide range of different units, from long-distance expresses to local multiple units.

  • Owning models of trains that I've been on or seen in my lifetime.

  • Computer control of trains leaving me free to play with just one on a busy network.

  • Improving RTR models by adding realistic lighting and weathering.

  • Scenic opportunities around the railway.

What I realised was that, despite it being my experience of model railways so far, I wasn't really that interested in having one or two trains running round in a loop chasing each other. I'd be entirely happy with a chair that faced the station area, having trains come in and go out, and ignoring where they went in between; even if it was just round the corner to pack in the fiddle yard.

At this point it's probably worth talking a little more about the lift I had planned. I experimented a fair bit with different components and circuits and had a pretty good working idea about how to do it, but the more I thought about it, the less practical it seemed to be. The lower fiddle yard would be cramped and difficult to access if necessary, and the length of the lift would absolutely limit the length of train I was able to run on the layout.

These two things combined I decided to rethink the layout a little. Instead of a tail chaser, I decided that one side of the oval would be the station area, and the other side would just be a non-scenic fiddle yard. This gives the maximum amount of space for both, leaving the sides free for some scenic work on one side and a motive depot area on the other.

An open possibility is for a "branch line" to leave the main station, continue round the layout increasing in height as it does so, and terminating at a local station at a different level. This would provide a little extra interest to the layout.

Next Step

The next step is to clear away the experimental trackwork and wiring from on top of the boards, and the boards themselves. Then actually get them cut to shape and screwed down permanently, at this point I'll know exactly how much space I have so that I can start some proper layout plans.

Hopefully all this will be done over the next month or two, barring some business trips, and my mission for Warley will be looking for some inspiration as to the layout and picking up a few kits for things.

Monday, January 23, 2006


It's been a busy few weekends for me up in the loft. Before embarking on any grand plans for the layout, I decided the first thing to get out of the way was the framework for the baseboards. Given the more restrictive nature of a loft layout, I figured it would be much better to have this up first and worry about the minutae of point locating later.

I knew roughly that I wanted a tail-chaser at about 3ft height, however once I got up there I realised that this puts it pretty close to the roof at the back of the layout. By lowering it down to just 2ft6 high it gave me enough headroom to build backscene and has allowed me to consider a higher level for one of the track pairs.

The frame itself is made from 34mm square planed wood, simply because there was a huge quantity of this at a discounted price at my local B&Q. This was initially cut to fit in the "V" of the trusses and screwed in with No. 6 wood screws.

Frame screwed into truss, and cut ready for its mate.

Half-cuts were made in each piece to fit the perpendicular framework to be added later. It was about this point I remembered being taught that builders aren't precision engineers. While checking the alignment I realised that one of the supports was slightly out of position compared to the other. It measured up the same as the others, it turned out that the roof truss itself was out of position compared to the rest.

Checking the alignment with a spare piece that was cut wrong.

Once all of the supporting pieces were screwed into their respective trusses, I measured up the longer pieces that were to go across them. The middle of the three pieces has extra cuts made on the top to allow for cables to reach the back where the electronics will go.

These were all left in position, but not finally screwed down until the very last thing, to allow for any changes-of-mind I felt like having.

Framework cut and lying in place.

For the ends I decided to use a much longer piece that extended out into the middle of the loft, where legs would come down to the floor. Initially I hoped the legs themselves would be sufficient, but once up I decided to add braces in both directions to hold everything in shape. It payed off, the arrangement amply held my weight while I was sitting on it to screw everything else in position.

Legs on the longer frame, before braces were added.

While putting everything together I started thinking a little bit about the layout that was going to be on top of it, and I modified my plans appropriately. I'd pretty much decided that the far reach was going to be where the station and town area was to sit, so I placed the legs closer together with a strong support frame under them. I also decided to slightly increase the depth of this, so added a further beam going between each side.

The opposite side I decided would be a valley, so spaced the legs out much further apart and added extra framework onto the supports onto which the valley floor and a river could be modelled. This side also had two trusses very close together which I'd decided not to place framework between.

One corner of the "valley" side, note gap between frame and wall.

My thinking about this gap started because the rear tracks at that point were to be elevated, and cross the track in a viaduct; therefore there was no need to create such a rigid frame at the base level, instead the work would be done at a higher level. After thinking about it, I realised that by moving the rear tracks forwards I could use this gap as a handy way of getting trains down to an even lower-level fiddle yard that I'd been thinking of.

The current plan for the gap is that it will hold a "lift" that will be able to serve both floors, or if that doesn't work, a traditional ramp down to the lower level.

Friday, December 30, 2005


There's no denying that for those of us who find interest in such things, the sheer range of ways to control a model train layout these days is staggering. Not just in terms of train operation, but in the increased realism one can obtain through correct signalling and lighting throughout the system. In fact, much of this is what's got me interested in the hobby again.

Given that I have a reasonable amount of room for track and the green bits, about 4m x 4m in fact, I anticipated wanting more than just a few loops of track with individual old-style controllers for each loop. Certainly it has the benefit of simplicity, but it lacks that certain something for those of us who enjoy spending a weekend with a soldering iron and multimeter.

I started looking at the various methods of splitting the track into blocks and detecting the trains that passed over them, initially I thought about having a controller-per-block but quickly decided on a controller-per-train; electrically connecting it to the right block through automatic means. This certainly allows multiple trains to operate on each loop and would provide for realistic signalling and junction management. My idea was to link this all into a computer which could handle much of the monkey work so as an operator I could just concern myself with running trains around.

At this point I started scrounging around for a convenient power supply to test my idea, and found an old computer PSU under the cupboard; surprisingly suitable for my needs, except with rather more ampage than your typical shop-bought train set controller. This got me thinking about using a single power source for all of the track and trains, and instead just having a voltage-regulator for each train rather than a complete controller; much easier to build by hand.

I deliberately hadn't done much research into "off the shelf" systems yet, I wanted to see what I could come up with myself and then compare it to that which was readily available and see whether it was worth the extra to work to do it myself or whether the existing systems would fit my needs perfectly.

However some of the information about DCC must have entered my brain because it's about this point a friend and I started wondering just how small we could get the voltage-regulator, and whether we could put it inside the locomotives themselves. Over a couple of days this blossomed into having the track permanently wired with 15V DC, and inside the train having a small controller connected to an 802.11 radio receiver much like in the WiFi card of your laptop.

The obvious advantages to this would be that each train could be commanded to go at whatever speed wanted, and more particularly trains could occupy the same block within a depot or goods yard. The fact that coach lighting would be always on would be a huge bonus too, as a self-confessed LED fetishist one of my aims for the layout is to have as realistic and prolific lighting as possible.

I hadn't yet fully fleshed out plans for train location reporting, one of the ideas of using the radio chips was that the train could report back to the controller just as easily; perhaps by using a barcode scanner cut into the bottom it could read its location from the track as it passed over.

I described the ideas to my father, who strongly suggested I go look at DCC as what I had described was broadly similar, at least in terms of effect rather than the technical details. And thus after a thorough research of DCC, I've dropped my custom idea and decided that it's exactly the right thing for me.

Point control pretty much comes off the shelf, all points would be motored and controlled through accessory decoders like the Lenz LS100. At this point it's worth bringing the computer back into it, to behave as the IECC/power signalbox for the layout. Because each junction can be controlled and examined using the DCC system, it's worth placing them under central control to ensure that points cannot be changed underneath a passing locomotive. This could be accomplished with much wiring and circuitry, but it's far easier to simply connect a computer. A desired route would be booked on the computer and placed in a queue, with each route taken off the top once the junction is clear.

Block and signal control is still one thing that doesn't quite come off the shelf, so should still provide me some soldering interest. My current plans are to split the layout into blocks, just as I would have done with a more traditional approach. Each block will use something like the Lenz LB101 and LR101 combination to report back the occupancy of the blocks to the controller, and also provide a way of hooking in a signalling circuit. These I intend to make myself, combining them an accessory decoder so that a manual override to red may be set from a controller or by the computer.

Junction signal control can be either controlled from the computer via decoders, or wired in a more traditional manner; currently the traditional has a fair appeal as this surprisingly makes feather and theatre indicators easier with a handful of TTLs to determine which way the junction is set.

Stopping trains at signals and in stations is a much more interesting challenge. The block can be split into two parts, with a distant signal at the join between the two blocks; a train entering the second block while the distant is at caution (so the following signal is at danger) would need to slow down and stop. One solution to this is the Lenz BM range of modules which use "Asymmetric DCC" to place a SLOW or STOP command on the track unless bypassed. This initially seems like an ideal solution, except for one minor point, which is that it takes away some of the operational interest; I'd rather like to have to do that work myself for trains I'm controlling, and only have it performed automatically for trains the computer is controlling.

Still keeping the blocks split and the distant signals, it's relatively easy for the computer to detect when one if its trains passes a caution signal and issue the appropriate stop command. It can also provide a kind of AWS for me too by buzzing if a train that I'm controlling passes a caution signal, and maybe if I don't press a key within a few seconds, applying the brakes for me. This could still use the BM range of modules to actually perform the brake application, rather than sending a command to the train, to take advantage of the automatic stopping distance parts of the decoders; it also adds a fail-safe so the train stops, even if the computer has got muddled as to which train is where.

I've mentioned the idea a couple of times of the computer controlling trains, this should be easy enough to achieve with the computer booking routes across junctions itself as well as stopping the trains at signals and stations. The appeal of this is to create a busy layout as a background for the single train I'm operating manually, and especially for the glorious feeling of having to stop one's slow freight at a junction because the computer-controlled Pendolino has priority.

Wednesday, December 28, 2005

Locomotives and Rolling Stock

There's much discussion and writing about picking the prototype you want to model, but the underlying recommendation to it all is to model precisely what you want to. My personal interest is to try and capture the essence of the railway that I've known in my life, rather than a precise location or particular time period.

Being born in 1980 and growing up in Sussex means that my boyhood was dominated by Network SouthEast and slam-door rolling stock, sadly this is pretty scarce in the Ready-to-Run market so I suspect some kit buying and building will take place later on.

However I now live in Birmingham, and for a couple of years commuted to London and back every day; so more recent years have been dominated by the West Coast Main Line and cross-country services.

The name of the layout reflects this mix, and is the name of the main station I hope to build; something of a mix of Clapham Junction in South London, firmly in the NSE region, and Watford Junction just north of London on the WCML.

Anyway on with the pictures of the locomotives and rolling stock I've collected so far; sorry in advance at the quality of some of these pictures, they've just been taken for cataloguing and insurance purposes, I plan to stage some proper pictures of them all later on.

No modern image of the WCML would be without one of these, the Dapol Pendolino; this one's 390010 "Commonwealth Games". Sadly they only came in 4-coach rakes, this model could really do with a couple of plain coaches to increase the length compared to the longer Bachmann Super Voyager model which is a shorter train in reality.

And here's that train, the 5-car tilting Bachmann Super Voyager; 221101 "Louis Bleriot".

Completing the set is the 4-car non-tilting Bachmann Voyager; 220001 "Maiden Voyager".

I still have a fondness for the old WCML trains and red/black Virgin livery, so have been building up an older-style train. The mix of Mk 2 and Mk 3 coaches here isn't a mistake, the Euston-Wolverhampton trains ran in exactly this formation. The mistake is that the Mk 2 brake was only used on cross-country trains, the WCML trains had an 82xxx DVT instead; Hornby made one for a while, though sadly discontinued, it's firmly in my wishlist of things to get.

Close-up of the locomotive on the above train, Hornby R2290: 86248 "Sir Clwyd - County of Clwyd".

I'll always remember seeing my first Eurostar at Tonbridge station, speeding through it looked positively alien and beautiful compared to the slam-door I was waiting for. This is the 4-car Hornby R1013; 373219/220, my wishlist includes the extra two saloon coaches to bring this up to a 6-car train.

Now we move onto the smaller DMU stock; here's the Bachmann 3-car 170637 in Central Trains livery, a common sight around where I live.

I've always liked this South West Trains livery, so had to get the Bachmann 2-car 170302; it also drags the stock back towards the old NSE region as we've been getting positively WCML and Midlands so far.

And one I've had for a while, the Bachmann 159001 "City of Exeter" DMU; I deliberately picked the NSE livery as much of the old stock was never repainted into the liveries of the new owners, so this could fit in today without much stretching.

This locomotive was picked just for its livery, they actually ran completely out of the area I knew but with a bit of stretching we could believe it's been reassigned as a Thunderbird or similar. It's the weathered Hornby R2350, 50045 "Achilles".

Shown here again with a rake of old Lima NSE-livery coaches. I suspect these will largely live in sidings to add a bit of sense of history to the layout, there are plenty of stored coaches around the region anyway.

The Bachmann model of the Virgin 57301 "Scott Tracy", complete with the "International Rescue" logo and name plate missing from the Heijan model of this locomotive. DCC control would let this be used in actual formation too, if the Pendolino and Voyagers had couplings, that is.

I like the level of detail on these Bachmann BR Mk1 Pullman coaches, so have been collecting them with just the Second Bar to go. One great thing about the modern image is that you don't need to stretch to find an excuse to run older stock, you can just claim it's a special or preservation working. This train could be taking people on a Hospitality trip to Silverstone, which is exactly how I rode on my first Pullman train.

I hadn't originally planned to include freight in my layout, instead just focusing on passenger services. Two things persuaded me to do so, the first is simply the increased operational interest of running freight and the shunting one has to do; the second was the Bachmann models of the Class 66 "sheds" which for me really capture the awesome size and presence of these locomotives; here's 66135 in EWS livery.

And again, pulling a rake of tankers, most likely between yards.

And 66200 "Railway Heritage Committee", also in EWS livery. I picked this operator simply because it's the one I see the most every day, especially as they have a large depot just north of where I live.

66220 again, pulling a rake of tarmac hoppers.

If you're going to have freight, you need a shunter. One great thing is that operators have been recently repainting some locomotives into "heritage liveries", which gives us an excuse to add a bit of colour and variety to the layout. Hornby R2418, 08402 doesn't look at all out of place in its BR Blue livery.

Here it is pulling a rake of various individual trucks, mostly given to me as gifts. On first glance, you don't notice that the adjacent Shell tankers are from two different companies (Bachmann and Hornby) but on closer inspection there are lots of detail differences.

Those heritage repaints can also give us an excuse to take some of our older stock out of storage and pretend that it's running today and has just been repainted; provided one doesn't look too closely at the running numbers, of course. Here's an old Hornby R073 D1670 "Mammoth", in real life the locomotive itself has been renumbered, renamed and scrapped; but we can pretend it's still about and has just been repainted again. Double-heading with it is the Hornby R288 47620 "Windsor Castle", in real life this locomotive was refurbished to a 47/8 then 47/7 and for a while pulled the Royal Train under the name of "Prince Henry"; since retiring perhaps it was repainted and given its old name back (even though that name is currently given to a different locomotive altogether).

This is perhaps stretching the excuse a little far, a Hornby R556 set I've had since I was a boy. I do intend to have an HST set on the layout, but I suspect I'll go for one in a current livery with more coaches than this older set.

Finally purely for novelty value, a Bachmann motorised model of a Plasser Tamper Machine. The motor is tiny, I suspect this will be problematic to chip.

Loft preparation

Deciding that the layout would live in the loft was the easy bit, like most houses my loft isn't quite the ideal place to simply drop in trains and run them, it first needed to be prepared; this means flooring, power and lighting. At this point most of the literature tells you seek profession advice and make sure the loft will take the weight. I've often suspected this is so you can't blame the author when your house comes down around your ears, that builders are thinking of a holiday in the Bahamas when writing you a quote, and that most of us just go for it anyway. Still probably an idea to at least run your plans past a friend "in the trade" just to make sure they don't choke on their tea or something.

Once I was sure I wasn't going to end up with good sky views, it was time to pop along to the local B&Q and get the various bits I needed to get the loft ready for receiving the railway.

The first consideration was access to the loft, with the best solution being one of the many styles of stowable "loft ladders"; I picked one to my liking and that would fit in with my requirements, in particular the one I chose has an interior vertical clearance slightly shorter than the minimal height for the baseboards saving me the trouble of having some kind of lift-up section or similar.

Installing a loft ladder required installing a new loft hatch; the house came with one of the "polystyrene plug" hatches that you simply push up and out of the way to gain access. With a loft-ladder and regular access I decided to change it one that hinged downwards instead, and found one of roughly the right size and build for the existing hole.

As you can see in the pictures the new hatch needed slight cutting to fit into the hole, and there was also a slight gap which needed filling in; this still needs to be filed smooth and painted. I also added extra cross-joists to form a frame around the hatch, and in particular support the flooring and the loft-ladder itself. Otherwise installation of both was pretty much done by following the instructions.

New loft hatch installed

Stowed loft ladder attached to new cross-joist.

Additional new cross-joist.

Ladder deployed.

Flooring was bought in packs and pretty easy to lay, taking care to stagger the joins and mark where both joists and cables lay beneath. The middle of the loft has been completely floored from side to side, where the bulk of walking will be; I also decided to add about a metre of flooring past the truss on each side to provide better storage than balancing things across joists.

Loft flooring down.

Flooring around the hatch.

Storage space behind the trusses.

Used storage space behind the trusses

Power to the loft was simple to arrange, the loft is plugged into an ordinary 13amp socket in the main bedroom with a heavy duty cable taken through the ceiling in the airing cupboard to keep it out of sight. A fused switch in the loft allows everything to be turned on and off in one place, rather than at the plug socket downstairs which is actually behind a chest of drawers; I chose the bigger switch style because it's simply easier to find in the dark than a typical 13-amp switch. This cable then ends with two plug sockets, into which are plugged 4-gangs to each end of the loft; obviously one needs to remember to put this on once the cable has been put into place.

Lighting is provided by a small fluorescent strip-light at each end of the loft, which are plugged into the gangs; these only take a couple of amps, leaving plenty free for the trains and even a heater for the winter months.

Power switch.

Light at one end of loft.


As most of the literature on model railway building says, one of the key factors in one's layout is where it's going to live. I guess I'm a little more fortunate than most as I own my own (small) house and live on my own, so certain factors such as "explaining to the wife" don't need consideration; but on the other hand, it's only a small house with not much space to spare (the second bedroom already serving as an office from where I work).

My initial plan was to follow the thinking of a portable layout made up of individual base-boards that could be detached from each other and stored when not in use. The advantage to this is obvious as it doesn't require any permanent home, except for storage space. However that turned out to be an issue, even building it from 2x4 boards, these take up a fair amount of space, especially when they have buildings etc. on them. In fact, about the only places to store them would have been a garden shed or in the loft.

My next thoughts were to actually extend the garden shed idea, as I'd need to build one anyway, and design the layout to be put up in the garden and stored in the shed. This was obviously at odds with the time I'd be spending on the hobby, ie. the winter and rainy weekends, so would have probably been a non-starter.

Finally I thought about the loft, which I'd dismissed until now as I didn't expect it to be large (or high) enough. A chance need to get at a box which had been pushed right to the back, and require me to actually climb into the loft rather than just push and pull things from the hatch, changed my mind. I discovered that the headroom was sufficient for me to walk around in the middle without needing to duck or bow my head, and that the size of it was larger than I thought due to the lack of water tank or similar trappings.

In fact, once measured up I realised I'd have enough room for a decent sized layout with plenty of room underneath and behind the baseboards for storing those things which previously lived there. Of course, it needed a little improving first which I shall touch on in the next post.


Probably the easiest question to answer first is how I came to decide to build a model railway layout; like most people, I was introduced to them as a boy by my dad and had a small collection of very-abused trains and coaches which ran around a board in my bedroom and later-on a couple of loops of track in the loft. His interest in them has continued to this day, and since moving he's restarted building a layout for himself, but mine waned and with all the business of moving out of home, finding a job, and generally getting on with life, it was put to the back of my mind.

Well, not entirely to the back, I still every now and then picked up a locomotive or train pack that I liked and kept them hidden away for when I wanted them.

The decision that it was time to get them out and build a layout for them happened a few months ago; by day I work from home as a computer programmer, which combined with the social parts of the Internet makes it very hard to "get off" the computer and do something else instead. To try and combat this, I've taken up a few hobbies and past-times this year to give me a variety of other things I can do. Constructing a model train layout and all of the building work and intricate detailing involved seemed like the perfect thing for rainy and winter days.

Then a few weeks ago I went, along with my dad, to the Warley exhibition at the NEC which pretty much fired up my imagination and also let me increase the size of my collection somewhat. Now with all the ideas in place and a good collection of trains to run, it's time to start planning and building it; being a geek, I obviously can't help wanting to write about it as I do so. I hope that this will interest those others who enjoy model railways, and serve as useful reading for anyone starting out like me.