The Late Shift

My film of the rest of the day at Port Deveron, August 1965.

My copyright for Trad as well as Trains 😉

Posted in GNSR in 4mm scale | Leave a comment

Train watching

I have been out early filming the first part of the day’s timetable operations at Port Deveron. Yes, we really did witness the sunrise….

No whizzy high-speed model trains here, just the ebb and flow of activity at a quiet provincial town station. Get the Thermos out, and settle in for the morning…

Sound levels are a bit hit-or-miss. It seems that my newly quiet locomotives don’t register well in on the recording, and it was not helped by the need to use two different tablets to record this, after the first one packed up. On the other hand, maybe this is closer to an appropriate level than is sometimes the case…

More instalments to follow in good time…

Posted in Uncategorized | Leave a comment

Towards a timetable: 2

The class 20 has just arrived from Cairnie Junction with a brake van, to collect the evening fish train to Aberdeen, which can be seen on the quay in the background. Once the 6.40pm to Elgin has departed, it will run round, leave the brake van near the goods shed and go looking for blue spot vans.

Work has continued on developing the timetable. In the previous post I outlined the assumptions made about likely traffic levels on the Port Deveron line. In the interim, I have been busy trying to place the required trains on the taktfahrplan. As suspected, it quickly gets out of hand: both in terms of ending up with too many trains, and with the general brain capacity required to hold all the necessary logistical issues and positions at one time.

Some of the key constraints are:

  • The general capacity at Port Deveron. The station is not large and will quickly become congested. Although there are in total seven sidings (including the bay platform), I don’t want to lose the uncrowded feel of the model, quite apart from the practical difficulties it will throw up.
  • Line capacity out of the station. PD is served by a single line, which notionally splits about 2 ¼ miles out of the station at Boyndie Junction, with one single line going to Portsoy and the other to Tillynaught. Both are about six miles from PD. The real journey from Banff to Tillynaught took about 15 minutes. This constrains arrival and departure intervals, since outbound trains need to clear Boyndie Junction before another can arrive from the other direction, while two trains working the same line have an absolute minimum interval at PD of 30 minutes between departure and the next arrival.
  • Attempting to mesh the timetable with the known moves along the Coast Line at Tillynaught, and in some cases the Glen Line at Cairnie Junction, so as to give connections to and from Aberdeen in particular. There is the possibility of an early departure for Elgin too, by changing onto a Glen Line train at Cairnie.
  • Wanting to incorporate a reasonable selection of interesting moves at the station without losing the overall realism dictated by the schedule.

Some decisions were made, and the proposed through Aberdeen service was an early casualty: the connections provided would already give PD a generous service as it is. I also decided it was permissible to shift the timings of some of the known trains along the coast line to make more space at PD when it came to connecting with them. It’s not unrealistic for timetable to be tweaked over the years, and I was starting with a composite in any case.

The end version of the taktfahrplan

So we have arrived at a service of five return trains to Tillynaught Junction as per the original Banff line service, which give connections to/from Aberdeen; the early morning one is extended to Cairnie to give the aforementioned Glen Line connection to Elgin. These will normally be formed of a dmu, though it is possible for substitution with a loco and one or two coaches, or even in the fullness of time, a railbus.

There are also three return trains from Elgin to PD, which are deemed to be extensions of the Buckie local service; these will usually be a type 2 on hauled stock. These travel by way of Portsoy.

In addition, there is one daily goods train from and to Elgin serving all stations en route, which arrives late morning and departs mid-afternoon. Scope is created for this to trip distillery wagons to/from Whyntie as a separate run if traffic warrants it.

Finally, we have the arrival of empty fish vans mid-morning which were detached from a coast line fish train at Tillynaught, and the return train departs mid-evening, having spent the day being loaded on the fish quay. I have yet to find out much about fishing fleet routines – though I suspect that loads were configured to travel overnight to reach the big city markets further south in the very early morning.

The sequence schedule that emerged. Some tweaks still needed.

I haven’t yet quite dealt with the options for low-catch days, though this could involve a few vans being attached to the afternoon Elgin passenger train as far as Portsoy, whence they could make their way to Aberdeen on a coast line train. There are light loco moves associated with the fish train, and one of these has the option of tripping earlier-arrived parcels vans up to the main line for their forwarding to Aberdeen.

Also to be resolved is what happens when the dmu needs to return to base for fuelling; it stables overnight at PD, but there is no accommodation for it at what is left of the local shed.

I have yet to figure out when PW trains might run – but it is likely they would run as required, or on days of timetable suspension in any case.

No doubting the location…

So PD has a service of eight passenger trains a day; over a fourteen hour period, that is perhaps a little generous, but not stretching the bounds of credibility too far. It might have been possible to cut some of the shuttle runs and rely on connections at Portsoy instead, but the effect of that would have been the dmu spending quite a lot of the day sitting at the terminus, which is pretty much what the real service did.

This exercise has certainly created scope for some more rigorous thinking about how to run trains on PD – but as I write I have yet to run the result to see if it works. Working backwards from plausible wider reality is interesting – we now need to see what real-time operational headaches it throws up.

I’m not intending to work to the clock, so this will be more of a sequence than a strict timetable. Judging by  my habitual operating sessions, it will probably take more than one day to cover the fourteen hours. I also have yet to decide whether to go the next step and employ some kind of chance card system to dictate loadings, motive power, special circumstances and so on. On the other hand, I don’t want to end up playing a board game here.


I’ve now run the sequence once; it mostly works, though there were a couple of occasions where I ended up in a spot, for example with a passenger loco that needed to run round and depart with its train, but unable to do so because the daily goods was standing in the loop with nowhere else long enough to recess it.

The 20 creeps across Station Road to the quay, to a chorus of squealing wheel flanges…

It has taken two sessions to work through the entire thing, though it did lead to extended running compared with my usual random dallying. That has to be good. I think it does give a much stronger sense of purpose to operations, not least because one knows what each train is supposed to be doing.

Waiting to head to Aberdeen to catch the overnight departure south.

The other benefit of working backwards from traffic needs is that one is in effect presented with an operational puzzle: the specific station moves are not specified – there is just a set of arrivals, departures and other requirements, and one can solve the puzzle in any number of ways. Building in optional moves, such as the trip working to the distillery and parcels van moves gives more permutations in this respect.

A little later, the last shuttle of the day from Tillynaught Junction arrived. It will shortly move across to the other road in the train shed to stable for the night.

All in all, a useful development, and one which is probably capable of more refinement yet.

Posted in GNSR in 4mm scale | Leave a comment

Towards a timetable

It was ever thus: a small layout inevitably has limited operational potential. Which is not a way of saying I’m tired of running Port Deveron already; more that I’m putting more thought into how to operate it to the optimum.

Rail lines in the Moray area. Thin lines are ex HR. The fictitious line from PD to Portsoy is shown, as is the “never-built” line to Macduff.

I’m definitely more of a layout builder than an operator, but that isn’t to say that running trains holds no interest, once all the gremlins have been ironed out. My childhood layouts all had supposed geographies, and (rather fanciful) timetables, probably as a rub-off from some of the timetable maniacs I used to spend time with, in my train spotting days. All the many since have at least had geographical rationales, theoretical traffic patterns and (mostly) appropriate motive power and rolling stock. The trouble is, I never stuck to the timetables that resulted for more than a few sessions before the prospect of endlessly repeating the same cycle paled.

So the issue is, can proper operation improve the running of a small railway, and can it be made sufficiently interesting that it enhances rather than inhibits the experience? There are plenty of books that say it can – though they still mostly deal with larger and more complex layouts than I have room for. The latest is Martin Nield’s Authentic Model Railway Operation from Wild Swan, which I’ve just finished reading. I’ve also been looking back over books on the subject by David Jenkinson and Bob Essery.

One of the problems is that they largely deal with steam age models; quite why no one has attempted something similar for railways in more recent periods is not clear. Perhaps it is assumed that they either lack interest or can simply be visited and watched. That aside, as Nield says, operations did not change very much between the turn of the 20th Century and the 1960s – which just about sees Port Deveron across the line.

Another difficulty is that there is nothing to describe systematically how a station like Port Deveron and its line would have been worked – I mean at the level of individual movements – in the early diesel era. What happens, for example, to the train engine of an incoming goods? Would it do some shunting and work the return straight back out, even if there was no traffic? Would it head off light engine either to the (imaginary) loco shed, or even back up the line to the junction, six miles away – or further? What is the correct sequence for moving a brake van from one end of a train to the other? Would it stay in the consist during shunting, or be put aside first? All things that would be useful to know in the interests of greater operational accuracy.

But all these imponderables aside, I have set out to see what I can do to produce something like a convincing working timetable. There are lots of assumptions, and I will be very pleased to hear from anyone who can correct them or provide additional information.

Firstly, I have considered the likely traffic patterns; since my original map of the area was drawn, I have decided that the line from Port Deveron (Banff) to Tillynaught Junction was built after all, and that later a third side of the triangle was added, allowing trains from the west to reach PD from Portsoy. (PD is, in any case, assumed to be a rather more important railway centre than was Banff – and such cut-off lines were built in the area, for example at Forres and Grange. I’ve also assumed that the line from Inveramsay to Macduff never got further than Turriff, thus making Macduff too, dependent on PD’s station…)

Therefore, passenger traffic is likely to consist of:

  • A local service from PD to the junctions at Tillynaught and/or Portsoy to connect with Coast Line trains between Elgin East and Aberdeen; it is conceivable that this service might have occasionally extended to Cairnie Junction to make a connection with Aberdeen trains using the Glen line via Dufftown and Craigellachie. The timetable for the Banff-Tillynaught shuttle for 1960 is to hand and consisted of five round trips a day.
  • An extension of the Coast Line local services that ran from Elgin East to Buckie. I have no information on how often they ran – but it is possible they could have sometimes extended to PD had that west-facing junction existed.
  • It is also conceivable that there might have been a daily through train or portion to/from either Aberdeen or Inverness.

All of the above combine to produce far too many trains to be justified, so some decisions will need to be made.

On the goods side, the most obvious flow is fish and seafood from the harbour at PD. This would need to find its way to Aberdeen for early evening to attach to the overnight fish trains heading south down the ECML. However, information gathered from books on the Buchan Lines (Aberdeen to Peterhead and Fraserburgh) suggest that traffic levels fluctuated, as presumably did train times due to the tide patterns – and hence the trains were irregular, and sometimes consisted of additions to passenger services when loads were low. The general flow is, however, time critical.

The second goods flow would be general merchandise, including coal for the town, and agricultural supplies such as fertiliser and machinery. Outbound, a major cargo would be vanloads of potatoes, this being a major area for growing them.

The third flow on the line is to/from the Whyntie Distillery, rail served by a trailing connection about three miles up the line. In reality, there are or were several small distilleries in the area, and the train serving this one would need to reverse at PD. There is no visible difference between loaded and empty grain hoppers, and likewise open wagons loaded with whisky barrels. Occasional vans and perhaps coal supplies would also be needed; it is probable that this would all run as part of the daily goods, though large volumes might sometimes warrant a second train. The recent fiddle yard extension will make longer goods trains easier to accommodate.

In addition to the above, there would be small numbers of oil tanks for the harbour fishing fleet (either in the daily goods, or even attached to passenger trains, Mallaig-style?) and the town might possibly justify the occasional short parcels working either to attach Aberdeen-bound vans at the junction, or through to Elgin. There might also be the occasional PW train.

So, all in all, without too much distortion of probabilities, there is a fair amount of potential traffic.

The next stage has been to gather what information I could about real workings – not easy in an area like this. I compiled a list of trains by year from the captions of as many photographs as possible, and was able to find a few documents online from sites such as Railscot. Thus the Banff-Tillynaught shuttle timetable is known, as are a few of the coast line trains. But nothing definitive as yet about goods workings. From this, I have started to compile a Taktfahrplan-style timetable diagram to see how modifications and additions might be made to arrive at an overall timetable for Port Deveron. More on that another time.

Early representation of the area’s passenger trains. The Thick line is the 1960 Banff line shuttle, which is known. The thin lines are other trains, times taken from photos etc, and given approximate trajectories using average speeds calculated from local timetables, and distances from same or Google Earth, where the route can be traced. The fact that the shuttle meets some trains at Tillynaught suggests I am getting it right – but there are clearly others missing, not known. Goods trains will be slotted between the passengers in due course.

In the meantime, here is a clip of the real thing which I recently came across – albeit of the Glen line rather than the Coast line. 10’00” onwards is a gem: what it is all about…

Posted in GNSR in 4mm scale | Leave a comment

As You Were…

Darned new-fangled North British must have failed again…

The title of this post was going to be “As WE were” – and then of course, I remembered that I myself have but a tenuous grip on the era that Port Deveron, as modelled, represents. Born 57 years ago tomorrow, I must have been an alert child, because I do have some distinct memories of the late Sixties. I remember, for instance, when the old-style traffic lights were replaced with modern Pelican-crossing style ones in our local town centre. I do remember, too, seeing others of the pre-Worboys signs – though of course they lingered for quite a while, and just a few are still extant. Geoff Kent’s new book Unconsidered Trifles from Wild Swan, which I recently purchased is a delightful collection of pictures of such small details.

Your carriage awaits…

Woe is me! I also distinctly remember – on the long and slow non-Motorway trips to grandparents in Leicestershire and Dorset, seeing the rusting remains of the Somerset and Dorset line as Holes Bay in Poole and Shepton Mallet. I remember too, the similar remains of the Great Central which still stood on our regular trips into Leicester. And I remember too, seeing a few locomotives in very faded green livery with the old British Railways totem, still present in about 1975. Not to mention the history-pieces I must have witnessed from my pushchair in about 1967, when my mother used to take me to Taunton Station to watch the trains. If only, as we all no doubt think, I had realised what I was seeing….

Anyway, I’m not much of a nostalgic really, for all that I am now modelling period railways. My main criterion for doing so was simply something that would create lots of opportunity for atmospheric and characterful modelling – and I’m sorry to say I see few bang-up-to-date layouts that do that. Maybe railway modelling really is an exercise in nostalgic recognition after all…

All of which is by way of bringing the news of PD up to date. There are still plenty of small scenic jobs to do, but I have had a pause from many of them, only recently resumed. In the meantime, I have been making a conscious effort to address the actual railway content of the model to try to bring it up to something more consistent with my aspired-to scenic standards.

The road sign is from Shire Scenes, with my own image scrounged from the internet and printed onto a sticky label.

There are now four sound-equipped locomotives, and I am still pleased with the impact of the added dimension, especially now I have found out how to reduce the volume significantly, and I’m getting better at using the various effects too. There are plenty that will never get a look-in however; less is definitely more.

I took delivery of a substantial selection of Hunt couplings, on the grounds that they warranted a more extended trial. I must say, however, that I am not entirely overwhelmed. There are now several variants of these, and they are not all compatible with each other. The latest ones have overcome the problem of repelling poles, but they still do not align well with the earlier versions, bought to accommodate different coupling mounts on the wagons, tending to push wagons off-centre. What is more concerning, even a slight difference in height means that the lower wagon is likely to be lifted a millimetre or two off the rails, with the inevitable consequences. Sadly, the rigid effect between wagons has not solved my shunting problem either, so I am now pondering a return to S&W across the board. I may have to accept slightly larger spacing between wagons as a result, though.

I have also been building: two ex-LNER planked fish vans have joined the three blue spot ones, and all have been weathered. The birthday tomorrow is expected to yield a couple more wagons, to build rather more diversity than the current stock.

Clutter and dirty wagons

Weathering has also been carried out on the rake of grain hoppers and the three oil tanks; I feel that practice is indeed yielding better results in this respect.

Perhaps the most significant work of the rolling stock programme to date has been the dismantling of the three Stanier coaches, and the painting round of the glazing to remove the prismatic effect, as described by Malcolm Nevitt in MRJ 167, a task I have been summoning the courage to try for some time. It has made a significant difference, as has the removal of the lurid seat colour, and the addition of passengers, just visible in one of the photos.

The loss of the prismatic effect in the windows of the Staniers is a big improvement, as are the uneconomically small clientele just visible inside.

Also on the motive power front, the arrival of sound decoders has liberated some Zimo ones, and I have used these to resurrect a couple of my elderly steam locos; the Standard 4 is now performing well, if occasionally, while the Black Five (which were regular performers on the ex-LNER Moray Coast route) is getting there, but a bit stiff, and in need of some cosmetic repairs after many years in store.

A small delivery of ModelU’s finest arrived a few days ago; these are destined for the class 105 when I open it tomorrow to give it its voice. It’s a devil to dismantle, so all interior jobs will be done in one swoop. For the Staniers, I settled on a collection of rather more budget-friendly white metal figures from Monty’s via Dart Castings. These have painted up sufficiently well for a darkish interior – but my, it certainly reminded me how far we have come in recent years with the quality of such details.

The Monty’s figures in the foreground are not bad as white-metal castings go, but Alan’s output in the background has brought us on by light years…

The same consignment also contained etched road signs and bicycles from the Shire Scenes range, the former of which can be seen installed in a couple of the pictures; a pleasing detail.

On a more prosaic note, there has been some work on the fiddle yard extension: the two cassettes have now been combined and widened to present one fold-over non-moving deck; this can still be removed and stored under the layout, but has finally solved issues with the curved track joint to the layout proper. We now have the potential for four/five coach trains or around 15 goods vehicles; what’s more, trains arrive on the layout at line speed, rather than accelerating as was the case – all in all, scope for a significant operational improvement, which will in due course be completed by adding turnouts and a total of three lines to the yard. I may cheat and use Peco points for these.

The new fiddle yard extension; this has doubled the useable length. Track will eventually be slewed to provide three roads. The fascia has now been finished across the upper buildings (this section is removable) and the baseboard framework blacked in to tidy things up visually.

Finally, I am giving more thought to the operation of the layout; whether some kind of timetable would be an improvement on my usual occasional shuttles. But those early memories have unfortunately not left me with much real insight into how a line like this would have been worked – other than the suspicion that typical model railway practice probably ain’t it. If there is anyone out there who can offer some consultancy on this, then I would be all ears….

Posted in GNSR in 4mm scale | 2 Comments

Getting hitched

Locomotives are the easy bit…

Even though I’ve worked in 4mm scale many times before and am finding it pleasing to be back again, switching scales has not been without its issues. Part of the reason for this has been the resolution to try to get things right that I have never done before, where necessary adopting new and hopefully finer standards in the process.

Generally speaking, I’ve maintained this principle with some success while doing the main building work over the last year. There remain, however, some issues that are not yet solved, and some which have only become apparent with experience.  The main ones are the fiddle yard with accompanying two-foot radius lead-in curve, and the thorny matter of couplings. This post is mostly about the latter – though it is worth providing a short up-date on the former too, as it has a significant bearing on the problem.

My plan for a hinge-out fiddle yard has generally worked on a structural level – it is indeed easy and quick to use, and locks firmly in place in the open position. But I rather under-estimated the poor visual impact of having a gaping void under the town when in the open position; this was rectified by providing a push-in blanking plate which restores visual completeness, but at the cost of access to what was meant to be the stock storage area. In the event, this part has not worked so well – the space under the board is not in any case big enough to store many stock cassettes, especially now that it is in part occupied by signal mechanisms and the mains power block for the layout. And the quantity of stock is mysteriously growing beyond what was anticipated…

More generally, I am constrained by a severe lack of capacity in the room for storing stock – mostly at present it lives in plastic boxes in one of the under-layout cupboards when not required – and the cassettes are really not being used. There is simply nowhere convenient to put the ones not in use at any one time.

I don’t particularly mind assembling trains one at a time, and this layout hardly requires a high-intensity feed. But more of an issue is handling and storage damage, to which couplings are particularly vulnerable.

The fiddle yard, incidentally, has also proved frustratingly cramped, and plans are afoot to increase its length in the near future – the dry run of which suggests it will make a major difference to ease of train composition, viable train length and use in general.

But the coupling issue remains. I know I’m hardly alone in finding this problematic. I have used Spratt and Winkle auto couplings before; the 4mm versions are grotesquely prominent, but the 3mm/finescale ones are more acceptable, and I have continued to fit them to more stock.

Spratt and Winkle couplings are a reasonably unobtrusive choice, and function well provided alignment is maintained.

The plan was to use these at points where vehicles need to be separated, and something else within cuts. I conducted major surgery on the coaching stock some time ago, which involved cutting the outer bogie stretchers, to body-mount S&W to them – a matter which the manufacturer neatly side-steps. It has worked, however – the bogies are fine, and the rakes couple and uncouple reliably when on a running round manoeuvre. I did however modify the paddle counter-balances by folding them over on themselves along their long axis, so that they did not foul the bogies on the curves. In the end I fitted only the hooks to the coaches, as they only ever need to couple to the loop on the locos. This is quite neat. Within the rake, I have used Hornby close couplings, in conjunction with Modellers’ Mecca corridor connections, all of which function very well, and prevent any slop or wobble.

Hornby close couplings are working well within rakes.
S&Ws are equally working well at the ends of rakes.

I also fitted S&Ws to a selection of the wagon fleet, and this is where things have gone less well. Even without the considerable amount of fiddly work in fitting the things, they have proved prone to misalignment – it’s not easy to mount the couplings reliably to a mixed wagon fleet, even with the use of 5522 models’ mounting plates, let alone keep the loops level and the hooks straight during handling and storage. What’s more, it is not unknown for a coupling to come adrift, despite what I thought was secure epoxy resin anchoring.

Getting S&Ws to function through the ploughs on the Shark was a challenge – but it works…
One-ended fitting has some benefits – but at the cost of needing to run all vehicles one way round. And they’re still vulnerable to distortion. Buffer heads awaited!

Reliable shunting round the lead curve (2ft radius) which also serves as a head-shunt, has not yet been fully achieved. Having done exhaustive investigations, I concluded that minor variations between couplings is part of the problem, as is the occasional buffer-lock between some permutations of vehicles – and especially the leading vehicle, where the hooks have a tendency to tangle with the locos’ buffer beam detailing. I am loathe, though, to increase the vehicle spacing to something ridiculous – the whole point was to avoid this.

I have been wondering what to do. Proprietary tension-lock couplings are notoriously visually intrusive; the modern small version are somewhat better, but still not ideal – and it seems that even combinations of various types of tension-lock are not without their problems. So much for the ability to plug them into NEM pockets; not all vehicles have these anyway  – and they are hardly within the finescale ethos I have been trying to work towards…

Above and below: tension locks of various sorts – visually intrusive and not as reliable in compression as one might expect…

I have even pondered using 3-link couplings but suspect they would make the problems worse – and I think Alex Jacksons are a non-starter given the constraints here.

A search online revealed a small outfit called James’ Trains Parts that offers 3D printed plug-in instanters and 3-links which look interesting but the choice on offer is confusing to say the least and risks the wasting of cash. So instead, despite some reservations I took the plunge and bought some of the new Hunt magnetic couplings. I bought ones with too-long stems, so we are back to silly vehicle spacings, though I have improvised a use for them on two new fish vans and more, shorter ones are on order.

Hunt magnetic coupling on two new fish vans. Any improvement?

It feels odd not to have mechanical couplings between vehicles, but it seems that the option to pull vehicles apart may be the answer to assembling cuts in my small fiddle yard, while retaining S&Ws at the ends. But as to whether they are acceptable visually, the jury is still out. I think they are no worse than tension-locks – but what looks like a hefty towbar between wagons isn’t exactly great either. In their favour, there are now versions to fit almost all known existing coupling mounts…

I know that Kadees have a good reputation and have used them between vehicles on the class 105 dmu – but they somehow don’t seem right for British stock of the era, quite apart from issues of cost, supply and the perplexing variations in which they come, which I have previously found.

One might have though that this issue should have been easily resolved by the collective might of the model railway industry – but it seems not. We repeatedly fall between what suits the train set manufacturers and those without the need to avoid scale links. I have even tried to conceive a bespoke solution of my own but as yet without success.

What to do…?

Posted in GNSR in 4mm scale | 7 Comments

In my mind’s eye…

If only I had more space.

I wrote some time ago about the need for the topography of the model to be consistent with how the real-fictional town of Port Deveron would look.

Way back last August, I presented the imaginary town plan, with significant input from the real towns of Banff and Portsoy. The proposition was that the GNSR provided Port Deveron/Banff with a rather larger station than the tiny affair that it actually had, and had extended the line a hundred yards or two out onto the harbour side.

In the interim, I have discovered that sidings did indeed run past the station down to the harbour more or less where the station supposedly is.

In the interests of explaining the layout of the model – and particularly the relationship between the two modelled stretches of water, this afternoon I had a play around in PowerPoint, to see what I would have modelled if I had taken over the whole room.

Or maybe not. In that scenario, I suspect rather more railway lines might have invaded. But it makes for an interesting prospect. On the left, we have the old town and old harbour, and on the right, the much enlarged new harbour, built when the railway arrived. The access road to the station and railway quay crosses a swing bridge at the exit from the old harbour. The new harbour gives onto the river estuary to the right of this view.

If I had modelled a prototype, all this would have been done for me of course – but it has served to visualise the wider scene quite usefully.

Posted in GNSR in 4mm scale | Leave a comment

In praise of small manufacturers – or how not to go potty…

My little forest of chimney pots. ModelU at front; Wizard in the middle and Model Railway Scenes to the rear. All in need of bedding in. The gulls are hand-carved (current population: 10).

I’ve only been to Pendon once – and it was many years ago. I remember an old guy – who I later concluded may well have been Roye England himself – pointing out minute features to me – such as the robin on the garden fork-handle, and some of the population – as he told me, all hand-carved.

I’ve always seen railway modelling as a personal practical challenge. Maybe it comes from being (just) old enough to have caught the tail end of the era when, Pendon-style,  making many things oneself was unavoidable. My father bought me my first edition of Railway Modeller in October 1974 and my early impressions were of those in their fifties or older who had come up self-sufficiently though the post-war gestation of the hobby. Equally, I know I’m not alone in finding the plant-and-play approach to modelling unsatisfying because of the uniformity it creates.

That said, I’m not really criticising, because we all end up making compromises at some point, and that is what this post is really about.

The current lockdown conditions nonetheless focussed my mind in this respect, partly because it was not easy to go out and buy materials and equipment on a whim (my ‘local’ model shop is a good 16 miles away), and partly because I was determined to use my shift from two to four millimetre scale to use the opportunity to make as much as I could myself, without going potty.

This image has an empty alt attribute; its file name is 4-02.jpg
I wish my hand-carved seagulls looked less like ducks. Daft, maybe, but getting it right is very difficult, but as 3D printing is showing, it makes a considerable difference… The tyres were laser-cut in three parts and assembled…

It is pots that actually tested the point, though; chimney pots in particular. Roofs are of course highly visible on any model, and chimney pots are often a critical feature. I made the first few by hand, mostly by shaping them from plastic tube using needle files. But it dawned on me that I was going to need over fifty, and as one of the last jobs to be done on a building, they just weren’t happening. What’s more, some chimney pots are of a shape and intricacy that makes them very hard to file up from scratch. I seem to remember Allan Downes having the same problem, so I’m in good company; I assume the Pendon brigade just soldier on regardless…

So in the end, I decided to buy the pots – and then encountered another problem: in order to keep the fineness of modelling where I wanted it, there was a significant cost attached. ModelU produce some very fine examples – and I have used a number – but the cost of bulk-purchasing was rather eye-watering. I have also used some cast white-metal square examples from Wizard Models, which are adequate, but rather less refined. In the end, I resorted to Googling “3D printed chimney pots” and came across a (new) company I hadn’t heard of before, called Model Railway Scenes. Amongst an interesting range of products, they produced some very nicely-crisp pots at a more affordable price, and a fair-sized order was duly submitted (sorry Alan…). As a result, I now have a veritable forest of chimney pots in place, mostly waiting to be bedded into small fillets of DAS and weathered. They do indeed complete the buildings very nicely. I think that the resultant variety has added a little something too.

Here’s a look at the Artitec crabber that is in an advanced stage of construction. Still some small parts to add; detail painting to do, and name/numbering.
I would have struggled to build this from scratch too; also Artitec.

This company is but one of several now producing accessories using 3D printers or laser-cutters, and I am pleased to see that such cottage-industries seem set to remain a fixture in the hobby as a result. It has (perhaps) never been easier to acquire the machinery and a website and set up a small concern.

The issue got me thinking about similar matters a little further – where we can draw the line, use pre-made parts, and still call something “scratch-built”. There is no doubt that industrial techniques can produce results that are beyond even the most skilled individual. I’m thinking particularly of the human form here, where scanned, 3D printed figures have brought about a minor revolution in realism in one of the last remaining untouched corners of the hobby, where even the finest models sometimes let themselves down. There is also the issue of quantity: is there really modelling value in producing 300 identical model fish boxes? But there is, of course, the cost in what one can reasonably claim to have ‘made’ oneself – and I suspect that the ready availability of these items (in some cases custom-ordered at that) is going to contribute to a further decline in the genuine self-builds with which one can sometimes surprise oneself, as I did with my GNSR gas lamps.

No commercial models available – and the scratch-building of two (so far) has given a couple of afternoons of pleasant activity…

Somewhere in the middle, we have those who are using these technologies themselves; clearly departing from ‘traditional’ model railway skills but nonetheless using IT skills to produce fine work. It saddens me that this will probably lead to the loss of manual skills, but it has its place – and I must gratefully acknowledge the work of Graham Speechley in producing the lobster pot frames and street lights to a standard I probably would otherwise not have managed.

Graham’s lobster pots in the foreground. 3D printed figures and bollards. And a skinny home-carved seagull (I have since switched to thicker plastic)
Graham’s custom-printed 1950’s style concrete lamp-posts. Unique to my model….

On the other hand, again, perhaps the decline of one skill may lead to the emergence of another, and I see that painting the printed figures is becoming a sub-discipline of its own.

Another technology with mixed implications is digital photography. Cameras are now so versatile that we can achieve instant close-ups of our work – handy for correcting errors, but I wonder whether it will ever be possible to create models that can withstand such close scrutiny without looking dreadful. Some of the accompanying photos mercilessly reveal my own failings in this respect – but I am going to claim, with reason, that the limits are those of my eyesight, and they mostly look fine from normal viewing distances.

This character remains one of my favourites. The photo cruelly exposes flaws not visible to the naked eye. Or at least, mine….
There is no doubt in my mind that scanned-printed figures are a major step forward…It still requires thought to position them naturally, and not to over-do the population levels, though… As for the money/time trade-off, I’m glad I don’t need a city’s worth…

In the end, each modeller no doubt makes their own decision; I am certainly not going to decry those who resort to more ready-made stuff than I do, though I would still argue that there is plenty that can be done to personalise ready-made models, and indeed to set them in a well-observed context rather than crudely plonk-and-play. And after all, all of my motive power and rolling stock is commercially-made; an area I’m simply less interested in developing, particularly now that the proprietary offerings are so good. Resistance even in Model Railway Journal gradually seems to be wilting in that respect.

Not sure whether this is a viable scratch-build either, though Geoff Kent might disagree….Stage 2 weathering awaited.

So this post is, above all, a call-out for those small suppliers who make so many bits and pieces that we might struggle to do ourselves. I like to reassure myself that my fickleness in who I buy from is actually spreading my custom widely….

In my case they include (in alphabetical order):


Hardy’s Hobbies

Lanarkshire Model Supplies

Model Railway Miniature Scenery

Model Railway Scenes

Model Signal Engineering/Wizard Models


Oxford Diecast

Railway Scenics

Road Transport Images

Scale Link

Scale Model Scenery

Scale Scenes

Severn Models

Shire Scenes

5522 Models

I just wish they would come up with more memorable and less confusing names!

I think the ability to support small producers no matter where they are, is one of the main benefits of the internet – it is also a pleasure, for instance, to buy CDs direct from independent musicians, which often arrive signed and with personal messages attached – and no middle-men.

But here’s also a shout for scratch-building – there’s still a place for it, and certainly for the modification and customisation of bought pieces to add individuality and personality to our models.

Posted in GNSR in 4mm scale | 2 Comments

Engineers’ possession.

Having pushed my luck with finishing that last building, I decided to have a short break from ‘serious’ modelling work, and just enjoy running some trains. I also had a purge on the room, now that most of the heavy modelling equipment can take a rest for a while.

Or so I thought. When I turned the power on a few days ago, some kind of surge took place in the servos. There is normally a corrective twitch as power is applied, but for some unknown reason, this time it was different. When I came to operate the points, I found that several were not working properly.

The worst was the king point, where the drive wire had clicked right out of its hole in the tie bar – and of course, being the king point, this effectively brought the entire formation to a halt. Something had to be done, but I really didn’t want to have to up-end the layout again to adjust it at the servo. Had this been a Peco point, the tie-bar extension could have been bent upward and the wire re-engaged; when I tried the same trick, I just broke the pcb…

So out came the files, soldering iron, flux and more. The blades are attached to the tie bars using pivots made of wire , bent underneath and on top in two 90˚ planes, with the blades soldered to them. I had little confidence that I would be able to replace this with like, while everything remained in situ. However, it proved possible; the pins were prepared away from the layout, and the new tie bar slid into place under the rails.

But the same lack of flexibility was clearly going to prevent the re-engagement of the drive pin. In the end, I decided to cut through the tie bar to the hole, so that the drive wire could be engaged from the side. This of course meant it would pop straight out as soon as pressure was applied – so I made a brass collar from part of a redundant S&W coupling, using the hole made for the chain. This was bent into an inverted U shape and dropped down over the wire and the tie bar and Araldited in place to the latter. The blades were soldered to the new pins and – touch wood – everything seems to be working again.

Luckily, most of the other turnouts only required re-setting on the drive limits on the microprocessor, though one servo driving a siding turnout appears to be dead and will eventually need to be replaced. I can live with it in the short term.

(I think there is some way to go to make servo functioning failsafe. While it is a minor problem, I am not entirely happy with the way the drive wires arc upward while in motion. Another time, I think I would position them to move under the board in the horizontal plane and push a drive rod fixed to the tie bar back and forth). Somebody must have already tried this…

Then I discovered that the same twitch had disengaged several of the signal drives. More fiddly work, which involved removing the drive wires, re-engaging them through the holes in the various cranks and threading them back through the baseboard to the servo drives beneath. Once again, it seems to have been done successfully, much to my surprise…

Anyway, after all that excitement, I decided I really did need a quiet life for a few days. Hence the only new work is the addition of warning signs to the accommodation crossing – printed out from a photo of the real thing, onto sticky label, given a very fine frame from the same source, stuck to 10 thou plasticard and a post, the latter, and the raw edges carefully touched up with paint, and planted. The only other additions are a streetlamp to the telegraph pole near the quay, and a small piece of fence to fill a gap near the accommodation crossing gates.

The population has swelled by another three individuals – and two cats – in past days, and I think we are approaching saturation in that respect.

I was in need of a gentle session with the camera after all that, so here is a bumper collection of images that resulted. Of all the models I have made over the years, I think this has turned out to be one of the most photogenic, and I am delighted with it in that respect – broken tie bars notwithstanding.

D6116 rolls into Port Deveron with a train from Elgin
As seen from the steps of the signal box
…and the platform.
D6116 poses in the sunshine
D5347 will depart first with a couple of parcels vans, which it will run up to Portsoy for attachment to the afternoon Elgin – Aberdeen coast line train.
What the signalman sees….
Distraction for drivers.
The NBL is left alone…
The line between ‘life’ and ‘cameo’ is fine…
Likewise. Especially when there is a session due to start at The Mariners…
D6116 still waits. As with the class 27, it has now been given its voice…

…and we’re away…
A little later, D8011 arrived in search of the afternoon freight.
Seen here collecting vans from the goods shed. They are probably full of potatoes, which are the area’s other main product. King Edward (the place) is just a few miles away. It may be noted that a little ‘progress’ is creeping in, in the form of a blue and grey mark 1…
Posted in GNSR in 4mm scale | Leave a comment

Topping Out

Today marks a significant day for Port Deveron. A little over a year since construction commenced, the final vacant plot was occupied.

There only remained a small space, about 8cm x 4cm for the last building, and it was a question of what should go into it; another cottage might have sufficed, but for a prominent position, I thought there should be something else.

In the end, I borrowed part of a terrace of shops from Buckie town centre, only modelling one end of the row. There was then the question of what kind of shop would occupy such a plot, right at one extreme of Port Deveron High Street. The obvious choice was a low-order convenience store – but in the end, I opted for my wife’s suggestion (again) and went for – what else? – a slightly faded men’s outfitter and Highland Apparel supplier.

So Malcolm Strachan’s kilt shop now graces the space, complete with formal-wear outfit in the window (carved and painted, once again, from plasticard). There is quite a lot of tidying up to do in the area, but nonetheless an important milepost has been passed, in that the layout looks complete.

It’s rather alarmng – but a sign of times – that a build planned for three or more years has reached this stage in just over one – but it has kept me well and truly occupied during lockdowns, and there is still plenty more to do, as various parts need further detailing, even reworking. There is quite a lot of weathering to do – and the small matter of the rolling stock fleet to attend to….

In the meantime, I’m quite please with the townscape.

Posted in GNSR in 4mm scale | 8 Comments