Junction Dock 2018

Well 2018 was a very quiet year for Junction Dock – I did precisely nothing! No, sorry, I take that back – I fitted one piece of ~A4 card to complete the curved back scene in, I think, April… Club commitments, and 12-inch-foot commitments, mean I’ve not done much ‘home modelling’ this year. I have had fun with a few wagon kits though, so I thought I’d round out the year by posing them in front of the half-finished ‘goods shed’ on Junction dock.

img_20190104_214642690_hdr (1)

LNER Fruit, GWR V16, LNER 5-plank, LMS 3-plank – in various states of not-quite-finishedness

Pushing the wagons about was useful – the trackwork is simply not good enough. I have serious gauge narrowing at the tips of the point blades (all hand filed, probably not finely enough, and with not enough set), there are steps between sections which are very noticeable (not sure why — to few sleepers perhaps?), and the turnout mechanisms aren’t positive enough. This is the real reason I’ve not done anything on the layout for a while – the fear that the fundamentals are not quite good enough… That and the terror at trying to cover the entire lot in scribed DAS setts!

My initial idea – misguided I now realised – was that I could mask my first attempt at trackwork in setts. The ugly soldering, lack of chairs, mis-shapen sleepers etc would all be hidden. Of course the opposite is true – if you’re going to encase your trackwork in clay, you’d better be damned sure it is up to snuff first! So I think Junction Dock needs some serious rework on the fundamentals. A shift of EM is appealing, given I feel I’m unlikely ever to have/want a big empire populated by a stable of RTR stock. Perhaps after a year dormant the project should be marked ‘DNR’ and I should consider something more classical?  I have a few ideas of things I’d like to do, but don’t want to just flit around.

The good news from 10 minutes of pushing wagons is that I do still like the concept of the layout, and can see what I want. So perhaps it is not DNR but WIFLI – When I Feel Like It! After all, the rest of life is stressful enough there really is no point getting worried about a lack of progress with little trains :-\


Turnout Operating Units

I finally managed to make some turnout operating units (TOUs) for Junction Dock, and all the switches are now switching. The whole process was really a massive pain, but I’ve ended up with an implementation I’m happy with. My original idea had been to use wire-in-tube to operate the points, but moving the layout operation to the front meant I didn’t have a good location for the lever frame. So I ummed-and-ahhed, and decided to go with rod operation under baseboard and out the front panel. As I’d never really planned to install much under the board, not much (i.e. no) thought had gone into the placement of the structural members; which in turn means I had a rather constricted space to build the TOUs in. In the end, I came up with a design which is ~50x50x10mm, and provides a nice ~8:1 gearing between the push/pull rod and the switch.

After much searching, I found some rather attractive concepts, made mainly out of plasticard, on the scalefour forum, here; https://www.scalefou…0&t=425&p=13048 . My first incarnation (no photos I’m afraid) was very similar to the linked version, but I used an external crank to provide the gear reduction. This worked well, but didn’t fit! So I had to change the concept to an internal lever.

JD_tou_mark3madeThis is held by an 8BA screw on one side (left), pivots on a 1/16th brass tube in the driven arm (which in turn holds the connection to the tie-bar itself), and then held by another 8BA in the driving arm (right). The ratio between the hole distances provides the gearing. A microswitch provides polarity switching. The image below probably explains it better than the text. This is an early incarnation and further improvements were made, but this is the best photo I have. The whole shebang is mounted on a piece of 20thou plasticard (40-60thou would be better, but that’s what I had), which in turn is mounted on a sheet of ply (to make sure the mounting screws don’t come through the top of the baseboard!). The two sliding arms are built from evergreen plastic, 1/4 rod telescoping inside a 5/16th tube, from memory. I drilled a 1/16th hole through the middle of one rod to take a piece of brass tube of, I think, 1/16th OD (cemented into the plastic for security). This brass tube holds a telescoping piece of tube (3/64th?), which connects up to the point tie-bar (as my track is inset, I use full moving PCB tie bars).

All the screws are 8BA, which I found self tapped well enough into a 1.5mm (or 2mm?) hole. There is a screw to adjust when the microswitch trips in the arm motion, to control when the polarity is switched. The lever arm above just had holes for the two arms, which I later changed to slots. The ratio between the hole distances provides the gearing. I went for about 8:1, which seemed OK. That means you can move the point actuator by ~10mm to get the 1.5mm throw on the point itself. It also means there is enough resistance in the system to avoid the point back driving against the actuator. All of this provides a nice positive action to the point actuators


Mark 3 shows the slotted tie bar, and a change to the drive connector to bring the drive rod up to the right height for the baseboard. This reused the drive bar from mark 1, hence the redundant brass tubes (in theory set to the gauge of the point blades, and designed to allow the blades to be driven directly without the use of a functional tie-bar “up top”). Connection to the drive rod is via a stripped down chocolate block connector. Mark 4 made some further rationalization of the design.

Below is an image of two of the units installed under the baseboard. JD_tous_installed1The drive rods come through to the front of the layout. I decided to put the motion limits on the drive rods themselves, and that is simply done by adjusting a pair of chocolate block connectors against front and rear wooden stops. The TOU in the lower left of the picture is at an angle to the board, and in a really awkward position; so I used a bit of wire-in-tube to transfer the rod motion to the TOU. This is also a bit of a “special” unit, to account for the baseboard (in)design — it has the connection to the point outside the bearings for the arm.

Here is the front side with the three drive rods coming through. More chocolate blocks as temporary handles until I find something more attractive…


I finished these a few weeks ago, but haven’t had time to post anything. The layout is now fully wired (v simple – single engine in steam policy…), tested, and sprayed with primer. Next stage is I think getting some building frames in position so that I can start with DAStarding setts.

More track, and rust…

I’ve now completed the trackwork for the scenic side of the layout, and done most of the “backside”. Attached here are some pictures of what I did, and some notes on what I wouldn’t do again!

JD_stickingdowntrackThe pointwork for the layout – which accounts for most of the front area – was built off-board, and shown in a previous entry. Here it is lain down on generous bed of PVA (I didn’t bother with any copydex/sound insulation, as the whole lot will be encased in DAS cobbles anyway).


Maximum possible weight was applied. I thought this was a good thing. It was not… By putting so much on top of the track (particularly with a broad section of contiboard to ensure a level surface) I managed to create a lovely moist microclimate; which is not good for the steel rail! Much rusting ensued 😦 This was far worse than any rusting I got from the flux. I’ve cleaned off the rust best I can with fine abrasive paper and wire wool. Once wiring is complete, the whole lot will get a spray of primer which should seal things.


Next time round, I was a bit more ‘precise’ with the PVA to fix the sleepers for the rest of the trackwork. That seemed to go OK. The straight track was all built in situ. After soldering, I flushed all the joints in warm water and scrubbed with a toothbrush before drying it all off with a hot air gun. No evidence of any corrosion on these (left) — apart from on the section I forgot to clean (right)!! :-\

JD_cleanedjoint JD_rustyjoint

The sector plates are angle aluminium. I don’t have a block gauge to set these, so I modified a 16.5mm roller gauge by filing flats on the outer edges at two sectors. Should leave the gauge usuable for normal code-75 track, and be usable at a specific rotation for larger code track (or angle ally!)


A little bit more to do on the trackwork at the back and the daughter board sidings — but I’ve run out of rail… On to wiring now (droppers installed along with the track), and them TOUs I keep putting off..

Final nice view along the dock front… The half-built coal wagon has been my testing truck, and happily runs through all the trackwork so far…




I’ve been making some good, but slow, progress on trackwork for Junction Dock. One of the main aims of this layout was to have a go at building trackwork; with the safe assurance that I’ll be able to hide most of the resultant horrors under a thick layer of DAS. I’ve found trackwork threads/posts on various sites incredibly helpful, so I hope me listing my first fumbling steps below might help someone else at some point. This is all PCB construction. The techniques here are pretty much pure Rice (steel rail, powerflow flux + syringe).

First step was to get some goodies, which I picked up from Railex back in May last year.

Attached Image

10m of bullhead, PCB, solder, and a set of gauges (OO-SF). What more could you ask for!? I had a crack at building one point, the classic B6L, which didn’t turn out too badly. The gapping is perhaps ‘enthusiastic’, but it actually filled out quite well. I’ve never checked this under power, but wagons seem to run through it OK. As you can probably make out, the set on the turnout road is not enough, and the flange at the crossing is perhaps a tad generous; but fairly happy as a first attempt.

Attached Image

Attached ImageEnthused, I moved on to junction dock. First thing was to make the crossing vees. These are not brilliant, but appear to be functional. I made a jig for a 1:6 vee when I made the first point, but then forgot about it and ended up just filing up the vees to a set of measured lines on a board (I won’t admit here that I had the board up-side-down, and didn’t realise the jig was still stuck to the other side…). After filing the tips to fit (just about), I held the rails down with blutac and soldered up with 180C solder. End result, 3x 1:6 vees, and a 1:4.5 curviform for the steeply curved warehouse siding. These won’t win any prizes…

Attached ImageThen moved on to laying up the turnouts. PCB sleepers chopped to size (just over size for the track gauge; stingy scot…) and then stuck to a template with thin strips of double-sided taped. The sleepers were gapped and electrically tested before I started adding any rail. As the track will be inset, I didn’t provide all the sleepers (or make them the right length). I tried to make sure there was one sleeper every 3-4 at least, but in reality the pattern is pretty much set by needing to have a sleeper at the ‘end’ of each distinct bit of rail.

Attached Image

(sorry for the v poor photo quality!)

Learning from some corrosion issues on the first couple of turn-outs (soldered up over the space of a week or so), I made sure I cut and prepared all the rail sections so that I could do the soldering in one run and clean-up immediately after. This means lots of blutak to check lengths and fits etc. Running is not as smooth as hoped at this stage…

Attached Image

Attached Image

Once everything was ready, the switch was soldered up in the order 1) stock rail 2) crossing vee 3) other stock rail 4) wing rails 5) check rails 6) closures + switches (one unit). Each piece of rail was mounted in place (with a combination of blutac+template and/or gauges where appropriate), and then soldered up a few sleepers at a time. I used powerflow flux in a syringe (about 0.8mm tip? green needle anyway) to put a tiny (~0.5mm) spot of flux on the outside of the rail. Then picked up a spot of solder (140C) on the iron (Antex 18W) and ‘fizzed’ it onto the joint. 1–2 seconds for the flux to fizz and the solder to flow; quick up-and-away with the iron; and I seemed to get a 95% reliable joint and a chair-ish shape (not that it matters for this track of course). I appreciate this may not be best soldering technique, but it seemed to work for me.

 Attached ImageFinal stage was to solder up the blade tips to a moving sleeper. This was the trickiest bit for me (perhaps ‘cos I ended up doing it late at night each time!). With hindsight, I might even take these moving sleepers off and use dropper wires to a sub-baseboard mechanism. Little gain with inset track, but I need to have the mechanism below the boards anyway…

Attached ImageOnce each whole piece was built, I took it off the paper and gave it a good rinse, scrub (with toothbrush), and dry to get the flux off. This seems like a decent approach, and I haven’t noticed much/any corrison on the later turnouts. I did try to minimise the amount of flux used, with a tiny spot on each joint. I think I used <1ml total for the whole shebang. I did also notice that the flux definitely ‘goes off’ if left in the syringe for any length of time (e.g. week+). Was a frustrating 20 minutes of not managing to solder anything until I worked that one out!

Over the space of a month or two, I put together all the P&C I need for Junction Dock. Below shows, my 1st — 4th attempts at track making. I think I got better as I went on, which is always nice.

The last of the common crossings seemed to come out nicely, and seems to be the smoothest running (very little/no drop apparent)

Attached Image

So now I just have to do the plain track connecting all the bits up, and then get on with the wiring. I am slightly worried though that I’m going to run out of rail just before I finish! Doesn’t look like I’ll be crossing C&L’s path at any shows in the near future, so I might have to stump up the high (if understandable) postage cost to get some via courier :( Pleased however that I managed to score 1000 sleepers and 36ft of timbers (old SMP paxoline type) for a grand total of £5.50 at a show last week. Should keep me in PCB track for a while :)

Next job though, before any more track, is to finish off the wood work and get the backscence and proscen(ium) made & fitted.