Large radius half curves

I never understood why HD went to all the bother of producing the standard curve in half and quarter sizes yet never applied the same level of attention to the large radius.

Not normally a problem until you want to do something fancy whilst maintaining the appropriate gap between tracks.

As the new layout has a "fancy" corner, it was out with the Dremel again to carefully cut down a large curve into two halves. Some minor soldering required underneath before this can be fitted but it should do the job and save me having to use 1/8 rails to space it properly.

Looks about right when compared to the standard radius anyway...

New layout: first scenery

The first signs of scenery appearing!

The last time I had a tunnel on a train set it was made of a shoe box, these have been formed out of the polystyrene sheets that I had bought to insulate the shed at our previous house, never imagined that they'd end up being part of my new layout!

Managed to clear some space in the garage to accommodate the new boards, although smaller than the last layout when assembled,  the new boards take up more space due to the extra height to accommodate the scenery, signals, station buildings etc.

Having finally stripped the last of the trackwork and electrics from the old boards, I finally made the tough decision to scrap the lot. Of all the old boards, only one section (two four foot boards hinged together) were really salvageable. Many of the derailments and nearly all of the uncoupling issues can be attributed to the old boards warping. They weren't reinforced and had to be screwed together to prevent further warpage, then we were bending sections of track to fit properly - not the best idea. Once all of the track was removed, the boards really weren't in a salvageable condition, so they've gone to the tip. Sad to see them go as many hours had been spent, not only constructing them but also modifying them to suit the changes we made to the old layout and repairing them when stock started to derail and uncouple. They've served me well over the years and hopefully the new layout will last just as well, if not a good deal longer as the new boards are of a stronger (and heavier) construction.

New layout: All in a day's work - part two

Prior to today's hectic all-day session, the turntable has been cut out so that it will line up properly with the track (the Marklin turntable is much deeper than the standard Dublo one)

Hole cut, turntable test-fitted.

All six boards were then primed and painted prior to today's session, here they are assembled again for the first time since they were made and checked to ensure minimal warpage and of course that the dowel pins and catches still line up. 

It's a snug fit but everything works.

Once a good fit was confirmed it was time to separate the front & rear sections ready to fit the electrical connectors.

After fitting the front section first, the rear boards were flipped and the breakout PCBs that I had shipped over from America were cut into position. A nervous moment as a lot of money has been invested in these components with the intention of minimising the amount of manually connected terminal blocks, which the previous layout had many of. These weren't really intended for this kind of use, normally you would plug a 37 pin cable into the socket, not a PCB into a PCB. The theory here is that once the dowel pins have been lined up, these will connect together when the boards are then clipped together.

And it appears to work! - Each section was tested, separated and re-joined and each board connected properly with no fuss.

As time began to defeat us, there sadly wasn't much time to wire up or fit track. We did however manage to fit a perspex protection panel along the front edge of all three front boards (it's hard to see it in this picture but it's there along the top) after loose-laying all the track along the front three boards, I managed to mark and drill the holes for the wires to the two pairs of points and fix down most of the track on the first board.

Just a few more pieces of straight and the corners to fit, then this board can be wired. End panels have also been fabricated and bolt to the exposed board edges so that the track is fully protected for transportation, also means that the boards can easily be stacked.

Good progress again after another long day, still need to lay track, wire up and then start thinking about scenery, then there's the upper loop to be added but this can't be done until the hills & scenery are done. All agreed that if work slips behind and the deadline is at risk then the top loop will be put on hold.

Fingers crossed we can still meet the deadline in October.

New layout: Wiring

I seem to be forever soldering.

Every spare evening is spent in the back room armed with a soldering iron. Thankfully the control panel is now completely finished, apart from two coats of varnish to protect the new paint.

Time to concentrate on the wiring for the track. As there are six individual boards I will need to ensure that all the tracks are electrically connected via the 37 pin connectors.

I also need to wire the points motors and modifiy the existing Dublo wiring under the points to allow for track isolation (Dublo never made isolating electric points sadly)

For the wiring I've opted for six core alarm cable - this stuff is stranded wire rather than the solid cores found in telephone cable. I used phone cable throughout the old layout and found that it didn't like being used in terminal blocks, didn't solder that well and became quite fragile. I'm also conscious of voltage drop through the extensive amount of wiring being used on this layout, alarm cable doesn't tend to suffer too badly in this area, let's face it, it's designed for 12v use and on a standard house alarm system you'll find miles of this stuff in between all the sensors - if it works for that then it'll work for this... that's the logic anyway.

Here I've cut lengths of cable ready to solder to the points...

Once all the points were wired, I labelled them (as some have a different configuration and some don't require isolation) and then soldered individual wires to some straight track sections to allow for all the board joints and the eight isolated sections in the main station...

Another tick off the large to-do list, just the back boards to do now!

Turntable test

First test of the newly "modified" Dublo / Marklin turntable using the new switch panel and control PCB...

New layout: Control panel completed

Possibly the most labour-intensive project I have ever undertaken!

The MDF casing needed priming before paint...

I found a reasonable colour match for the blue, I wasn't going to go all-out and get the colours professionally matched, no point and this, for a cheap shop own-brand paint, does the job quite nicely. While the paint was still drying I loosely fitted the main panel for the first time... 

Next day, and after the paint had hardened, I began wiring the Duette controllers into the panel (should call it a Quartette now I suppose)

40 LEDs had to be soldered and glued into position - hours of work involved in this and when I finished, the light was way too powerful as this picture demonstrates...

After a bit of head-scratching, the problem was resolved by adding two more resistors in the 0v return.

The 37 pin breakout PCB was wired, plug-in terminal blocks have been used so that the panel can be removed easily for maintenance. 

Everything in position... finally!

There were problems - many of them.

First of all, two of the brand new switches were faulty, thanks to my wiring diagram and schematic diagram I was able to trace and rectify the problem with relative ease. Just annoying that new switches should fail like this.

Then I discovered problems with the Duette controllers. First of all, two of them, when energised, operated at full power on minimum and minimal power on maximum. After more head-scratching the problem was traced and resolved, two wires on the potentiometer reversed!

After resolving that issue, another Duette problem. If controller B was used in forward and controller C in reverse (the two main loops) a short circuit was detected by the ammeters. This one took a fair bit of tracing, taking a whole evening alone to resolve.

Finally the fault was traced to the centre of the potentiometer for controller B. Basically the centre shaft (the bit that the control knob attaches to) is live at +12v dc. On the Duette, this is insulated from the body of the controller by using a cardboard spacer and a large, square cut-out which ensures that the shaft can never come into contact with the controller body. Although I'd re-used the cardboard spacer, I'd drilled a round hole only slightly larger than the diameter of the shaft itself, and this rubbed the centre shaft. After a quick re-drill of all four controllers I refitted & tested. Job done!

Having all of the major components on plug-in terminals made the process of fault-finding much simpler so already this decision has paid off.

A much more sensible brightness level. Just need to secure the main panel and, once varnished, the job is done.

I'm now in the process of testing the PCBs I made for the boards. So far I've tested the eight relay track isolator and the first, two capacitor PCB for board one, just need to test the turntable control - already confirmed that there's power from the panel to the PCB but as it's getting late I'll test tomorrow. The turntable is a tad loud to be testing at this late hour.

Although it's taken many hours just to build a switch panel, I think the end result is stunning - far better than I could ever have hoped. I particularly love the new panel for the four Duette controllers. The printed & lacquered vinyl looks amazing, albeit still with some small air bubbles. It was annoying that I couldn't use the same process successfully on the main panel but this, for now, will do.

Next weekend we'll hopefully commence track laying and wiring.

New layout: Control panel - inside a Duette

I was always intrigued as to what was inside the A3 and the H&M Duette controllers. At least one of these mysteries is now solved.

After drilling out the rivets, the lid lifts off. I've already removed the left hand controller by this stage.

The rivets holding the rectifier and cut-out for the left hand controller have also been drilled out ready for removal.

A 16V AC twin coil transformer sits in the middle, supplying independant power to each controller.

The 16V AC auxillary output is simply connected straight to the standard output of the transformer coil, after the cut-out device, with the other 12V DC supply tapped from the rectifier.

The auxillary outputs are not required so wiring is slightly simpler. 

The components are stripped, the potentiometers are mounted to the new panel and new, colour coded wiring used to make it easier for me to identify which wires go where. New toggle switches replace the old H&M ones for resistance and half wave.

All that's left is a shell, two outputs and a bank of switches.

After stripping the second Duette, all four controllers are mounted to the new panel.

And the cut-outs & rectifiers mounted to a spare piece of aluminium left over from the control panel.

This is now ready to drop into the new control panel, I'll use pluggable terminal blocks here so that if a fault occurs, I can remove either the control panel or the transformer and rectifier assembly easily without the need to de-solder anything first.

New Layout: Electrics

My evenings seem to be taken up with soldering, lots of soldering.

First PCB to be constructed is for the first board, this comprises capacitor discharge for two points, track isolation for the two points (as Dublo points aren't electrically isolated) and turntable control - the relay & diodes of which were reclaimed from the small control PCB that I built ages ago for the turntable.

After many more nights of soldering and making lots of mistakes (ie having the layout map upside down) I've finished all the PCBs for the layout.

This little lot needs to be fitted throughout the front three boards...

The shuttle control is still in its bag and the signal matrix control was purchased a few years ago. The rest is self-built.

The eight relay bank is for the track isolation controls within the four tracks of the station. Elaborate but allows easy connection and disconnection of locos & multiple units etc.

I have to confess that the smaller, two capacitor PCB on the left was actually the first PCB that I produced for board one but sadly I didn't buy bread board large enough to accommodate the second bank of terminal connectors and the large capacitors. For board three there are only two pairs of points and track isolation is not needed so I de-soldered the second bank of terminal blocks and soldered in the capacitors. In reality I probably could have produced a completely new PCB from scratch quicker than it's taken me to re-hash this one but I don't like seeing things go to waste!

I've tested everything as far as possible using my multi-meter to ensure that there are no dry joints, so far so good. Now to test the capacitors using a pair of points.

The idea of using home-made PCBs is that this should make wiring the boards much simpler. On the old layout, the switches were mounted in the boards with individual wires and capacitors all over the place. One stray wire could take ages to locate. This makes the whole process a lot simpler and should (hopefully) make fault detection much easier, plus I have circuit diagrams to refer back to - previously I've done everything by memory, the result being that each board ended up being wired differently with different colour combinations used for the points wiring. Everything across this layout will be standardised.

Much easier.

New layout: Control panel progress

The control panel is making slow, steady progress.

The lower panel is still waiting for the rest of the components to be swapped from the two Duette controllers. Some drilling of rivets and rewiring required first.

I mocked up the display, printing the design onto ordinary A4 paper, the idea being that I could then mark out the positions of all the points switches and LEDs, rather than trying to drill through a freshly lacquered panel and risk ruining the final finish.

The left panel has been redesigned as I wasn't happy with the look.

Mocked-up and refitted to see how it looks...

Time to mark out the holes, I centre-punched these in an attempt to stop the drill from moving when I cut the holes. Once all the holes were punched, I removed the template...

After lots of drilling (pilot holes, followed by 3mm LED holes and 6mm switch holes) I ended up with what looks more like a pin cushion. The ammeters were also cut in using a series of drilled holes and a reamer.

It's fair to say that using 2mm aluminium sheeting was way over the top for this, 1mm would have done the job. Drilling aluminium creates lots of swarf, which needs to be removed from both sides or it'll ruin the final look. I found that the easiest method to remove the excess metal was to carefully re-drill each hole with a larger diameter drill bit. After all the drilling, filing and reaming, I ended up with this...

The original plan was to finish this panel in exactly the same way as the transformer panel, however there are problems with the method I used - basically printing onto glossy vinyl and lacquering in place. The vinyl isn't too happy with the inkjet printer, the ink doesn't fully dry and smudges with ease - even if left for a few days.

I tried several times to print and stick the vinyl overlays into position but found it impossible to do this without creating air bubbles underneath, I can't apply any pressure as this smudges the ink! - I've experimented with applying a thin layer of lacquer prior to application, this worked reasonably well but I've now run out of vinyl. Not to be beaten, I've printed all of the control panel again onto photo paper and have experimented with using PVA glue to stick the panels to the aluminium, once dried I'll apply several coats of lacquer which should eventually give me the same look as the bottom panel. This has also proved troublesome and for now I'll have to wait for the glue to dry to see if this was a good idea or not.