Thursday 22nd October

Worklist:
Fit new knobs
Run trains and test points

The knobs (guitar effects knobs) finally arrived from China (10 in a pack with a screwdriver - £4.41). They are only 19mm in diameter and are very neat.

Guitar effects knobs from China, via eBay

The layout was assembled ready for another test run, following Geoff's rewiring work, which comprised linking across track joints with poor sleeper adhesion by adding droppers and joining underneath. However, after running over virtually all of the track, Laurence found a section with a short circuit. We spent the rest of the morning trying to isolate the problem,  We tested for faulty micro-switches, stray hairs of copperclad sleeper, cross-feeds from other points etc., all with no success. After lunch, Geoff took the decision to apply a heavy-duty jolt, from a 12v battery, and the problem disappeared (there was no 'bang' or spark as we normally find, so we were not sure what the problem was).

We also tested the electro-magnet uncoupler, which had failed to work during the previous tests. The wiring was double-checked and the switch unwired and rewired. It was then remembered that the original control panel had a separate power supply for the solenoid points and uncoupler. The original supply was found and installed, solving the problem. Trains were shunted and uncoupled on all of the magnets.

Around this time, Geoff proposed that another section be introduced at the end of the platform, next to the refuelling area, so a loco could remain in there while the coaches were removed. It made sense, but there were no spare ways through the loom. Thoughts turned to why the coaling stage was sited there at all, as the coal and water were the wrong way around and it looked a bit 'clumsy'. There was also a bay on the end of the head shunt, which could be used for coaling. One for further discussion.

John's Peco Hunslet on a scratch chassis with a Gakken coach on a scratch chassis.

We enjoyed the investigative work, but we vowed to do a lot more running to get a better idea of how the layout worked. No hardship in that!

Thursday 15th October

No work this week as we'd decided to test the new control panel with a running session......

Geoff had assembled the layout and connected the new panel, but one problem remained - the missing relay hadn't been found. Rather than 'work-around' the problem, he popped down to the local Maplins for a replacement.

In the meantime, Laurence and Stephen cleaned the track and, as several bits of new ballast needed to be scraped off, the hoover was pressed into use as well.

One last job was to trim the actuator wires to the correct length, which Geoff completed with a cutting disk in the Dremel.

Cutting the actuator wires to length

The panel now complete, Laurence laid out one or two loco's and we played trains for a few seconds..... until the loco stopped. A short on the power supply indicated a problem. Disconnecting the loom showed it was on the boards. What seemed like ages looking for likely problems eventually focussed on a faulty micro-switch, which was replaced.

Back 'on track', we then discovered a rail-end had come loose from the PCB sleeper, which was repaired. Then another, and another. By lunchtime, our patience was wearing pretty thin, so we stopped for a break. Fortunately, just one more repair was required and we finally ran a loco across the whole layout. A few pieces of stock were added, to test the magnets, which seemed to work OK, except for the electro-magnet, which needs further investigation.

It was good to see the layout in one piece again, and we decided that a regular running session would be a good idea to make sure everything was still in working order and have a bit of fun.

Why so many track problems? They had certainly all appeared since the last running session. We think that the man-handling of the boards in-and-out of storage each week has probably played a part, but the age of the system must also be a factor.

Decauville on a Fleischman chassis

Thursday 8th October

Worklist:
Control panel

We were eager to get the panel finished this week and got straight down to work (after our first cup of tea).

After a brief logistical discussion on how to lay out the relay and power supply units, Geoff and Laurence made-up a strip of plywood to form a sub-assembly. Meanwhile, Stephen made up all of the connecting wires together with the 18v ac supply 'posts' taken from the old control panel.  A bridge rectifier was added to convert the supply to dc. The power units (which can take anything up to 30v) were trimmed to 1.5v and 5v. All of the units were assembled on the strip, the panel dropped into place. Geoff then made all of the final connections and we carried out an initial test.

Geoff in his element

The final assembly

Two of the relay boards

Power components

The 18v ac supply enters the box on the right and terminates on two bolts. One pair goes to the bridge rectifier that feeds to two point power supplies - one unit set to 1.5v for the point (servo) motors and the other (set to 5v) for the relays.

We connected the first layout board, to test the system, and it worked, much to Laurence's delight. It only has point, but it was very pleasing:

Laurence 'tickled pink' that it was working!

The other two boards followed, but it was found that a few of the points needed rewiring on the boards as, in their old configuration, points were wired and operated together, whereas the new system was one-switch-per-point. Furthermore, not all of the switches were correctly aligned, but a note was taken of all the anomalies and Geoff swapped the wiring later. One of the relays was also found to be faulty and will require replacement.

We were so pleased to complete the system overhaul, that we intend to have a running session next week, to give everything a good work-out, remind us what have achieved to date and where we are going.

'Our man' removing the hinge from the old panel.

Thursday 1st October

Worklist:
New control panel

Full house this week, albeit L had an early appointment elsewhere.

We had realised that the donated box had a sloping top, which wouldn't work in all of the positions we intended to use the panel i.e. at the front, for workshop operation, or rear, exhibition use. Stephen set about cutting the box edges level. Meanwhile, Geoff made up his preferred design of tag strip to join the loom and panel wiring. Brass nails in a strip of plywood.

First, the loom wires were fed in through the hole in the box base. But then a decision was required as to how to lay the wires out. It's not often you see the 'group leader' stretching his grey matter to the limit, but John's wiring system did the trick.

We also discussed the brilliant solution Geoff had conjured-up to control two points, sequentially, with one switch. More importantly, would we use it? It was considered that to have one set of points controlled differently to the others would require a different switch layout, which could cause confusion in the future. It was decided to adopt the standard format - one-switch-per-point, which required an extra wire in the loom, which Stephen fitted. There was one spare pin in the board connector, so not a problem.

We had decided to create a temporary panel-top, to test the new switches and knobs. Geoff had prepared and printed a layout, which was glued to a piece of hardboard. Geoff drilled the holes and Laurence then populated them with switches.

The old toggle switches were used for the sections, but temporary 'birds beak' knobs were used for the points.

Recap - Instead of the stud contacts used for solenoid control, we now needed to change polarity so DPDT switches were required. We had copious rotary switches that we could use, but the operating angle was quite small. Stephen had been researching alternative switchs and found one with a 45 deg. movement. However, it wasn't available in DPDT. When no alternative was found, we decided to use the ON/OFF switch in conjunction with a relay, to provide the polarity change. Geoff had suddenly realised that he had several of the switches and dug them out - perfect! What was even stranger was that the switches were identical to those in John's original control panel - spooky?

Knobs - we had the ubiquitous 'birds-beaks', but Stephen and Geoff wanted something different, something asymmetrical with a point/indicator at one end and no overhang at the rear. Stephen had researched knobs and circulated his preferences, together with designs for home-made units. Geoff had even fabricated his own idea in brass, which was tested today - the switch and knob/bar worked well. However, Stephen's continued research had produced a small knob made for guitars/effects boxes - not a bar, but a circular knob with white-line pointer that (somehow) had the same effect we sought. A mock-up of the size (19mm dia.) was played with and agreement given for Stephen to purchase some for testing. Watch this space...

---

Homework

Geoff continued with the wiring, completing both loom and panel elements. We think he's done this sort of thing before.

Stephen completed manufacture of the relay boards - one relay per point. There are three boards to allow for fitting in the box. A three-wire bus-bar loops through the boards - a 1.5v supply for the motors and the common return to the rotary switches. The feeds go straight to the relay.

This is the unit used to provide the 1.5v dc supply. We are using them for a number of applications on the layout as they are cheap (eBay / China / Hong Kong), and provide a regulated voltage.

LM2596S Step Down Adjustable Power Supply Module Input 3V-40V Output 1.5V-35V