Tinkering with a Piko Hercules

[Ian Morgan]

I thought I would do a quick little job as a distraction from working on my 2mm scale layout 'Freshwater'. Nearly 3 months later and I have just completed it. A few years ago, I picked up a Piko HO train set and a couple of extra carriages, second hand at an exhibition in the Netherlands. It was a model of the Allgau Express (Alex) that I used to see when I was living in Munich. I bought it with the intention of running it on 'Fallerstadt', a layout being planned at the time by my mate Dave.

The train set was cheap because the loco was running very badly. When I got it home and opened it up, I was surprised to find a DCC decoder installed. DecoderPro failed to communicate with it, so I popped in a Zimo MX631C and away it went, sweet as anything. It was only equiped with headlights, illuminated by a bulb at each end, and some plastic light-pipe to the three headlights. They switched with direction. It was put into its box awaiting a layout to run on.

Following my success with refurbishing a second-hand, sound fitted N gauge triebswagen, I thought fitting red tail lights to Alex would be a nice, easy job. How hard could it be?

I opened it up, and there was just so much room inside to play with compared with the models I usually tinker with.

First job was to replace the bulbs with surface mount warm white LEDs, with a surface mount resistor attached. That was indeed quick and easy.

There are plenty of function outputs on the MX631C, so I thought a couple more surface mount warm white LEDs should fit to the cab ceilings to provide interior lighting. Having done that, I had to add some paint and printed speedos and displays to the cab moulding which could now be seen more clearly. Each cab light is controlled by a seperate function button (1 and 2). Again, this job was completed quite quickly.

 

 

The decoder came with a 470uF 25V electrolytic capacitor for 'stay-alive' use, however the motor seems to need a lot of power when running, so more capacity was needed. Using the smallest 470uF 25V electrolytics I could find, I managed to glue in 8 of them by the bogies without inhibiting their swivel or tilt. This produces less than half a second of running after cutting the power feed, which should be enough to give reliable running. When not moving, the lights remain on for several seconds after cutting power.

Now it was time for the red tail lights. I wired up pairs of very small surface mount red LEDs with very fine enamelled copper wire (salvaged from an old Nokia phone charger). The LEDs were glued behind the red lenses of the headlight clusters and wired up with surface mount resistors. Power on, and the whole of the light clusters and headlights lit up red, along with the inside of the loco. The least amount of red light was actually from the red lenses, which has a thick blob of red paint on the inside.

 

 

So, plan B. I had some fine metal tube which fitted nicely in behind the red lenses, after removing the LEDs, glue and a lot of the red paint. I glued short lengths of the tube (about 4mm long) to the front of some more surface mount red LEDs.

 

 

In an attempt to stop light leakage, I tried to apply a layer of Milliput around the LED. This proved to be impossible. Also, too much manipulation broke the fine copper wires, again and again and again.

 

 

I tried again, this time coating the LED and wire with UV cure glue, to insulate and strengthen them, and tried wrapping them in aluminium foil to stop light leakage. I found wrapping to be impossible. Also, I broke the fine copper wires, again and again and again.

Instead of the aluminium foil, I applied several coats of black paint to the UV cure glue blobs (again breaking wires every now and then) and eventually got them light tight. Fitted and working, they looked good, but I had applied too much UV cure glue and the blobs were now to large to fit the light-pipe and shrouds for the headlights. Start again, new LEDs and metal tubes, sparing use of UV cure glue, black paint, more broken wires, but eventually two working tail lights at each end.

 

 

Having got this far, it seemed only sensible to add one more feature. The ability to inhibit head and tail lights at the end of the locomotive that is attached to a train. Functions 3 and 4 seemed logical for this, but on the MX631C, these functions are logic level only, and not normal current sinking outputs. Also, small holes need to be cut into the plastic sleeve around the decoder to access the solder pads. A couple of small PNP transistors and some resistors allow these logic levels to switch current to the LEDs. The assembly was wrapped in Kapton tape. There are now 7 wires connecting to the LEDs attached to the body. I glued a set of 7 header pins in the roof of the body, and attached 7 wires from the decoder and transistor bundle to a header socket, so the body can be detached with relative ease.

 

 

The loco was reassembled and after playing with DecoderPro, all seems to be working. I found that functions 6 and 7 can be set to dim the headlights at each end, which was a surprise added bonus. The only slight problem is that functions 3 and 4 default to off at power up, inhibiting the lighting. I have to press function 0, then 3, then 4 to get the directional lights illuminated. I had hoped that the function outputs could be inverted with a CV change, but it appears not to be the case.

So, the loco can go back in its box until I have a weathering session, or 'Fallerstadt' is able to run it, whichever happens first.

 

Edited June 23, 2023 by Ian Morgan
re-loaded images

[Ian Morgan]

Oh, and this is what it should look like: