There are five bars per platform: one of a rather large diameter in the centre (I chose 0.5 mm to scale), and four near the access doors, thinner, which I estimated 0.3 mm (0.4 may also be suitable). There was no question of giving them the rather convoluted shape of the actual bars. I was content with rectilinear bars.
Photo J.-P. Demoy
For the material, something was needed that resembled chrome or stainless steel. The best, for the appearance, would have been of the piano wire; It can be found in detail at L’Octant, but it is hard enough to cut, bend, and solder, and its appearance is not so gloss. I preferred the nickel silver, bought at micro-modele, in spite of its somewhat yellowish colour. After all, it is also used to reproduce steel rails…
Here are the dimensions of the different bars, calculated for the assembly we will see below (quantities per car):
OK, I spill the beans: these bars will be used to drive the lighting power from the conductive couplings! This is why I recommend tinning them on the 7 mm long end before they are assembled.
The difficulty will be to obtain the best possible verticality and parallelism, because any defect will be very visible through the windows. We will start by gluing the central bar which will serve as a reference for the others. Here again, a template made of a plate (wood or thick plastic) drilled perfectly vertically will be very useful.
So, from below the fittings, I slip the ø 0.5 mm bars into their holes, with the horizontal part of 8 mm well-placed in the groove; I slip the template over the whole and I glue underneath with cyanoacrylate.
After drying comes the mounting of the ø 0.3 mm bars; this is the most delicate point. First, place the transparent support plate, which is inserted into the ø 0.5 bar, and rests on the toilet wall (or its equivalent at the other end). As the balance is unstable, I place a 20 mm high shim on the stairs side (I use extruded polystyrene). I hold it all with a not too strong spring clamp or a good elastic. I could then glue this part on the partition, but it would be very difficult to ensure the parallelism of the bars. That is why I start with a blank mounting of the ø 0.3 bars, as follows.
I stick small Patafix pellets on the transparent plate, above the ø 0.3 holes. Then, I slip the bars from below, successively into the fittings, then into the transparent plate. They will then perforate the Patafix and be thus retained without being able to fall back. I leave a minimum of overhang at the bottom, the maximum available space being 1.2 mm.
Note: in this picture, the gluing of the transparent plate on the fittings is already done, and the shim is removed.
With all the bars in place, it is easy to see if they are all parallel or not, and correct by modifying the position of the transparent plate. When it is satisfactory, I glue the transparent plate on the wall with liquid model glue.
Once the glue is dry, I can glue the ø 0.3 mm bars underneath the fittings, with cyanoacrylate, then remove the Patafix pellets, and then glue all the bars over the transparent plate. But I do this only on the cab side of the driver car, because this allows to keep a certain clearance and sometimes to avoid bending the bars.
This paragraph would be more appropriate in the next page devoted to wiring. However, this must be done before reassembling the fitting in the chassis. That’s why it is discussed here.
This wiring consists of placing two wires along the chassis: the first is the positive polarity of the decoder (standard blue colour), the second is that of the decoder C output (standard green colour). Here, the two wires are red because I do not have too much choice of colour.
The wires will run into the original grooves. To do this, if it’s not already done, the ballast must be removed.
I pass two 240 mm long mini-wrapping wires stripped at each end in the grooves, between the bean-shaped holes (distribute the lengths by marking the middle of each wire with respect to the centre of the chassis). I hold them by a few pieces of adhesive tape, and I make them out by the beans. The advantage of wire-wrap wires is that they are both thin and rigid, which makes it easier to hold them in place, but flexible wires can be used, provided they are fairly thin.
Later, other wires going to the conductive couplings and to the interior of the car will be connected.
I put the ballast back in place (glued with Pattex 100%). Some ballasts being well rusted, I applied a rustproofing before reassembly.
It is better to replace the bottom fitting in the chassis before gluing the top fitting, to avoid stresses on the latter that could take it off. Do not forget to reassemble the elongation drawbars first!
Reassembling poses no great difficulty. Be careful, however, that the parts are keyed, and that there is therefore a direction of mounting to be respected. Once the fitting is snapped in, I check that the clips on the ends both hitch. If only one does, it is possible to slightly slide the fitting to better centre it. If it is not hitched, it will rise and this will be very visible by the windows of the platforms, because the holding bars will be leaned.
Then we have to glue the upper level fitting to that of the lower level. I verify that the top floor is not bent down (“banana shaped”), because there will be nothing to support it. If so, I fix it with elastics on a rigid board by interposing a wedge that deforms it — moderately — to the opposite direction, and I let it marinate for at least one night. One can try to speed things up with a hair dryer, but be careful: the plastic may release internal tensions and deform even more (lived)!
I then glue it at the foot of the stairs. The best is still to use cyano. To ensure alignment, I glue one side at a time, with two clamps: one that ensures alignment (A) overlapping the two elements, and the other for tightening (B). Another possible glue, also based on cyanoacrylate, which seems interesting because it allows an adjustment of the parts for about twenty seconds, would be the Uhu Strong & Safe.