The problem of height

The dimensioned diagram indicates that the actual coach height is 3940 mm, i.e. 45.3 mm to scale. It can be seen that the REE model is missing almost 1 mm in height at the roof: the model’s height is only 44.4 mm. At the level of couplings and buffers, the chassis is too low by about 0.5 mm from NEM 303 — it is actually at the lower limit of the standard — which is probably more troublesome for operation than the lack of overall height. And, in fact, the coupling tends to unhook, as it leans downwards (the NEM casing has a reduced bottom wall, which does not help to hold the head).

Curiously, on some forum, many agree that the DEV AO coaches of the HJ competing brand are too high. These have certainly many defects, including their price, but their height is almost correct (45.5 compared to 45.3 to scale).

You’ll say, no problem, just put 0.5 mm thick washers on the pivots… Nay, it would be too simple! REE has provided spherical ball-and-socket joints for the pivots as well as for its bogie tanks. This choice is certainly sophisticated, but embarrassing: did you get washers shaped like a spherical bowl?

Of course, there are bosses on one side to have a 2-point suspension. They have however the fault of raising (very slightly) the chassis on their side.

One — half good — solution

On the side of the bosses, it is not very complicated: it is enough to glue 0.5 mm thick spacers on the bogies, at the location of the bosses.

On the opposite side, we must find a solution which retains as much as possible the 1-point suspension, that is to say that allows a rotation of the body in the longitudinal axis (roll).

First, we must find out the spacing between the body chassis and the bogie chassis. By successive tests, I found a value of 3.5 mm.

I made rectangular spacers in 1 mm thick plastic card, maximum width 6.8 mm (this is the space between the two inner bogie beams), height 3.5 mm. The upper side is rounded to allow rolling. On the other hand, this provision impedes the freedom of the bogie in pitching; the most satisfactory solution would be to cut the pivot pin on the chassis side and thicken it. Tough job! In fact, not that much: see below.

Dimensions of vertical spacer

Raising spacers

There is another disadvantage to this solution: there is no more limitation in pivoting of the bogie, because the rotation stop is not long enough.

Here’s what it looks like in front of an LS Models travelling post office, which has a height exact to within 0.2 mm. The bellows are practically at the same height, as well as the buffers and the coupling heads. For the latter, the remaining shift is due to the fact that the NEM box is tilted downwards at REE, and upwards at LSM! Unfortunately there remains a space of about a millimetre between bellows.

Buffers and couplings

According to the standard, the retraction of the NEM box from the bellows must be 7.5 mm. On the LSM coach it is 7.8 mm, while it is only 7.0 mm on the REE. I do not quite understand why REE has provided heads that are supposedly long in addition, while its short heads are not short enough, and that a tight coupling could have been done thanks to the spring buffers!

Someone told on the Loco Revue forum that after shortening the drawbars, he had noticed a buffer mess when crossing a curve followed by a counter-curve (turnouts). One should try by blocking the buffers back from the plane of the bellows.

Two years after: the (better) solution

In some cases, there are derailments due at least partly to the pseudo 1-point suspension. Moreover, as already seen, this spacer could prove to be troublesome on the beginning of a slope. So, I’ll try to maintain the current spherical ball bearing, but by raising it. What made me hesitate is that we must cut level with the frame, with risks of damaging the reliefs.

Now, I noticed that sawing was possible with a circular saw blade, provided that the blade screw head is not more than 1 mm high. So, I ground that head. Another condition: the disc must be less than 20 mm in diameter, and must be as thin as possible (mine is 1/10 mm). Finally, it is necessary to operate at low speed to prevent the plastic from melting.

Cutting the ball pivot

Click the image to see the details.

The raising spacer should be about 0.75 mm thick. It will be made of polystyrene to be glued with model glue. Indeed, considering the method that we will see below, I want to avoid cyanoacrylate or epoxy glue, which would be however stronger, but which would possibly block the assembly.

So, the spacer is cut into a 0.75 mm Evergreen card, with a ø 4.5 mm punch, and then drilled in the centre with a diameter of 1.6 mm. A 1.5 mm brass tube (or other metallic round bar) is mounted in the mandrel of the mini-drill. The coach is installed in the vice. The position of the tube is adjusted with the XY table to be exactly above the pivot hole.

Pivot parts with the thickening spacer

Click the image to see the details.

The ball pivot is slipped onto the brass tube, coated with glue, then the spacer meets it, also coated with glue, and the whole is lowered until it leans on the chassis. The mini-drill is locked in position (see Stand enhancements) during the time required for the glue to dry. This method makes it possible to obtain the best possible alignment for the assembled parts.

Gluing the pivot

Click the image to see the details.

The bogie can then be reassembled (not without having cut the old spacer — that of the not so good solution — for me), without too much tightening the screw for fear of unsticking the ball bearing.

While I was at it, I shortened the drawbar boxes by 0.5 mm to conform to the NEM standard concerning the distance between buffers and bellows.

To hold the coupling heads in their new position, I drilled horizontally through the NEM box and the coupling tail at 0.6 mm diameter and inserted a small nickel silver pin. Why horizontally? Because the REE NEM box has high and low walls that are not opposite to each other, as can be seen in this picture, and it was therefore impossible to properly hold a pin in a vertical position.

Shortened NEM box

I finally glued the buffers in the plane of the bellows, to avoid the buffer mess often observed.


Performed on my Fleischmann test layout with unlikely curves (Rmin = 356 mm). Test train set: the two DEV coaches at the head of a convoy composed in addition with a Vru (dining car) LS Models and four Heris USI coaches very braking in these tight curves. Observation: it passes the curves except for one configuration, in pushing, for which a bogie derails. But I suspect the connection wires of the lighting (see page 4), poorly tidied and which are somewhat opposed to free rotation.

So, I’ll have to rectify the route of these wires before they enter the coach body.

Visual result

Here is a picture of a DEV coach coupled to the LS Models ex-CIWL 3345 dining car. There is still a light gap between the bellows, but the heights are perfectly compatible. The height of the couplings, on the other hand, is not!

Buffers and coupling with CIWL dining car