About two years ago Peco added a curved turnout to their O gauge range. At the same time we, as a club, were looking to see how we could (a) improve our test track and (b) better adapt it for use as the hidden sidings for our new layout. Peco’s new turnout looked to be something of a heaven sent opportunity as it allowed the test track pointwork to be moved onto the curved end sections. That would allow the loop tracks, of which there are two in each of the test track’s four circuits, to be extended to the full 20 foot length of the straight section, which, for layout operation, would allow two trains to stand on each loop. In addition, since the test track comprised four circuits, and the layout only two, crossovers were needed to get between the inner and outer pairs of circuits. We had already produced hand built curved crossovers to fit within the notionally 4 foot length of each of two of the test track baseboards, but never actually commissioned them. In fact, we had doubts that we could make the lengthy switch rails work reliably without having to provide a secondary drive.

In the circumstances, adopting the new Peco turnouts looked to be a worthwhile option and, as a copy of the Peco general arrangement drawing for these had by then appeared on Tower Models’ website, plans were drawn up for the new track layout. On paper, it worked, although when it came to reality, the actual geometry of the Peco turnout was not quite what had been deduced from the drawing. Critically, the section through the crossing area is not curved, a result, we believe, of Peco having used the same sleeper moulding as used in their standard large radius turnouts. The reasons for this are understandable, but the result of inserting the best part of a 5 inch long straight into the curves is that the radii through the centre part of the turnout are tighter than expected, and the angles for each of the two legs are smaller than expected. That creates particular problems for crossovers where both ends are on the curve, particular if, as in our case, the separation between the two tracks is not constant.

Once we had laid the turnouts, twenty of them, and coped with the difficulties of their geometry, operating the test track turned up a whole range of design defects with the Peco turnout, not all of which are readily curable, or curable at all. Any test track has to be capable of accepting a wide range of locomotives; the discovery of these problems was something of a step backwards, particularly when it was visiting locomotives that were exposing them. We will not have been the only people to have used, or be contemplating the use of, these turnouts hence it seemed appropriate to set out what we discovered, and what can be done about it. In no particular order, they are:-

1 As manufactured, both switch rails are electrically connected to the crossing as per Peco’s normal practice. A consequence of this is that for the open switch, full track voltage is present across the gap between the switch and stock rail. For the straight turnouts, this gap is large enough not to be a problem however the same is not true for the curved turnout. Peco’s use of the existing switch rail design, whilst understandable, together with the curvature of the turnout combine to significantly reduce the flangeway gap. Because there is so much lateral slop between British finescale wheelsets and the track in 32mm gauge, contact can occur between the back of the flanges and the heel joint of the open switch rail, creating a short circuit. The fact that on curved track the leading wheelsets of any vehicle will run with the flange in contact with the outer rail simply turns a possibility into a certainty.

Fortunately, the solution is simple. All that is required is to remove the bonding connections that Peco fitted underneath the crossing and reconnect the switch rails to their adjacent stock rails. To be fair to Peco, they do mention this in the instructions (which we might have read first, had we appreciated the problem). Strangely, the sleeper mouldings do allow for such bonding connections to be factory fitted; why Peco didn’t bond the switch and stock rails is a puzzle.

2 One of the problems inherent with 32mm gauge and modern day finescale wheel standards is that there is far too much lateral slop between wheel and rail. Consequently it is inevitable that as the wheels arrive at the check rail they will strike the flared entry and be pulled sideways. As manufactured, the entry flare is anything but gentle with the result that instead of being guided into the flangeway, wheels were simply riding up the check rail and then, in the absence of any guidance, going the wrong side of the crossing nose. Locomotives with side control on a leading bogie or pony truck are particularly susceptible to this, as the action of the transverse springing is to cause the weight on the inner wheels, which are being engaged by the check rail, to be reduced, making it easier for them to climb out of the check rail. Simply bending the check rail ends to create a more gentle flare did effect some improvement but not a complete cure.

In the end, we have replaced the Peco check rail on the inner curve with one that is considerably longer. At the toe end, the rail has been extended by three sleepers, by which time the entry is wide enough to remove any need to bend the check rail; the flare is effectively provided by the curvature of the running rail. Some consideration has been given to replacing the standard bullhead section check rail by one made from 4 x 1mm brass strip, but since this requires a destructive alteration to the sleeper moulding, we have so far held back from this option.

3 In investigating the above, it has also been found that the check rail flangeway gap can vary between 1.75mm, which is appropriate for 32mm gauge, and almost 2.0mm, which is not. Added to that, the as-fitted check rails were also found to be loose on some turnouts, allowing them to tip inwards and exacerbating the problem.

4 Despite the fact that they had all but disappeared by the time Queen Victoria had departed, Peco have managed to incorporate several inside keyed chairs in the turnout, on the last two bearers at the crossing end, where they get struck by the wheel flanges. These are, fortunately, easy to resolve simply by cutting or filing down to the key.

5 For the first time, Peco have designed these turnouts to incorporate a micro-switch (obtainable separately) for controlling the crossing polarity, actuated by the profiled inner edge of the stretcher bar. Unfortunately, the geometry of the design is such that the roller on the micro-switch follower doesn’t always make it past the step on the stretcher bar. Its spring force is sufficient to overcome the over-centre spring provided by Peco in the turnout mechanism, so that it can become difficult to get the switches to stay over in the absence of an external point motor or other mechanical actuator.

6 Useful as the micro-switch is, it requires a more substantial housing between the rails than on the normal Peco turnout. In theory, the cover shouldn’t be proud of the rails. In practice, however, we have found that unless the micro-switch seats perfectly in its mounting, the top cover to the micro-switch housing stands sufficiently proud to catch low hanging parts of locomotives such as ATC shoes and tender water scoops, with predictable results.

Apart from very careful installation, which is particularly sensitive to the size and attachment of the wire tails to the micro-switch, the obvious alternative is to mount the micro-switch at the side of the track to be operated by the end of the stretcher bar. Leaving the top cover off also helps, but creates the risk that the over-centre spring provided by Peco can come out.

7 As with their other turnouts, the switch rails fit into joggled stock rails. The snag is that whereas the joggle in a real switch rail is forged, with rounded corners, the Peco version is machined, and thus square ended. This is another feature that was unimportant on the ordinary turnouts but becomes critical once the turnout is curved. The problem is that on the curve, the leading outer wheels run in flange contact with the rail. Once they reach the “joggle”, instead of being guided round offset, they end up striking the end of the recess, resulting in derailment. The remedy has been to file/grind away the sharp edge of the recess, giving a flared face to guide the wheel.

All in all, we have ended up being more than a little disappointed by these turnouts, to the extent where (a) it is arguable that they are not fit for purpose and (b) had we known, we might well have commissioned hand-built turnouts instead where we would have had control of the geometry. Having discovered, via the Yahoo 7mm e-group, the contact details for Peco’s engineering manager, I put the above comments to them, and had a useful but inconclusive discussion with them at Telford last September. All I can say is that the comments, which I was not alone in making, were well received, but subsequent feedback from them has been nil.