Society Badge Cheltenham Society
of Model Engineers
Next Public Afternoon
(1330 - 1600hrs)
Class 52 Project FacebookTwitter
Home The Society Calendars Galleries Projects Attic Links

Class 52 'Western' Diesel
by Barry Leach
Barry Leach

It was during a Public Running Afternoon in September 1999 that I decided that I would build a Diesel outline locomotive. What we need (I thought) is something different, something that will appeal to the younger visitors and something that I could have running by next season. A diesel seemed the answer - relatively uncomplicated (therefore, easy and quick to build), clean, and with no weight problems - cus it breaks down into several manageable parts.

I made some enquiries and was given the name of 'Dan Jeavons' as a supplier of parts for diesel locomotives. A phone call to Dan revealed that he was just completing the design of a Class 52 Diesel - a 'Western' class locomotive (since my 'train spotting' days I had a love of the Western Region 'Western' Class Diesels.
Western Pathfinder 1962

I purchased the drawings. However, they were basically only details for the powered bogies; there was very little information on the chassis/front end and next to nothing on the bodywork. However, there are two 'Western' Class diesels preserved at Bridgnorth on the Severn Valley Railway so getting pictures was not a problem.

The first bogie was completed in about 80 man-hours. No difficulties were found other than the usual 'lack of meat' on some of the castings, although allowances could be made and excess metal left on other items to compensate. However, one unfortunate 'problem' did appear. The bogies have six wheels with a main casting carrying the centre, non-powered axle. This axle has sprung axleboxes to keep the wheels in contact with the rails.

The outer axles are each carried on two plates that maintain the axle/motor gear mesh. These motor/axle assemblies pivot on the main casting with suspension being via a crossbar that connects to the main casting via a pair of resilient rubber bushes. The problem was that on each motor/axle assembly one of the two motor connection terminals exited the motor directly behind the suspension crossbar. This caused great difficulty when making the motor connections.
Terminal Obstruction

Rotating the motor or turning the motor assembly made no difference. My only option was to reduce the size of the suspension crossbar in the vicinity of the connection in order to allow the electrical connections to be made. With regard to the electrical connections, I do have some doubt. I question whether the push-on motor connectors are adequate for passing 10/15 Amps with negligible volt drop. A good soldered joint is the obvious answer but to do this means butchering the motor end caps to allow the wiring to pass through - something one doesn't really want to do on �40 motors. Advice states that the connectors are OK - but I remain sceptical.

Bogie Complete
The second bogie was completed in about 60 man-hours. Of course, the same 'problem' was found but at this point it became obvious what the cause was. The main casting is machined with an offset on each end, the offset being necessary to align the wheels of motor/axle assemblies on the correct line.

Each end of the casting has the offset on opposite sides so that the 'driven' wheels on each axle are on opposite sides of the bogie. If the offsets were reversed, the motors would have to be rotated 90� to realign with their mounting holes.

This would move the connection terminals away from any obstruction. A quick referral to the drawings showed that I had machined them correctly (or at least as per the drawing). The drawing view was upside down. The only view was from the bottom - it should have been the view from the top.

The aluminium buffer beams had been cast with tapered buffer stocks in-situ. The appearance of these is not prototypical and, therefore, not acceptable. The complete removal of the stock was contemplated but the fixing of any separate stock would be difficult due to the cast webbing on the rear of the beam. An external cutter was used to cut the outside of the buffer stocks into a parallel shank. Mild Steel rings were machined, with the correct outside profile, and loctited over the remaining aluminium 'stumps'. Small amounts of body filler were then used to put fillets in the corners etc. Hey Presto - correct looking buffer stocks.

Correct profile buffer stocks call for correct looking buffer heads. These are not nice, easy round ones but relatively large oval ones. Modern day computers now played their part in the project. I found a photograph with a full frontal view of a Western with a perfect shot of a buffer head.

Using the computer X/Y co-ordinates were measured to generate the required shape. X/Y co-ordinates were also calculated for generating the front/rear face radius. Mild Steel bar was turned to produce the buffer shank and a round head complete with the front and rear face radii. Using a Dividing Head on the Milling Machine one quarter of the buffer head shape was machined using the X/Y co-ordinate data. This was repeated for the other three-quarters and for the other three buffer heads - a very time consuming exercise not recommended if making oval buffer heads for a living. The 'steps' around the periphery of the head were then blended in. All in all the manufacture of these was a very satisfying task.

I wanted this locomotive to be as safe as possible for all potential users. Advice was sort from other society members and I purchased a '4QD' NCC35-24 controller, which has many built-in safety features. The unit has electronic ramping which only allows the locomotive speed to ramp up (or down) at a predetermined rate (user set). Also, if the locomotive is moving forward and the driver throws the reversing switch (which someone is bound to try), the speed will ramp down to zero before ramping up in the opposite direction. In addition, the hand controller will contain a "dead man's throttle", so that if released (by the driver dying or, perhaps, the locomotive breaking away from the train) the locomotive will decelerate to a halt. The controller also has regenerative braking, which passes all energy generated during locomotive braking back into the batteries.

The only 'safety' feature lacking is that of the speed controller failing 'short circuit', when the locomotive would continue at full speed until the batteries became exhausted. I covered this situation by fitting a high current relay, controlled by a light action toggle switch, in the wiring link between the two 12V batteries (it's a 24V system).

This switch is accessed through the large hole in the left hand brake box and, if pressed, effectively disconnects the batteries from the system. In addition, this feature protects the system should the batteries be connected the wrong way round. All in all I think every eventuality has been covered (unless anyone knows different?).
Emergency Switch

The speed controller came ready assembled in an aluminium case and this has been fitted below the chassis in space directly between the left and right hand brake boxes. This has made the space inside the body available for other things. Two 42 Ampere Hour Deep Discharge batteries fit neatly within the chassis. Hanging brackets were made and fitted, as at 15kg each I felt it best to keep the batteries as low as possible.


Whilst manufacturing the first bogie enquiries had been made with Diane Carney regarding name and numberplates. In the 1960's I had 'copped' all of the 'Westerns' except one - D1025 'Western Guardsman' - so as far as I was concerned there was no choice of name. The plates had been ordered and they arrived just in time for work to start on the body shell itself. They are etched in Nickel Silver and perfectly replicate the cast aluminium originals. A home of '81A' (Old Oak Common) was chosen for the shed plate carried by the Westerns in their early years.

Work on the fibreglass body shell was long and messy. Cutting and tidying the holes for the windows, footwells, handrails and exhaust vents produced copious amounts of dust and glass fibre. Lots of body filler was applied to mask the joins in the body sections and to improve the general shape. Bodyshell

Much filler was put on, and much was rubbed off with abundant amounts of 'wet & dry' and water being used. Two coats of primer along with coats of BR Maroon - it really looked a Western.

The final remaining item required to complete the bodywork (except for those 'rivet counting' details) was the transfer that locates on the side of the cab that doesn't carry a number plate (Westerns only carried one number plate on each end of the locomotive). However, this was not one of the many 'standard' transfers available from most Model Engineering Suppliers. Luckily I found the name of a specialist transfer company - Cambridge Custom Transfers - and they produced just what was needed (I'm not sure whether they still produce them).

The 'Western' was wired with the two 12V motors in each bogie wired in series, with the two bogies being wired in parallel across the speed controller. This gave the diesel a top speed of about 5mph which dropped to about 3.5mph when pulling about four adults on the 1:120 incline on the Society track. Whilst corresponding with another model engineer, who had also built a 'Western', it was found that his diesel had a top speed of 9mph with a load of 12 adults! This was with a gear ratio of 9:1, whereas my 'Western' had a ratio of 7.2:1. It transpired that he was running all his 12V motors in parallel across the 24V.

The 'Western' has now been re-wired to put the four 12V motors in parallel across the 24V controller. This has been tried on the Society track and has made a world of difference. The 'Western' pulled six adults plus myself around the track at a top speed of 10.5mph (at that speed it was most definitely not boring). At worst the controller was taking 40Amps. This value will have to be watched as the controller originally purchased is only rated at 35Amps (Hot)/50Amps (Cold). With the present performance the 70Amp controller would be preferable. This higher rated controller should be able to supply all the power demanded by the 'Western' in its new configuration.

Trials of the locomotive on the track have taught me two things. Driving a steam engine with the constant attention required is good fun and really keeps you on your toes. However, driving a diesel can be a bit boring - because there is little to do and more especially because there is no noise, no audible feedback - it is dead silent. What it needs is a diesel horn and, more importantly, simulated engine noise.

A standard 'Trax' diesel sound unit has been fitted inside the locomotive. This is connected to the 'wiper' of the speed control and increases the engine speed sound as the locomotive increases in speed. The output from the sound unit was originally connected into the two inputs of a 50W+50W Stereo Amplifier and the outputs from this connected to a pair of 5" diameter Kenwood speakers. The speakers are mounted into 'shaped' boxes made from 0.25" MDF. Everyone considers the sound to be very good, although I know it is not a genuine 'Western' sound. Recently the amplifier has had to be removed in order to fit some cab detail. The 'Trax' output now feeds direct into both speakers connected in parallel - and the sound seems just as good.
Western Guardsman, May 2001

Cab Detail
Cab detail has been added. Moulded plastic and cast aluminium items were purchased from 'Pheonix Locomotives' - many thanks to Keith McDonald for finally 'giving-in' to my badgering. General public do like to see a 'driver', even in our small locomotives.

The 'Western' has now become somewhat of a workhorse for the Society so it has been decided that further detailing of the locomotive would be unwise (i.e. it only gets knocked off with all the manual handling).

At least for now, the project is complete.


Hatherley Lane, Cheltenham, Glos, GL51 6PN, UK Tel: TBD Legal Notices
Home The Society Calendars Galleries Projects Attic Links