HELJAN CLASS 23 HEADCODE BOX MODIFICATIONS   -   PAUL JAMES
of D5901 at Doncaster works courtesy of nigelmenzies on flickr
Originally I had not intended own a Heljan class 23 Baby Deltic. Kier had three for Hornsey Broadway, and Greg Brookes had a couple lurking on his Shenston Road layout. With only 10 in the class I didn’t think about getting one. However a lot of talking had gone on about the mistakes on the model, especially the size, or lack of it, on the headcode boxes. Eventually the challenge was set. If they are the wrong size, what size should they be? The gauntlet was thrown down and a model duly bought. After a bit of thought, and a lot of searching on the internet, a head on view photo was finally found of a Baby Deltic, D5901 at Doncaster works awaiting scrapping. 13 June 1976. The photo was reduced down as best as I could to 4mm scale. The headcode box was cut out of a print of the photo and stuck onto the nose of the Heljan model to see how it looked for size. This was briefly eluded to in the April 2016 update.
Class 23 nose comparison - headcode photoshopped onto model.
After discussion about the effectiveness of the “paper headcode” was favourable, the next question was size and position. A quick look through my archives found an article by Monty Wells in the May 1983 edition of the Railway Modeller on making a class 23 from a Hornby class 37. This gave a lot of information and drawings regarding the Baby Deltic body, in particular the headcode boxes. A comparison between my paper one and Monty’s in size was close, my paper one being slightly narrower than Monty’s. When comparing the two sets of measurements on the actual Heljan body I decided to stick with my paper headcode measurements. These are set out below. (An amendment of Monty’s original drawing).
Once I had decided on the size of the new headcode box, what to make it out of? I had a couple of headcode box fitted Lima Class 20’s sitting in the pending box. The cab end headcode box on one measured up pretty close, being alright for height, but a little bit wide. They proved to be easy to get off the cab end of the loco, just being held in position by a couple of spigots into the body. They were rounded at the corners, but in reducing the width of them this would disappear. The sides of the headcode boxes were filed down until they were the right length, care being taken to make sure the loss was even on both sides.
The before and after can be seen below. I had originally intended to cut out the original headcode aperture from the Lima 20 headcode box and then use that as the new headcode, but soon realised there were a few problems with this idea.
I decided to cut a new front plate for the headcode box from some thin plasticard I had in the spares box....
.....and stick that on the rear of the Lima headcode box, and then cut off the original front part of the now reversed headcode box.
New headcode in position (under the masking tape). Note the original profile of the nose where it meets the sides. With the replacement headcode boxes finished, the next question was exactly where to fit them on the nose. Looking at as many photos of Class 23 nose ends, and more studious measuring resulted in me deciding that the top of the new headcode box would be the same position as the Heljan original. This would mean that the bottom edge of the new headcode box would be lower down the nose. This seemed to tie in with the photos and my measurement, as well as what was in Monty Well’s original article. The width of the new headcode box was just about the same as the Heljan one, so that would help in fixing the centre line of the new one. The original Heljan headcode box was drilled out and carefully filed back so that the top of the new headcode box would be in the same position as the original. The sides of the headcode box were done in the same way, filed back until the new one fitted centrally. With the top and sides of the new headcode box fixed, the replacement was placed in position and the new bottom of the headcode box marked up on the nose. More careful filing and frequent rechecking resulted in a slot in the front of the locos nose that the new headcode box could fit snuggly into. With careful lining up, I used the step down in the size of the buffer shanks as a reference for how much the headcode box stuck out from the nose, and when satisfied all was square, the new headcode boxes were glued in position. The photo below shows the new headcode box in position, and also the original profile on the corner of the Heljan nose.
Another thing that flagged up with looking at all the Class 23 photos was the profile of the nose where the top meets the side. There had been some discussion as to Heljan not getting it completely right with their first model, but correction it for the later model of the early disc version of the loco. This seemed to be possible to fix with the careful application of a fine file, some sanding sticks and finally very fine emery cloth designed for getting scratches out of Perspex. The patient application of the latter resulted in barely noticeable damage to the overall plastic body colour, which is green, and means I can get away without having to respray the nose, any slight difference in colour being easily masked with a bit of weathering.
Once I was happy all was good, the yellow was re-applied to the nose and glazing fitted to the headcode boxes using some scrap perspex sheet in the scrap box. New headcodes, as seen on prototype locos in photographs were manufactured on the computer and printed off to fit into the headcode boxes. The resulting look can be seen below and in the photos below.
At the end of this work I was happy that I had improved on the look of the Heljan headcode fitted Baby Deltic model, and with some test running under its belt was all set up the weather the loco. After discussing some of the other faults on the loco, the main one being that the cooling fan grill didn't line up with the radiators on the side of the loco, and was I going to do something about them. Challenge number two was set in motion. All will be revealed in Part 2.
Image of D5901 at Doncaster works courtesy of nigelmenzies on flickr