How to … change DMU split gears

Good morning all

Thought I would share with you something I have been working on since yesterday. As we know, the press-fit 25 and 16 tooth gears on the armature and drive shafts of the DMU (full chassis version) are notorious for splitting and, to my knowledge, the said gears are no longer available (I’m not sure about NGS, but since the shop is closed, I know of nowhere else who has them for sale). UPDATE: 25 Jan 2016: Please see the comment from BR Lines below; the 25 tooth gears were always available and the 16 tooth gears are now back in stock.

These 25 and 16 tooth gears are the same as the ones you will find in your bogie towers for various Poole built diesels, except for one quite important difference. They have a smaller diameter hole in the middle. This is because they are designed to be “press-fitted” onto the armature / drive shaft for a tight fit. You will immediately see that a 25 or 16 tooth gear designed for the bogie towers will not fit on the armature / drive shafts, since the whole in the middle is too big.

So, now we have a problem. With these press-fit gears having split (I’d hazard a guess at saying on nearly all DMU’s) and with no replacements available, currently available from BR Lines, see comment below this post what is the solution? Having thought about this for a while, and knowing that the bogie tower 25 and 16 tooth gears were otherwise identical, I focused my efforts on using these somehow.

After examining my guinea pig DMU I was pleased to find that the 25 tooth (larger) gears were fine, still snugly press-fitted on to their shafts. It was the two 16 tooth gears, attached to the armature, which had split. These were duly removed.

Now, the solution I came up with, even to me, sounded as though it wouldn’t work. However, I had nothing to lose and so thought it must be worth a try. My solution was to take two 16 tooth (normal) gears used in the bogie towers and, due to their larger diameter hole, epoxy them on to the armature shaft.

My reservations as to this not being successful were as follows:

1) Longitudinal alignment – will the glued on 16 tooth gears align longitudinally with the 25 tooth gears?

2) How would I get the 16 tooth gear concentrically placed around the shaft (would rotation give a wonky gear)?

3) Would the epoxy hold?

Below you can see a picture of the two gears in question. These are the ones at each end of the armature.


So, on we move to actually attaching the gears. I used two-part epoxy (Araldite) and the plan was to let it set overnight for a secure bond. the first attempt resulted in slightly too much epoxy on the shaft and this migrated to the inner white bearing. Absolutely no good, so required removing, wiping clean and starting again on that end. Second attempt, with less epoxy, was successful. I glued both gears at the same time, but it might be better to do one one night, followed by the other the next. The reason for this is that if you do need to make adjustments at one end, it is very tricky not to disturb the other end, which might otherwise be fine.

So, with epoxy applied, it was time to let it set. Whilst this was happening the rest of the chassis and mechanism had the full Rolls Royce treatment – complete strip down, clean / wash etc. Indeed it didn’t really need it, since the chassis was clean and looked quite fresh. I bought this particular DMU some years ago second hand, but I suspect it was sold on after the gears split, which was most probably not long after purchase, hence I doubt it has seen much running. The brushes and clean condition of the pick-ups would also suggest this.

So, with all the chassis parts nice and clean, and the epoxy setting, there was nothing else to do until the morning. A quick check every few hours to see that the shaft was freely rotating in the bearings was all that was really needed. I didn’t want to wake up in the morning to find that the bearings had bonded to the shaft – that most definitely would be the end, or at least return to square one, of this experiment.

So, this morning I dropped the armature back into the chassis and checked that it aligned with the 25 tooth gears. Hey presto! Also, twiddling the pinions with my thumbs, I was happy that there was free rotation. Still, I wasn’t assured of success and I still didn’t know what would happen when the chassis underframe was reattached and power was supplied. With great anticipation I proceeded. To my joy the armature turned sweetly, freely at low speed and without any evidence of gears binding, biting, or any other symptom that I thought might be likely.

It was drawing 120mA without the bogies attached, which to me is more than acceptable and what I would expect from comparable large diesel chassis (class 37/47). One thing I did notice, however, and I cannot be certain that this is related to the gear change or not, is that slight slop in one of the long pinion drive shafts (let’s call it that) meant that the end of the shaft could possibly make contact with the chassis end. The picture at the end is of the “good end” but it shows that there is a small hole at the end of the chassis block (it’s actually for the screw holding on the metal bogie retainers) but it also doubles up as giving clearance for the shaft ends.


Given this recess at the ends, I was happy that the shaft end was clear of the chassis. I oiled very lightly the shaft end anyway, just in case there was the slightest contact (although this would have ultimately meant a strip down and reassessment).

So, forward it was with the reattachment of the bogies. I still didn’t know for sure if this whole thing was going to work. Perhaps the load of the bogie gears would somehow fail the project. With the bogie gears reattached, power was applied. Hey presto! She purred sweetly, almost as if new. She was now drawing between 120mA – 130mA, barely any change from bogieless, which suggested to me that the bogies were not introducing any undue load. Even I would have expected a jump of 20mA or so with bogie reattached (certainly I would on other diesels, but perhaps the gearing design on these chassis helps).

Anyway, popped on the track, and after an intial stutter, she was running smoothly. She’s now being run in on the rolling road and I’m listening for any odd sounds, feeling for heat, and looking for any wobbles. So far so good!

The most important point to note, actually two, are these;

1) The 16 tooth gears had to be glued flush with the end of the armature shaft for alignment with the 25 tooth gears

2) Do not use too much glue so as it migrates to the bearings. Carry out periodic checks as the glue sets, and if you notice it has migrated, strip and wipe clean. I would say if it is going to migrate, it is going to happen early in the setting process, whilst the glue is less viscous.

Now, I know this isn’t a conventional fix, and it may have its detractors. However, with the lack of availability of press-fit gears (perhaps forevermore) available from BR Lines, see comment below I needed to come up with a plan which would enable me to run these DMU’s again. The adverse effects, as far as I can see, are so far none. It might just be that I was “lucky” with my alignment – I won’t know until I try this on a second DMU.

The point is, as certain spares become scare, those few of us who still enjoy tinkering with Poole models will be faced with more and more such challenges. The enjoyment is coming up with a solution. So far, in this case, I think I have found one.

Leave a Reply