|Vertical Bevel Assembly
In this section I will cover the last main area of assembly for the
SOHC engine, the vertical shaft and oldham coupling assembly.
When Arthur Carroll designed his new camshaft engine for Norton in 1929-30, the most visible difference from the earlier Walter Moore CS1 engine design was in the area of the vertical shaft design, and in the utilization of two new aluminum castings, one top and one bottom, that were designed primarily to house the vertical bevel gears, and allow accurate and captive location of the Oldham couplings, which in turn transfer drive from lower to top bevels via a vertical shaft between both.
These castings have a wide locating flange (approx 2 inches in diameter, by quarter of an inch deep) that allow them to positively locate into either the crankcase bevel chamber (for the lower) or the cambox bevel chamber (for the top), ensuring there is no possibility of lateral movement, which could result in the bevel gears being misaligned during assembly. The advantage of such a wide flange is that there is less chance of wear developing over the years, due to constant strip-down and reassembly. Although difficult to put into words, I always feel a strange sense of satisfaction when assembling these castings into either the crankcase or cambox, which I can only describe as feeling ‘engineeringly right’!
These bevel castings are secured into the parent castings by the use of four 5/16” studs, that pass through each corner of the casting, and then use special reduced head nuts to tighten down, as there is not enough room to allow normal width nuts (as a blatent advertising plug – I supply both the studs and the correct reduced nuts in stainless steel!).
Alumininium Bottom Bevel Casting.
|Pressed into each bevel casting is a plain phosphor bronze bearing that provides the main bearing support for the bevel gear, not only for the bevel gears shaft, but also as a thrust surface on which the rear of the bevel gear head itself can act. It is worth pointing out though that this latter bearing surface should not have any force bearing against it, which would mean the bevel gear is being ‘pushed’ into the bearing. This being the case, and the fact that the bevel gear is meshing with another gear would indicate that there is negative clearance or backlash in the bevel gears, which is very likely to result in premature failure of the bevel gears, or heavy wear at the very minimum. This is something that should always be checked carefully when assembling these components|
As well as the plain bearings, the castings also hold a second bearing this being a slim 2 row caged ball bearing. These twin row bearings do not actually support the bevel gears themselves, but act as a locator for the Oldham couplings that slide into the slotted end of the bevel gears.
A lot has been said about these bearings, much of it blasphemous!, due in main to the tendency for the bearings to shed its tiny balls as soon as there is any wear in the phosphor bronze cage, the result being that the balls can drop down into the crankcase bevel chamber where they can catch in the bottom bevel gear and result in broken teeth. The tongue in cheek view of my old friend Titch Allen (who was around when these engines were current technology) was that Norton’s considered this one ot their best design features, as it kept their maintenance depots in constant activity!
I think the idea of these bearings was that they were self centering, so they would have some lateral movement when you were removing the cambox from the engine, without having to remove the whole engine from the frame. They allow you to tip the cambox forward slightly, so it can be angled out from under the top frame tube. The problem is, that although the bearings allowed the cambox and vertical shaft to be moved ‘off centre’ so to speak, if there was any wear at all in the bronze cage, out would pop a ball or two and drop into the bottom bevel chamber.
These bearings have been obsolete for many years and trying to find them is becoming increasingly difficult. Because of this most people replace them with a rigid single row ball bearing alternative. Although these bearings do not lend themselves quite as easily to removing the cambox when the engine is in the frame, they are much easier to find, do not loose balls like the other bearing and I am told work just as well.
Next component in the bevel drive train is the Oldham Couplings themselves. These are tanged connectors in various lengths and they are designed so that the tangs on either end are 90 degrees apart from each other. They were originally supplied in various lengths, to take account of various compression ratios being used, the difference in length always being accounted for by the centre section width, the tangs themselves are always of a standard size. If you happen to come across an Oldham coupling with much shorter tangs (I have a couple of these in my collection), then chances are they are from the later DOHC racing engine and are not compatible with the SOHC engine.
By the way, the bottom Oldham coupling should always be of 3/16” thickness, only the top coupling can be varied in length, of which more in the next section. Also, it is very important that they are a good slide fit into the bevel gears and vertical shaft, without any traces of wear or looseness which may result in eventual failure. Quite often you will find that one end is slightly tighter than the other (and I don’t think it being tight does any harm), but there should be no perceivable play on the other side of the tang at all. I guess the reason they go like this is that inevitably they do wear in, and the wear develops in the direction of rotation.
Footnote Feb 2008: These Oldham couplings are not easy to find at all, and particularly in the longer lengths the chances of finding them at autojumbles is remote. I have been asked by lots of people if I can supply them? I am currently testing a test set in my own racing engine, and if all goes well I hope to be able to offer various sizes later in the year. The original type were normally machined then hardened afterwards. I am experimenting with making them in a much harder (heat treated steel), which means they should not need further hardening after machining.
Note also that the vertical shaft is hollow. If truthful I am not entirely sure if the original intention of this was to a) aid oil passage, b) aid breathing or c) make the coupling a stronger unit. Maybe someone out there knows more than I and can let me know so I can update this section?
Assembling the Vertical
Bevel Gear Assembly
Note : at this point, both sets of bevel gears should independently have their backlash set correctly. If after final assembly of the engine this has changed – something is wrong!
5. Trial assemble top half of engine with desired compression ratio.
At this stage do not fit Oldham couplings or vertical shaft. Measure
gap between end point of both top and bottom bevel gears (see below
on how to achieve this)
|More details of Vertical Shaft Assembly
As you can see from the summary above, setting of the correct backlash for each set of bevel gears is covered in other articles, so I will not bother going over that ground again.
If we pick up from Step 5,
this stage is intended to measure the correct gap between the two vertical
gears, so we can calculate the correct width top Oldham Coupling. I
am sure you will have noticed if you have tried this, that once the
engine is assembled, it is actually very difficult to get a ruler between
the two bevel castings without fouling on the castings themselves. I
suppose if you have gauge blocks or a similar set of engineering measuring
components that could be used to stack between the gap, then it would
be possible to quite accurately measure the gap.
Having gone through all the grief and pain of making sure that you have got this most critical part of assembly correct, everything from this point onwards is pretty straight forward, and more about final assembly than tolerances etc.
First of all, if like me you have done this operation with the engine out of the frame, the next step is to strip the cambox and head off again and do all the other smaller jobs associated with final assembly.
Once you are ready for that final assembly, and you have the head and barrel locked down for the final time, you should be ready to do the final assembly of the vertical shaft, and fitting of the cambox.
|By this point I am assuming you have each bevel casting correctly shimmed and assembled in their relative castings (i.e. crankcase and cambox) and have the Oldham couplings fitted into the bevel gear slots as well. As always, I like to ensure that there is plenty of (castor) oil swimming around at this stage and I have the engine at TDC and the cambox set with both cams off the lobe (i.e. TDC at compression stroke). It is then time to fit the vertical tube into the cambox, using the large shouldered gland nut and rubber seal. When I first assembled this engine back in 2005 I used the original type gland nuts, which takes a large single rubber seal (as shown in the photograph of the complete assembly out for inspection, above). These large rubber seals have a reputation for leaking at the best of times and I gather are now unobtainable, meaning the chances are, if you are using one, it is probably very 2nd hand by now! I have now started manufacturing these gland nuts in stainless steel, rather than the original chrome plated brass, and if I say so myself they are particularly nice, better than originals actually and I have made them with one big advantage – rather than the original single large rubber seal, my versions will accept 3 modern ‘O’ rings which are both readily available and almost totally eliminate the likelihood of leaking oil.|
Anyway, assuming you are still using the old original type (. . . and my blatent advert above has not convinced you otherwise!), the top one now need to be used to locate the vertical tube. I found before fitting, it is a good idea to smear the seal with a high melting point grease, to help assembly, then you need to slide it over the open end of the vertical tube and slowly move it up the tube until it comes against the beveled shoulder on the far end. Be very careful at this stage to ensure you have no sharp edges or nicks on the gland nut, or have picked up any grit on the rubber seal, as these will very easily find their way onto the vertical tube and leave you with a very prominent scratch the entire length of the tube – remember this is the central focal point of the entire engine!
The vertical tube on Norton Internationals and their other SOHC derivatives were all of similar manufacture, being a steel tube, chrome plated, with a pressed (or spun – not sure which) bevel at one end). The beveled end fitted to the cambox end, the final length then being determined by the size of the engine. There were commonly two types, 500cc and 350cc, but there were also much smaller quantities of the 596cc sidecar units made, which had a longer tube again.
These tubes look externally the samed as DOHC types, but actually the DOHC type fitted to Featherbed engines was wider and not interchangeable.
When first assembling the engine I found it very difficult to find one of these tubes that did not have some corrosion or pitting, and even if re-chromed, the re-polishing of them caused the tolerances to be removed even more, meaning they would be more inclined to leak oil. To jump ahead slightly, having originally assembled the engine using the original type single rubber seals, and a re-chromed tube, I found that the engine regularly leaked a small amount of oil from this tube/gland nut seal, so as well as manufacturing my own nuts, I also looked at having a batch of tubes manufactured from stainless steel.
|This turned out to be more difficult than I imagined, as the tube is very difficult to find in the original imperial size and now almost unobtainable. I did originally look at having the bezel spun (I even went out to a specialist ‘Spinning’ manufacturer – extremely interesting to watch), but eventually found that the same size tube is used on another well known motorcycle I have some involvement with, so decided to invest in having specialist pressing tools made, to make the lip exactly as the original in construction. I found that there is only one company in the UK that can now supply the relevant size in stainless steel (or so I am told!), and this I purchase in ‘ground polished’ finish – beautiful to look at, but unsurprisingly very expensive. Anyway the upshot of this, is that I am now able to supply both tubes, nuts and seals that not only look beautiful, but don’t rust and don’t leak oil!|
|Fitting Gland Nuts and Spacer Rings
Going back once again to the original build, using the original type gland nuts and large rubber seals, the other two items that will require fitting before screwing the gland nut into the cambox bevel casting, is the copper/asbestos ring that fits over the gland nut and the drilled spacer ring. The copper/asbestos ring, part number A11/762, should be replaced every time the nuts are un-tightened, as they crush down and therefore can only be used once. This was another of those items that you could never find one in good condition, and are of a special size, so again, I am able to supply these exactly as per the originals.
The final part that needs fitting before screwing up the nut is the pressed spacer ring – A11/757 and A11/759. I think these serve a couple purposes. First, as you screw down the gland nut against them, one side presses up against the outer race of the ball bearing on the opposite side of the spacer. I found when I was making the castings that these bearings are not really supposed to have more than 0.0005” - 0.001” (between one half and one thou!) interference at most, which is very little. I would therefore conclude that it is probably this spacer acting against the bearing that stops it from spinning in the housing.
Second purpose of the spacer, on the gland nut side, is that it acts as a stop for the vertical tube (top casting) or locator for the tube and stop for the nut (bottom casting). It should be noted that there are at least two sizes for these spacers, the smaller one being fitted on the top casting. However, I have seen various sizes of the larger one. Not sure why this is, but unless you know your spacer came out of the engine you have built, it is worth doing a trial assembly first, to ensure the gland nut does not lock against the spacer ring, before the copper/asbestos ring starts to crush – indicating it is too tall.
| As another point of interest
with these spacers, if you read the 1950 Norton Spare Parts Catalog (Yes
– I will be stocking it shortly!), you will see that this is the
only year where the catalog shows both the single knocker cambox (SOHC)
and the original customer twin knocker cambox (DOHC) together. These early
double knocker cambox’s were substantially different to the later
Featherbed engines, and they used a top vertical bevel casting almost
identical to the earlier SOHC type (i.e. very different to the later DOHC
Featherbed cambox bevel casting), but with a slightly thicker base flange.
Anyway, the 1950 catalog lists that for the 30M/40M that year (meaning
the DOHC, although both single knocker and twin knocker Manx’s were
made that year), the thicker bottom spacer ring is also used on the top,
which I assume is because of this thicker flange.
Just before I move away from this early DOHC cambox (not my specialist subject!), these early cambox engines are very rare now, as they are the favourite engine of the Cooper car boys. I have been asked if my bevel castings will fit these early cambox’s? Well yes they will, but I gather the original ones were slightly thicker, so the Oldham coupling may need to be a bit longer. If I ever have another batch of top castings made, I may see if the un-machined castings have enough extra metal to allow me to machine one of these later types.
To avoid damaging the bearings, I find
the best approach is to feed the vertical shaft into the bottom coupling
as best as possible, loosely tighten the cambox bolts, then put a spanner
on the camshaft nut (or engine drive nut) and gently rock it ‘to
and fro’, while jiggling the cambox and vertical tube. You should
find that eventually this results in the vertical tube sliding down
the lower coupling slot of its own accord, so it does not lock up and
cause any unnecessary stress – much more satisfactory.