Positive Stops and Final Drive Sprockets

I needed 3 positive stops for the 3 immediate projects – KTT 55, KTT 305 and the Cammy Special. Both KTTs have a 3-speed gearbox and the cammy special will use the 4-speed ‘box.

This is the collection of positive stops and assorted parts that I have.

The one on the top right of the picture is a new-manufactured item that I bought on eBay Australia many years ago and is the only one that I have that will fit the 4-speed ‘box. The backplate, GC-2/2, is different for the 3 and 4-speed gearboxes – the top central hole is in a different place, although I could only find one part number in the spares book.

The other 2 complete positive stops are from KTT 55 and KTT 305. The latter has the wrong mounting bracket but luckily there is a correct bracket in the collection of bits at the bottom of the picture.

The new positive stop had been expertly crafted – a real work of art and superbly manufactured. This, for example is one of the pawls.

And this is the motley assortment of pawls from my old positive stops

and which were all pretty loose in their holes in the main body, GC-21 (referred to as the “Centre Piece” in the spares book – makes it sound like a floral table decoration for a dinner party!).

I decided that I would make some new pawls by copying the new ones that had come from Australia. These need to be hardened steel and so the first step was to order a bar of O1 tool steel of an appropriate size that I could subsequently heat treat.

I then chopped a piece off and after cleaning up the ends, set it up in the 4-jaw chuck in the lathe to machine the spindle.

The nominal ¼” diameter (actually 0.245”) spindle on both the Aussie pawls and my copies is slightly longer than the originals and as the longer spindle fits into my GC-21 main bodies, I retained this feature because I believe it is a better design as the pawl is better supported.

After measuring the angle of the tapered surfaces with a digital protractor

The pawl was set up in the rotary indexer in the milling machine to complete the machining.

After a few hours I had 2 new pairs of pawls – Aussie new ones on the left in the picture below.

These pawls turned out to be longer than the originals that were part of the 3-speed positive stops – 0.82” versus ~0.78”.  While I was making these, I compared a few other parts of the new positive stop with the originals. Interestingly, I found that the 3-speed ratchet lever, GC-22, was different: the picture below shows the new Australian manufactured one on top of an original.

I concluded that whoever made this in Australia redesigned much of the mechanism. Anyway, with the 4-speed ratchet lever fitted, it worked perfectly on the 4-speed gearbox with positive selection of all gears and neutral.

So, back to the 3-speed positive stops.

Before fitting and testing the new pawls, I needed to sort out 2 new linkages that connect the ratchet lever to the Striking Lever, B-65. I believe that I mentioned in the past about the importance of eliminating any sloppiness through wear in the linkage on hand-change vintage Nortons using Sturmey Archer gearboxes – it invariably results in difficulty in selecting or staying in gear. When rebuilding these Velo gearboxes I had found on all of them that the part referred to as the ”striking lever double knuckle joint”, B-69, was badly worn and introduced a lot of free play and I replaced all of mine. This part is (or, at least was) available from Grove Classics.

I also replaced the rest of the linkage: new ball joints, B-29, at the top and knuckle joints, GC-92, at the bottom. The original ball joints are not available but these,

which are a pretty good approximation, can be bought on eBay for 15 GBP for 4! The GC-92 parts can also be bought on eBay under the name “clevis fork yoke end”. These parts, together with making new threaded rods, replace the entire linkage connecting the gearbox and the positive stop and avoids any problems with gear selection arising from worn out parts.

Having copied pawls that were longer than the originals it was not surprising that the new pawls were too long for the original ratchet levers. The picture below shows the pawls in the 2^nd gear position where it can be seen that the one on the right does not engage with the notch.

Surprisingly, the longer pawls also didn’t fit in the new Australian manufactured GC-22 ratchet lever, again confirming that although the Aussie positive stop was the same in principle it had been completely redesigned in detail.

By removing 0.010” at a time from both pawls they fitted perfectly.

I found it important that adjustments are made, half a turn at a time, to the length of the connecting adjuster to test the gear selection and after a few iterations I managed to get perfect selection of all gears. It’s quite satisfying when it eventually works! …at least with the bike on the bench.

The pawls were then heat treated. I’ve described what I do for heat treatment in previous blog posts so won’t repeat it here.

The last little detail was to make a new cover plate, GC-33, for one of the positive stops to replace one that had rotted through.

It's not necessary to make these aluminium discs - they can be bought on eBay for a couple of pounds. 

 

Final Drive Sprockets – SL-91

The last job on refurbishing the gearboxes was to replace 2 of the final drive sprockets. These go under the Velo part number SL-91/X where “X” determines the number of teeth.

I had this collection of used sprockets.

Superficially, they might look OK but that’s only because I’ve cleaned them up and put them in the tumbler overnight. Apart from the number of teeth, there is one fundamental difference between them and that is the “insertion length” or depth of engagement of the sprocket onto the gearbox mainshaft. If you compare the depth of slots indicated by the red and blue arrows it can clearly be seen that the red one is much deeper – twice as deep in fact. I found that both of these fit perfectly onto my gearbox mainshafts and so one has to ask the question why use 1/8” of engagement rather than ¼ “? (I actually measured 0.265”).

I found the same characteristic in those available for sale and as I saw no reason to use the shorter engagement depth (or, conversely, I would consider the longer variant to be a better design), I opted to make new sprockets with the greater depth.

The starting point is 2x duplex sprockets with 21 teeth, pictured below with the worn out one to be copied.

These duplex sprockets are made of EN8 steel – quite tough enough for this application and are pretty good value at 25 GBP for 2, including postage. The first step was to hold the sprocket on the hub in the lathe, remove one set of teeth and form the small hub that will go onto the mainshaft.

The sprocket was then turned round and the entire large hub removed, creating a workshop full of cutting oil smoke and a small mountain of swarf in the process.

After repeating the lathe work for the 2^nd sprocket

they were set up in the rotary indexer on the milling machine to machine the slots

and drill the ¼” holes for the adjuster peg.

Incidentally, it would not be so easy to setup the sprockets on-centre in the chuck of the rotary table if they didn’t have a number-of-teeth that was divisible by 3.

After chemical blacking

they are now fitted to the gearboxes of KTTs 55 and 305.

And that’s about it for gearboxes, clutches, positive stops etc. About 4 months of work in total.

In fact, that’s about it for this year. As soon as the effects of Christmas have worn off, the grandchildren have gone back to their homes to play with their new toys and, hopefully, Father Christmas will have deposited a 4-axis CNC machine into my Christmas stocking it will be time to get back in the workshop and start work on the engines.

In the meantime,

Happy Christmas

 

4-Speed Gearbox for the Velocette Cammy Special

I have so far rebuilt 4 early Velo 3-speed gearboxes – those for KTT 55, KTT 305 and 2 flat-tankers. I had one last gearbox to prepare for all 5 Velo projects and this was my one-and-only 4-speed ‘box. It’s marked as #2 in the picture below.

I bought this ‘box many years ago and it was advertised as being rebuilt and in good condition. Superficially it looks fine and all the gears selected positively, however, I have learned never to take these assurances at face value. So, off with the cover to see how it looked inside.

Pretty good. Bearings had been renewed, teeth and dogs on the gears were good, there was a small repair to one of the lugs on the front but I would consider this gearbox to be in excellent condition and did not require any attention inside.

However, there is always something to do! When the positive stop was fitted it was immediately apparent that the “Outside Striking Lever” (Velocette parts book terminology), part number BK-65/3, that was fitted was wrong.

There are 2 fundamental differences In comparison to the B-65 component on the 3 speed gearboxes: BK-65/3 has a spline fitting rather than a square and it is 1” shorter. The lever that should be fitted is an earlier BK-65 that has the same spline but is longer.

I didn’t have a correct BK-65 lever but, luckily, there is an easy solution: add a 1” extension piece

made from a tough steel, EN24T, threaded internally and that positions the linkage to the positive stop in the correct place.

The only other part that I had to make was the cable adjuster for the clutch, CK-21, that screws into the top

by copying one from one of the other gearboxes. This, I found, had another weird thread of 19 TPI (the other one was the oil filler that I had to make, see here, with 13 TPI) which I cannot find on any thread chart for this diameter. Luckily, my 76 year old Harrison lathe can cut these obscure threads.

On the clutch side of the gearbox, a new thrust bearing and the wire clip for the thrust cup, BK-52, were replaced.

I have found on 2 of the gearboxes that I have rebuilt that the back of the clutch has contacted the wire clip and this had ground a flat on the wire – in one case it had nearly severed it into 2 pieces.

I have not yet decided what chaincase to fit but in the meantime, I have lock-wired the 4 screws that would fix the chaincase to the gearbox.

One of the 7-plate clutches that I rebuilt was fitted with new C6/2 Thrust Pins

and 16 new clutch springs

before putting on the sleeve gear nut with a new C8 washer.

All 5 gearboxes that I need - 3 for the immediate Velo projects and the 2 that I plan to work on subsequently, have now been rebuilt.

There are a couple of minor things to finish off but, essentially, this part of the project is complete. This particular gearbox is now in the bike.

This is the pile of bits that has been replaced (and there are a few more parts already in the bin) from the gearbox rebuilds.

Before moving on to the next part of the “transmissions sub- project” (which is to get 3 working positive stops) I should mention that it is essential to have a few special Velo tools to assemble and adjust the clutches, shown below.

Without these, it is pretty well impossible to set up a Velo clutch. I acquired these many years ago for my Venom and Thruxton ….not much changed from the 1920s to 1971.

I also have a lot of clutch parts left over.

If you want any of these bits then let me know (email address at top right of page) as I have no further use for them. There are some 3-plate clutches in this pile of bits. Ideally collect from my home close to Worthing, West Sussex or I can bring to a Kempton autojumble, Founders Day or the Banbury Run in 2025.

There is also a complete set of standard ratio gears in good condition available that I replaced with a CR set in KTT 305s gearbox.

 

And finally…..

For those of you that are members of the Velocette Owners Club and that read the clubs magazine, Fishtail, you will be aware that there has recently been a series of concise and well-written articles by Brian Agnew about flow and combustion in engines.

I spent 40+ years of my working life engaged in exactly this topic, namely developing, validating by way of experiment and applying Computational Fluid Dynamics software for in-cylinder flow, fuel injection and combustion in the design and development of IC engines. In his most recent article (Fishtail 504) Brian refers to “Barrel Swirl”, more often referred to in the industry as “Tumble” – flow rotation about an axis parallel to the crankshaft. To illustrate this, I poked around in some old directories on my PC and found an animation of an in-cylinder flow calculation for a modern 4-valve engine (well, reasonably modern – I retired 9 years ago) which I have shown below. The surfaces of the intake and exhaust ports with the piston at TDC in the cylinder look like this:

and the flow generated during the intake and compression strokes are shown in the animation below.

The flow structure in earlier designs of 4-valve engines, such as the Rudge, would be completely different. Intake ports in modern engines are designed to achieve separation as the flow enters the cylinder to generate the rotation, seen in the above animation. The rotational speed increases as the air is compressed during the compression stroke (conservation of angular momentum - spinning ice skaters decreasing their moment of inertia by moving their arms inwards) and dissipates into small-scale turbulence as the piston approaches TDC. As Brian mentioned in his article, it is the small-scale turbulence that determines the turbulent burning velocity - the speed that the flame propagates; without these processes engines simply wouldn't work.

Such calculations were carried out routinely for competing designs, more often with fuel injection to assess fuel-air mixing, in the R&D departments of major manufacturers. Whether that is still the case I doubt very much ...the engineers that I used to work with are now either working on electric vehicles or pensioned off.