CL012 CLUTCH

In rebuilding the 7-plate clutch (inner end plate, outer end plate, chainwheel, two insert plates, two plain plates) on my '50 rigid MAC, I came up with a few questions. For reference, I went through David Childs' article, "Identifying the Right Clutch Parts," and the measured dimensions of all the components agree exactly with his list. On that basis, it would appear that I have all the correct parts.
1. David's list states that the correct clutch springs are C12/3. The free and bound length for these springs are specified as 7/8" and 1/2," respectively. My measurements indicate that when the clutch is fully assembled, the free length of the springs (i.e. when the clutch is engaged) will be on the order of 0.600." I measured the full travel of the thrust bearing (when the clutch lever is pulled completely up against the grip) to be about 0.0625." Therefore, the compressed length of the springs (i.e. when the clutch is completely disengaged) is 0.530."
The amount by which the springs are compressed is a bit more than I expected. The processes of assembling the clutch compresses their initial free length of 7/8" by nearly 31%. Further, when the clutch is fully disengaged, the springs are only 0.030" longer than their bound length of 1/2." Is this really how everything is supposed to be set up?
2. I replaced the original ball-race type thrust bearing with the redesigned roller-type. The original bearing was badly worn, so it is difficult for me to assess whether the grooves in the races are purely a result of wear, or perhaps a groove was intentional when these thrust faces were originally manufactured.
The end result is that the new thrust bearing assembly is about 0.020" thicker overall than the original. When I pull the clutch lever hard up against the grip, the front face of the new thrust bearing actually touches the surface of the clutch inner end plate. I don't believe this situation would damage anything, since the end plate and thrust surface would already be rotating at the same speed. Is there typically some amount of clearance between these two surfaces? I haven't checked the alignment of the engine sprocket and chainwheel, but I suspect I could install a small shim between the distance tube and the clutch inner end plate to add a small amount of clearance if it is advisable. - Paul S. Greenberg 5/3/2003

You are right to have detected the tight clearances of Velocette clutch!
It really is too tightly engineered for a road bike, but as we all say "set it up right and it is the nicest, least variable clutch".
There should be no grooves in the thrust bearing surfaces.
If you modify the lifting mechanism to get more lift, as some of we racers have done, you can easily make the "frying pan" contact the backplate - and they don't rotate at the same speed so metal-to-metal wear is rapid.
The VSL thrust races are the same dimensions as the originals (without grooves!).
There are some deflections in the operating mechanism that reduce the eventual full motion of the spring carrier, however the springs are indeed nearly coil-bound when the clutch is fully disengaged. (It's just as well that the design of the Velocette clutch means that this dimension isn't affected by wear of the friction plates!) However, it is affected by the length of the distance piece and the length of the thrust pins. A longer distance piece such as B35/3 for the later 9-plate clutch (with its correspondingly longer thrust pins, C6/3) would reduce the pre-compression if the springs weren't changed, which is the longer spring C12/4 is required to compensate.
You could also increase the compression by using a non-standard clutch lever with more travel than the Velocette original (which is unlikely since the original had more travel than the standard lever on other makes; or you could reduce it slightly by not adjusting the shimming of the thrust cup to the required "parallel to the gearbox face at half lift".
There remains the possibility that you have a mis-matched set of components. There is a puzzle that we have been trying to resolve for some time. Velocette specified the longer spacer, B35/3 for the early 7-plate clutch on the S/A MAC, but no other components appear different from the 7-plate clutch on the rigid models which required the shorter B35/2. I covered some aspects of this in CL007 which I have just revisited and realise I made an incorrect reference to "shorter" when I should have said "longer", so I have inserted a correction as follows:
Some corrections are required to the distance piece info in the following para, mainly for "shorter" read "longer"! - The distance piece for the S/A 9-plate clutches (and also for the early S/A MAC according to the manual), is the B35/3 which is longer (11/16" or 0.6875") than the one for the rigid 7-plate clutches (B35/2, length 5/8" or 0.625"), not shorter as guessed at by TW at the time of writing. This presumably is to move the backplate on the 9-plate clutch out by 0.0625" to keep the chainwheel teeth in the same position because the inserts on the 9-plate friction plates are documented as C25/2 at 0.155" versus C25 at 0.195" on the 7-plate clutch (my measurements differ as below, but who knows if an insert is 'original' nowadays) . Since there are one-and-a-half inserts to allow for (one friction plate plus a half of the insert in the chainwheel), the difference needed is about 0.060". However, the early S/A MAC is a puzzle to which no one seems to know the definitive answer. The Velocette parts manual for the early S/A MAC specifies the older, thicker inserts C25 but the later, longer B35/3 distance piece........ The puzzle is that this eems to give the correct chain-line on many early S/A MACs, such as mine.... With so many variables I think the main thing is to use the distance piece that makes the primary chain alignment correct. - TW 7/3/2003
There should be some clearance between the thrust race and the backplate when fully lifted, though we don't have a figure. Perhaps you have a mixture of components that needs a longer distance piece (much better than a shim), but I think first check your primary chain alignment and let us know what you find. - TW 6/3/2003

We have since confirmed identical dimensions on a similar MAC, so although the clearances appear curiously tight, and the springs are compressed disturbingly close to their bound length, things seem to be set up as designed. Again, all dimensions agree with David Childs list of specifications.
Excusing my relative naivety of Velo terminology, but I'm not quite sure which part is referred to as the "frying pan." 
However, the following is clear: 
i) the front and back clutch plates are splined together, and therefore always have the same rotational speed. 
ii) Once it comes in contact with the thrust pins, the face of the thrust bearing now rotates at the same speed as the clutch. Hence, if the face of this bearing does undergo enough travel to touch the inner clutch plate, the two pieces will both be rotating at the same speed as well. 
iii) Since the front face of the thrust bearing protrudes past the face of the thrust cup itself, it doesn't appear that there is any other part that could eventually come in contact with the inner plate, that would not be rotating at the same speed (this latter comment refers to your statement: "as some of we racers have done, you can easily make the "frying pan" contact the backplate - and they don't rotate at the same speed so metal-to-metal wear is rapid," which obviously lost me somewhere along the way). 
The two of us presently rebuilding our MAC clutches now feel that we have everything in order. As we went through everything, some off the dimensions struck us as a bit peculiar. However, having sought guidance from a variety of sources, it appears that we have all the correct components. David Childs compendium of specifications is quite useful, so I hope he has received adequate compliments for his efforts.
The fellow here in the states that has proven to be of enormous value is Ed Gilkison. Not only does he know these machines intimately, his supply of parts is daunting. We are also pleased with the re-engineered clutches available from Dave Smith. His units appear to be well designed, and we are looking forward to exercising them strenuously when everything is finally reassembled. - Paul S. Greenberg 6/3/2003

As you've worked out by comparing with another clutch, yours is probably OK so if the primary chainline is good then I would leave things as they are.
The "frying pan" is C29/26 Thrust Cup.
Your conclusions i) and ii) are correct; and if you overtighten the clutch cable you can eventually feel a distinct limit to the motion when either the frying pan or the thrust bearing contacts the backplate.
Your point iii) is correct in saying that the thrust bearing is proud of the thrust cup/frying pan but the frying pan is tilted by the operating mechanism whereas the thrust bearing is self-aligning (though only after significant revolution). So the extreme end of the stationary frying pan can and does rub on the back of the rapidly rotating backplate and both interferes with the clutch action and wears away. So we racers ground a profile on that end when we used a mushroom headed pushrod to get greater movement of the thrust bearing to reduce drag caused by squeezing a 21st spring into the clutch with increased pre-load on the springs to cope with 30% more power while still having no drag so our push-start would get us away from the start line ahead of the all the others............ Daft, really, in retrospect. - TW 7/3/2003