Clutch Adapter Differences

[BagONails]

Hi gang,
With the gearbox back together I am back on the primary drive side.

Comparing the replacement clutch adapter to the old one while dry building and checking alignments etc.  I started by trial fitting the new one on the main-shaft taper and it didn’t feel like it was fitting very well. After lapping in without much improvement I blue’d up the main-shaft and found only 40% contact between the two tapers. I repeated this check with the original part and it showed at least 90% contact which would only improve with the nut being done up behind it.
Following John CL’s excellent write up I also measured the amount of thread showing at the end with the new adapter in position and compared this to the old one, my results are set out below:
Results
Threaded length showing
Old – 15.3      New 11.75    ( so the old one is at least in the same ball park as John's checks showed)

Taper Dimensions (mm, sorry!)
                   Old               New              Difference
Small Dia       16.27              15.8  (16*)   0.47
Large Dia      18.8              18.15 (19.5*)   0.65
Length         18.4                17.9              0.5
Angle (calc)    3.93                3.75         0.18 (in degrees)

Sorry the editor doesn't seem to like my attempts to make a table...In summary the new part does not sit anywhere near far enough onto the shaft, it will cause the clutch chainwheel to be misaligned to the engine sprocket and there will be insufficient threads engaged in the nut.
 
The  figures in brackets marked * are the suppliers dimensions from their page. It is interesting to note they also sell an alternative adapter with larger diameters namely 16.6  x 20.7 stating “The bigger taper allows the clutch hub to move IN towards the gearbox by a couple of mm, helping with chain alignment.”  Fact remains that neither of these parts are remotely similar to my original and the new “standard” one I have does not even correspond with the dimensions given by the supplier, not even close. I find it strange that this feature of the adapter is so wrong when the rest of it seems to be made very well. It is a perfect fit in the hub and seems to have been ground with high precision…
 
Does anyone know what the taper angle is actually supposed to be on the main shaft.
I have set the original up in the 4 jaw on the lathe and got it running true , I then clocked the internal taper to set the compound round to suit. I’m going to use the original adapter as my master to get the taper angle and machine the new one. Boring the taper on my lathe is a bit dicey because it is hardened and there’s a dirty great keyway right through the centre which carbide doesn’t like very much.  My calculations tell me I need to go 0.4 -0.5mm larger on diameter to move the clutch in by 3 – 3.5mm.

and there I was thinking I’d be out on the next club ride…nah!

[chaterlea25]

Hi Ian,
I would bet you got a Triumph adaptor?
When I went about correcting a t1aper on a pre war BSA I had to set up a tool post grinder

John

[BagONails]

I don't know John this is the exact part I ordered unless they sent the wrong one, possible I suppose but all the other parts seem OK, all ordered together.

https://www.feked.com/bsa-4-spring-clutch-hub-adaptor-standard-taper-a-and-b-group-models-circa-1954-63.html

Yes I might still have to do that, I've been test cutting the old one as it is effectively scrap now (bearing race worn), I'm making swarf which is a start, I will then re fit it to see how it goes and then try to repeat with the new one.

What did you use a Dremel type grinding spindle?

[chaterlea25]

Hi Ian,
Well for starters A and B models need different adaptors, the single cylinder bikes need an adaptor that fits further  onto the shaft
I think You got a bum part 

I set up a die grinder in an attachment that fitted onto the lathe tool post,
To true up the stone wheel, I held the grinder in the lathe chuck jaws with the airline through the head stock and the diamond dressing stick in a tool holder, I could not get a any kind of finish until I dressed the stone wheel

Some time ago I modified one of the new repro adaptors to use with a Pearson clutch, they are a lot softer than the original BSA ones and turned relatively easily

John

[BSA_54A10]

John has nailed it you have been supplied the wrong part.
Unless you are very familiar with them it is an easy mistake to make
BSA used the same taper on all of the main shafts up until the OIF models .
The only read difference is how deep the taper is cut into the adapter .
BAck a while when we had specialised parts suppliers these sorts of mix up were very rare.
However now days when the few  remaining parts suppliers cover almost every brand of British bikes  mix ups are a lot more common .
Give Fecked a ring
Good chance one of the store persons put the wrong parts into the bin and of course the order picker only knows to take 1 part from Bin 3 row 5 column 2 and put it in the bag for you.

[trevinoz]

You will find that the angle should be 4 degrees.

[BagONails]

Thanks for your interest guys, my thoughts as follows:

I set up a die grinder in an attachment that fitted onto the lathe tool post,
To true up the stone wheel, I held the grinder in the lathe chuck jaws with the airline through the head stock and the diamond dressing stick in a tool holder, I could not get a any kind of finish until I dressed the stone wheel

Some time ago I modified one of the new repro adaptors to use with a Pearson clutch, they are a lot softer than the original BSA ones and turned relatively easily

John

I can see how this would work John but unfortunately, I don't own a die grinder although I have been salivating over a Milwaukee cordless DG...I can't in all fairness really justify one for my home shop but I'll keep working on it! My test runs on the old adapter have been successful using a 12mm dia boring bar and a very small and sharp carbide insert.

John has nailed it you have been supplied the wrong part.
Unless you are very familiar with them it is an easy mistake to make
BSA used the same taper on all of the main shafts up until the OIF models .
The only read difference is how deep the taper is cut into the adapter .

You are both most likely correct.  I have been in touch with Feked and they say they sent the correct part. (as ordered) All the others are the same apparently and measure 16mm at the small end but they have no way of checking the large end... It sounds like they are using a rule. Not really adequate when 0.1mm (.004") on diameter on this taper changes its position along the shaft by 1.4mm! I measured the small end at 15.8mm with a vernier caliper (itself not the most accurate method) the size I came up with seems to correspond with it sitting over 3mm out of position made worse by the taper being too shallow (acute) and the contact is all at the top end (gear box end) so I end up with only 11mm of thread visible. It does seem the angle is different so not right for any BSA if they are all 4 degrees.  The thing is what other bikes use an adapter for the 4 spring clutch and do we know the angle on these mainshafts?

You will find that the angle should be 4 degrees.

I believe you are correct sir!
However, I am investing a fair amount of time here and I have no idea where my mainshaft came from so I have done some checks of my own! This entailed turning up a bush to check the angle.  You need two known diameters and the distance between them to accurately check the taper angle. The longer the distance the more accurate your result will be.
I came out with an answer of 4.045 degrees so pretty close. Due to the difficulty of taking really accurate measurements off the end of the m/shaft I checked the sensitivity to +/- 0.05mm on axial measurements and this changed the angle to something between 4.14 degs and 3.947degs so I think we can be confident setting the compound to cut a 4 degree included angle.  I can use the same method to check the angle inside the new adapter by  turning a pin with two different diameters and of a known length then check the heights it sits at and do some calcs to find the angle.  I might do this before a turn it to the new sizes just to prove a point to Feked but I doubt it will get me anywhere tbh.

[BagONails]

Hi Guys, been a bit quiet of late but plenty has been going on, so here's an update. (including being back in lock-down which has given me an ideal opportunity to progress things despite it being horribly cold and wet outside!)

I decided returning the adapter to the UK supplier would be a waste of time. Instead I'd try boring the taper out so it would sit 3 - 4 mm inboard and correct the angle to 4 degrees to fit nicely on my MS.  I had a practice with the old one which was OK but to my dismay I found the new part to be considerably tougher! I mucked around for almost a whole day but we got there in the end. This became a real trial because I had no way to gauge how much I had taken off the diameter without removing the part from the chuck and trying it on the MS. The thing was so hard I was doing little more than scratching away at it.  The cut I was putting on the dial was not what I was getting as result so it was touch and go as I didn't want to go oversize and only needed to open it up by about 0.4mm max! Couple this with having to clock the thing up in the four jaw every time it was taken out...anyway we got there in the end with a 15mm length of thread visible up from 11.5mm and an excellent taper fit after a bit of fine lapping to take the peaks off. The photo shows the general fit up now with full threads coming through the nut. (nut is tight but not fully torqued up)

Both new nuts gearbox sprocket with seal (SRM) and the clutch nut (Feked) in the picture were very tight on the threads and I had to spend some time with my rattle gun and some more grinding paste to get them fitting properly but all good now.

Now I can see the offset between the crankshaft sprocket and the clutch.  With the cush nut preloading the spring and my straight edge firmly against the crank sprocket it is laying parrallel to but 3.5mm off the side of the clutch chainwheel, pretty much on the chain wheel centre line confirmed by the view from the front.

I guess my biggest question now is how much offset can I safely accommodate by fitting a spacer behind the crankshaft cush drive adapter and will this result in excessive pre-load on the cush drive spring?

[berger]

i have not been to the pub for ages so i am swigging on the back garden and my personal opinion is bang it together after that effort, my dad hated people wanting a latherer job doing when he had set the 4 jaw up so cheers to you, good work . also i ended up with the grinding paste thread problem and bought taps and dies that will be used only once by me because of shoddy aftermarket bits and bobs. talking of bobs can anyone lend me a score for my next sesh on the piste you know you want to

[muskrat]

G'day BON.
I know what you went through. I had similar boring iron head valve spring collars to fit SR cotters (long story).
It will be a "suck and see" with a 3.5mm spacer. As long as the spring isn't coil bound when the cush nut is fully home and the sprocket can rotate 1/8th of a turn. I doubt the extra preload will damage anything, just meaning the "cush" will be a bit firmer.
Time to get out the verniers, compress the spring in a vice and measure it's length when coil bound. Put the nut on without the spring and measure  the distance where the spring should be. Now back to the spring in the vice and open it up to that measurement. The total amount of gap between the coils will give you how much the sprocket will move on the ramps.
Cheers

[RDfella]

Am I missing something here? Surely if a spacer is fitted behind the cush drive (as opposed to behind the sprocket) it will make no difference to the amount the spring is compressed?

[muskrat]

G'day RD.
My mistake is that it's the sleeve that moves.
If the spacer is behind the cush drive bearing it won't have any affect on the spring, the nut will just sit out 3.5mm further covering the split pin hole.
If between the engine sprocket and cush drive bearing it will.
Cheers
ps 5am and only 1 coffee.

[trevinoz]

Ian, from memory, 3/4" of thread protrudes from the adapter.

[BagONails]

G'day playmates!
Bergerman, you crack me up. I've not been to the pub for what seems like ages either although only locked down since last Tuesday night...hopefully we will be out of this by Weds.  All new cases found to date have been directly linked and already in quarantine so we are going well.  I would gladly buy you a beer if it was possible maybe when we next get to travel home though it could be sometime...In terms of just "banging it back together" I think this is what the PO did and it hasn't gone well..anyway I shall continue my investigations for now.

i have not been to the pub for ages so i am swigging on the back garden and my personal opinion is bang it together after that effort, my dad hated people wanting a latherer job doing when he had set the 4 jaw up so cheers to you, good work . also i ended up with the grinding paste thread problem and bought taps and dies that will be used only once by me because of shoddy aftermarket bits and bobs. talking of bobs can anyone lend me a score for my next sesh on the piste you
know you want to

Musky and RD, thanks, good thinking and all valid comment.

After posting I carried on rereading and searching old threads and realised what had been noticed by Col early this year when he posted in this, John O R's treatise on the adapter variation etc. here:
https://www.a7a10.net/forum/index.php?topic=9646.msg70166#msg70166

and by Simon (Mosin) in 'Touchy Feely" a decade earlier! here:
https://www.a7a10.net/forum/index.php?topic=2689.45

I think I now have the culprit identified; my sleeve bearing is wrong. It has only a 5.15mm wide shoulder and it should be 8mm or thereabouts. Interestingly it is also 5.3mm longer overall than the one John measured in his report. It also has a distinct undercut machined into it and I can't see any reason for this either.  This makes it different again to the ones Simon, John and Col have unearthed and potentially a 3rd different design part! (Unless it is a spurious part off something else but that seems unlikely) see attached photos.

This is I believe the root cause of my primary drive self destructing although I have found numerous other issues nothing has been conclusive until now 3.5mm misalignment is pretty significant, either with or without oil!

So now the plot thickens further, do I make it work with this or do I buy yet more replacement parts???
The second photo shows the assembled condition only just possible to get the split pin in.
Third photo shows the cush spring being compressed to being coil bound measures 24mm (free length 42.5mm)
Final photo shows the gap at 4.85mm between the sprocket and the cush drive sleeve when set at the 24mm (coil bound spring) distance. This indicates I can potentially space the sleeve bearing out the required 3.5mm (additional spacer between the existing one and the shoulder on the sleeve brg.) and remove this amount from the end of my sleeve brg (which is 5mm longer than it should be anyway) while still maintaining a gap of 1.35mm (4.85-3.5). i.e. on max cush deflection the spring will become coil bound 1.35mm before the cush drive sleeve is hard up against the nut. If I can do this then I will maintain the current nut position, have all threads engaged and be able to fit a split pin. I could even pinch an extra mm to make fitting the pin easier! I could even take the whole 5.3mm extra length off the sleeve and this would prevent the spring from ever binding up on the coils, maybe a good thing?

I think my remaining question now is how likely is it for the cush drive spring to bind up and will this damage the spring over time? Effectively I'm not changing the condition I had previously if I go this way but that doesn't mean it is correct!

[Swarfcut]

 Bagga   That drive sleeve looks like one from the earlier plunger motor. I have one to hand, the shoulder for the oilseal is 5mm wide, overall length is 56mm, and like yours also has the cutaway on the splines, which are 25.5mm long. OD where the duplex sprocket runs is 35.8mm Measure yours for comparison.

 By contrast a drive sleeve from a 1956 S/A bike is 51 mm long, oilseal shoulder 5mm wide and the splines run to the sprocket  position without the cutaway. Here again assuming my part is actually the correct one for my bike. Sprocket position  and spline are 35mm. OD.

 BSA changed the drive sleeve design early in the development of the S/A models. 67 1097 (1954/55), then 67 1134 for 1955. The early part has the "spacer" behind the oilseal shoulder incorporated. The later part uses the separate spacer. Later models from 1956/57 use 42 0069.

 Bear in mind possibly none of the parts you have for this cush assembly are strictly correct. P.O. Antics etc. Plunger and S/A Cush spring sleeves have marginally different splines as detailed  above and do not interchange.

 Swarfy.