Crankcase bearings

[Steve.S]

Aha! I think I have just had a Eureka moment. More of which later.
Since you're doing the 1930 engine, I believe you are using the incorrect roller bearing, which, of course, gives rise to your question about the possibility of the inner race moving towards the ball bearing, and the consequent need for an inner collar.
I think, if you search on the internet, you will see what N305W and NF305 look like.
Yours has only one flange on the inner race. Mine has 2 flanges. So can you now see that the inner race of the NF305 can't go anywhere, and there is no actual need for the inner spacer?
The Eureka moment I had, was that I discovered that for the 1929 engine, BSA transposed the bearings. i.e the roller bearing goes in first followed by the 2 collars and then the ball bearing. Hence the need for the inner collar on the 1929 engine, but not on the 1930 engine. Ta Da! I suppose you could phone the Factory for advice, but I tried the other day and there was no answer. (Sorry about that.)
So, if it was my engine, I would swap out the N305W for the NF305, and forget about making a collar. It's up to you, and I seem to be in a minority of one.
I couldn't find the Sloper piston on their website either, in fact, I couldn't make much sense of their website at all, but if you email them I'm sure they will put you right.
As I tried to suggest, checking the endfloat is really very simple. Just assemble the crankcases and push and pull the mainshaft in and out with your hand. To be accurate, it really needs to be done with everything dry, as oil will take up clearance. No need to put the cush drive on, or the 1/2 time pinion and nut. Simply push and pull on the drive side mainshaft. That is your endfloat.....if any. There is no thrust face to pull up against (as you say), because the cush drive is on a taper and not up against the bearing (as Mr. Chapmast says,) and I'm fairly sure the 1/2 time pinion butts up against a register on the mainshaft, and does not tighten against the bearing. But it's minus 10 degrees here at the moment, and I don't want to freeze my balls off going outside to check.  So, NO, NO, NO, the cush drive does not lock up against the bearing (again as Mr.C says.)
As you say the Sloper crankshaft can float around, and that is why BSA recommended zero endfloat (less 0.002"), as per Service Sheet 15. Personally, I wouldn't bother about the 0.002" nip.
It would be good to know if you have now got your head round it all.

[chapmast]

If you have a look at the engine drawings held in Roll 1 of the Technical drawings section of the library then you can see that BSA were all over the place on this.

Sheet 25-0019 shows a 1928 OHV engine with two ball races on each side of the fly wheels and a roller on the drive side towards the Cush drive. Both inner and outer spacers are fitted.

Sheet 25-0036 shows a 1931 OHV engine with with two roller bearings on each side of the flywheels and a ball race towards the Cush Drive. The outer lip of the rollers are away from the crank ( the only way they can be fitted) and the inner spacer is missing.

Sheet 25-0029 shows a 1930 SV engine with two ball races fitted on the drive and timing side. The roller bearing is fitted up against the crank on the Drive side. There is also a strange looking inner spacer.

Sheet 25-0028 shows a 1931 SV engine which you would think followed the same pattern as the 1931 OHV engine but instead follows that of the 1930 SV engine of the previous year.

As previously stated, my 1928 engine had 3 ball races fitted.

Essentially BSA kept changing their minds and seemed to try almost every possible permutation.

Service sheet 15 ( dated Sept 1930) is actually quite contradictory. It seems to imply that it refers to all models with some caveats with regard to 1927-1929 Sloper engines. It refers to fitting shims on the drive side between the ball race and flywheel engine shaft boss - from 1930 onwards none of the engines fitted a ball race against the flywheels.

I think that you have to assume that they are shimming up the roller bearing. On the later configuration I think that they are relying on the lip of the roller bearing so stop any drift. Although the ball race in this configuration can act as a thrust bearing I don't think that it is really behaving in the same way as on the later B/M series models.

[Steve.S]

Although Service Sheet 15 is dated September 1930, I believe it only refers to pre 1930 models, where of course, the ball bearing is against the flywheel, and it's talk of Glut Pins, which were deleted for 1930.
It's interesting to note that Bernal Osborne was also a little confused about the positioning of the bearings, in his article for Motor Cycling Magazine.
He mentions that a previous owner had swapped the position of the bearings, (i.e. from the 1929 arrangement to the 1930 arrangement.)
He goes on to say that the Artist has reproduced the BSA drawing, but fails to mention that the drawing is clearly of the 1929 engine, complete with inner spacing collar and Glut Pins.
It seems that BSA themselves got it wrong, because if you study the OHV engine line drawing on page 13 of the 1930 BSA Motor Cycle Instruction Book, you will see that the arrangement depicted is of the 1929 engine, with the ball bearing against the flywheel, and inner spacing collar. So it seems that the different departments of BSA were also confused.
In practice, I doubt if it would make any difference how you fitted what bearings, as long as you didn't use 3 roller bearings. If fitting a ball bearing against the crankshaft with an outer roller bearing, you would need the inner spacing collar.

[Oval57]

OK Gentlemen, I truly appreciate your input and research into this subject, I think I can safely say that my opening line on this thread "I'm Confused" now has some reasoning behind it!
I had studied the article from Mr Bernal Osborne, he is the one that started the confusion , I had Googled any information I could find on the subject and being new at pre 40s bikes, did not, but quickly found out like the BSA factory (RIP)that it depended on the day, the weather, and the person as to how they were going to build their engines!
So, as it appears to be no right way, I have assembled mine as follows starting from the drive side.
Ball Bearing, Spacer Outer, Inner spacer (machined up by me) Roller Bearing, jumping over the crank to another Roller Bearing... everything turns beautifully, no noticeable end float, and the rod is perfectly central.....
End float will change today though as I noticed the contacting face of the crank pin on the timing side is a bit ugly, so I am going to dress it up in the lathe and replace the removed thickness with a shim.
then hopefully by the week end I will have the bottom end all sealed and locked ready for the top end..
Cheers