Superblends VS Ball rollers

I am not advocating sloppy work, just making an observation that tolerances on old bikes are sufficiently wide to make the job easy, even for those without high tech measuring equipment. Crank shimming is a job that many competent do-it-yourself mechanics shy away from, because of perceived difficulty, I hope this discussion may persuade more of our fellow Norton enthusiasts that they can tackle this job as well, regardless of which method they opt to use. I concur that using a blow-torch or an oven is a recognises and effective way of removing and replacing outer races, however I would rather do it once than two or more times. The method of shims behind the inner race is "by the book" as per the Norton manual and uses original Norton parts (notwithstanding the fact that many of the exploded parts diagrams show the shims in the wrong place). I manage to get my end-float within factory spec every time with this method. The use of a special inner race makes the procedure quick and hassle free. I don't insist that anyone uses this method, but it's accurate and it works.

I'm not disagreeing with you at all.

Furthermore, measuring end float couldn't be easier, just a set of common feeler guages is enough to get the tolerance in range.

I don't own a dial indicator for run-out although I probably should. I just use a jig and a set index; anything "iffy' goes to my machinist, period.

If I had a shop twice the size of the one I have now, I'd have at least a lathe and boring rig, plus a chest full of the cool stuff that I just gawk at whenever I'm at the machine shop. I LOVE to watch those guys work.

It was always my belief that any bearing shims that are used are placed on the side of the bearing that will not come into contact with any rotating part. So the shims don't spin. Placing the shim in the case first behind the outer shell acomplishes this. But. it does take more trial and error. This is why it may be a good practice to measure the existing end float first before dissassembly, then measure the shims & bearings that are removed, the new bearings and with a little bit of math you can come to a fairly good result of what the shim requirement may be. Not an exact scientific process but what the hey, it may just be close enough the first time so you don't have to try and try again.

GrandPaul, what do you consider to be the major advantage of large OD shims behind the outer race ? (You mentioned it in capital letters).

It would seem best practice to me to have the outers seated as fully in their housings as possible. Bearing in mind that some of these cases need 25 thou or so of shimming, I would rather not have the outers sitting proud of their seats by 10 - 12 thou.

Neither outside of the inner or outer race would be a location where movement should be present but I have to say that I've seen more evidence of outers moving than inners. As Dave mentioned as well, the less often that they are removed, the better.

I have never found shims present when dismantling engines. It seems most people don't bother.

We should of course be shimming to achieve centrality of the con rods in the bores. This is relatively straightforward in a single cylinder engine but somewhat beyond me on the Commando. I have never been able to achieve sufficient accuracy of measurement to be confident of basing a shimming decision upon it.

I have found shims behind the inner race on some engines that I am quite sure were undisturbed from new - most bikes in Hong Kong did fairly limited mileage before the importer went out of business or they were taken out of service.
In conversation with Les Emery on this subject, he stated that given the amount of side play on pistons and conrods, the crankshaft end float was only really an issue with keeping the primary chain in line and that with a belt drive it doesn't matter at all. I am personally of the opinion that the factory spec is a good reference to aim for and should achieve more complete race and roller interface. Hopefully this discussion demystifies the whole procedure for those who haven't attempted it before. The more things we can do ourselves the better chance we have of keeping our bikes on the road.

To me, the logic of Joda's explanation concurs with Fred at Old Britts, and I looked to him when starting out with Nortons a few years back.

At .017 of shims on Chip's bike, the outer race did not stand proud of the crankcase casting.

I guess I've been lucky.

The three Commando engines I've worked on never needed shims. And should I ever encounter a Norton crank that does need shims, I would worry about shimming it in the wrong direction.

Having watched Mick Hemmings video on heavy twins restoration I hadn't even considered the idea that I might need to check for end-float on the crank. :shock: Again in the video, Mick simply heated the cases and dropped the new races into position.

Being this is essentially the first time I'll have attempted to rebuilt an engine (Yeah I know I kinda like jumping in at the deep end), once the new races are on the crank/cases, are nipped up... How can you check to see if the race halves are fully engaged?

Great thread this, shame I missed it earlier.

H

I've got a couple of thousand miles on a pair of FAG X-Life spherical roller bearings. They're self-centering, have a much higher static and rotating load capacity than ball bearings and tolerate more shaft deflection than "superblends". Think they were 21306-E1's. They are a captured race bearing, so I had to ream the inner race for a sliding fit on the crank to get proper endplay. Neat stuff. If they blow up, I'll let folks know.

For my next engine I'm going to try Saint Gobain full compliment Cerbec ceramic ball bearings. Ball bearings do spin up better than rollers (lower moment of inertia), but they can brinnel and gall under higher loads. Unfortunately they don't make a 6306 type, so it'll have to be a custom order.

Dave M, I like your approach with using an inner race from a used SuperBlend, but my understanding of SuperBlend bearings was that a large part of the extra cost was for hand fitting of the component parts. What research did you do to confirm uniformity of the inner races?

RS

For what it's worth, Kenny Dreer was using Koyo 9-ball bearings for mains on the last of the VR880s.

Anyone have any input on failures with these?

Roadscholar, I haven't done any research at all, it simply seems to work for me and I have built about 10 engines with this method. I do stress that I only use my special inner-race to get the end-float within spec and then the new inner race is substituted and reunited with it's original outer, so I'm not worried about mismatched components. I don't think any bearing manufacturing tolerances are relevant here given the fairly liberal end-float parameters that we have to work within and I have so far not found any measurable discrepancy between the special race and the correct race in this regard.
DaveM

Dave M,
Thanks for the wisdom, I kinda figured it to be that way, precision and all that, I just wanted some confirmation, and I have a set of cast off superblends so I'm on my way to make such a tool.

Dave Comeau told me that shimming the end play behind the outer race was asking for trouble in that the race is moved further away from contact with "solid ground" and may end up spinning, which won't do your case any good. I was very lucky in that my end play was right at .017, just shy of mid point for a MkIII bottom end according to the WSM; I didn't need any shims. Mr. Comeau also suggest that the drag set up by the oil pump is sufficient to haul the crank over to the right side of the end play at engine speeds that make a difference; I'll leave any controvesey to the masses; I was just plain happy to have my end play come in without needing any shims.

RS