Britbike forum Forums British Motorcycle specific Brands Bsa bulletin board A65 End float with needle roller/ball conversion

Does anyone out there have experience of A65 crankshaft end float on an engine with the timing side roller/ball conversion? The NKIB roller/ball bearing is supposed to hold the crankshaft in place such that end float is limited to the ball race clearance and so shimming is not neccessay. I have such an engine that has done around 30k miles since the modification and whilst changing the triplex primary drive chain recently, I noticed very exessive end float. I also saw the excessive movement when checking the iginition timing with a strobe pointed at the rotor. The still image produced by the strobe can be seen moving in and out as the engine rpm is changed. Having had a look behind the timing side covers, the balls of the NKIB bearing look like they are brand new. Nothing obviously wrong and I took the opportunity to re-torque the worm drive nut and lock nut. Any ideas what is going on?

I forgot to mention....it's a late motor that has the roller bearing on the drive side as opposed to the earlier ball type...

JCC welcome to the board. Many members have fantastic knowledge and experience. I'm pretty new too, and have 2 A65s. Being a backyard mechanic for decades does lead me to venture a guess as to what you got: the inner race or the outer race is loose and sliding on the crank or the housing.

Thanks for the input. The inner race of the NKIB bearing is a split arrangement and the two halves are held together by the the oil pump worm drive nut which holds down the timing gear which, in turn, holds down the inner race. From the position of things I can see that things seem to be o.k. with the holding together of the two halves of the inner bearing race because there is a gap between the timing gear teeth and the crankshaft end. The gear will not go on any further because it is prevented from doing so by the two halves of the inner race being contacted together.

Is the outer moving? Why I use a plain heavy series needle roller with no side load and a bearing in a plate bolted outside the alternator. Watch the T/side and move the crank. The outer should be held by the balls to the inner. Something has to be moving. It may move more when hot, but try pushing it, then levering it back.

There is only so much crush you can fit the bearing with. Checkout what is going on and what moves.

+1



………


One thing I have noticed when I have seen some of these end feed conversions (one of my A65s had the NKIB Bearing but I changed it to an RNA bearing and out rigger bearing outboard the alternator) is that often the shim-spacer is not fitted between the timing gear and the oil pump drive worm. What then happens is the hex of the worm is now too close to the oil pump. Check you have that before doing a torque check on the timing side.

Mark, Allan.... Thanks for your inputs. The washer/shim/spacer is fitted underneath the worm drive as intended. The end float is not easily detected by manually pulling or pushing on the end of the crank. I noticed it when removing the alternator coils. They were a bit stuck in place so removed them with a three legged puller that was pushing on the end of the crank. When turning the screw, what happened first was that the crank moved in before the coil assembly moved out. As suggested by Mark, movement is easier with a hot engine, which I noticed when timimg it with a strobe light. The crank was clearly jumping sideways by quite a lot. We are not talking a few thousanths of an inch. More like a mm or therabouts hence it was visisble under the staic images produced by the strobe. My thinking is that the outer race on the timing side is moving in the crankcase and, the more I think about this, the more I think that there is a design flaw as follows. I assume the bearing is an interference fit that is inserted from the inner side of the main case and that it abuts against a lip on the oil pump side of the case. The lip stops it moving to the RHS but there is nothing but the friction of the interefernece fit preventing it from moving, within the clearance limit, to the LHS. From the drawings of the bearing assembly I see that it has a very thin outer race, which suggests to me that the interference fit of the outer race will not be very tight in terms of friction..... hence the movement....One of the claimed benefits of this roller/ball bearing combination is that end float is completely eliminated and no need for any shims, as the balls of the bearing assembly hold the crank in place laterally. That is only true if the interference fit of the outer race is tight enough as well, which I suspect can't always be the case with a bearing that has such a thin walled outer race. I'm begining to think that for such a conversion, a plain roller with a shim system would be better because, in my case, the balls are doing the job inside the bearing but the outer assembly of the bearing isn't staying in place so the balls of the bearing have become redundant. Anybody else seen this problem? I've never heard it reported previously.

Two ways to fix it without opening up the cases, use the outrigger bearing as Mark suggested or use a roller bearing thrust collar which requires narrowing the sprocket. It makes the main bearing like the 5 speed high gear with guide lips on both sides of the rollers. Assuming you have a RMS-11L or MRJA 1-1/8 with the single lip on the inner race.

I understand the outrigger bearing idea but I'm not sure I follow the last comment about using a roller bearing thrust collar. The crank movement to the right is limited by the timing side outer race hitting the machined lip on the timing side case. The crank movement to the left is limited by the roller guide lip on the drive side RM.11L bearing. To take up the end float slack in system on the drive side (without opening the main cases) would presumably involve moving the drive side outer race further inwards to the right so that the lip is closer to the rollers and the outer race would no longer bottom out on the casing lip. How does a roller bearing thrust collar with a narrower sprocket do this?

When using the nk1b type bearing it's still not a bad idea to put a spacer
on the crank at the drive side to prevent the outer of the t/s bearing moving
in the case. Having used NK1A, NK1B, RNA, and NK1X types over the years
i still like to set the crank up so it's unable to move more than a few thou from
side-side. Relying on a needle race fit is not ideal IMHO.

Nick

It's looking like serious disassembly is required as the bottom end is not in good shape after 30k miles. I have to say it has not turned out to be the "fit-and-forget" system that I thought it was as the mileage it has done is less than what was on the original, as-bulit, factory motor, which was getting a little tired but still running well enough for general, everyday, use.

Hi

I'd suggest that it is actually the outer race which is moving in the casing. If the Brg is assemble correctly, then the oil pump screw holds it all together on the crankshaft, therefore the outer race has to move in the casing to allow any end float. If this has been happening it should show up on the oil pump and drive gear scroll not meshing correctly.
As Nick said adding a spacer will prevent this. I have used these types of Brg's in a number of engines and never had them walk about.

My tuppence worth

John

I think the same. That it's the outer race is moving about when it shouldn't but in my engine it seems that the intereference fit is not sufficiently tight enough to prevent movement, which doesn't really surprise me with the thin walled outer race of the NKIB bearing and the vibrations that it must experience. It looks like a spacer and shims are needed, as long as the moving about hasn't opened things up too much, but then the balls of the bearing are redundant. Perhaps some Loctite bearing retainer might add something positive to the situation?

Hi

While you have not posted a photo of the set-up, I assume you have a lip on the casing which prevents the Brg moving out board? Just asking as I have seen one engine where there was no retaining lip on the casing.
I would heat up the casing and remove the outer race, add some loctite as you say and replace the outer race. Rebuild the crank into the cases and use the Brg as intended with the ball bro's. Then as Nick said in an earlier post, space the crank out using a spacer/shims as necessary. This should then allow you to use the engine.
It's a pity this site doesn't allow a simpler method of posting a photo, to help members explain what their problem is.

Hope this of help to you.

John

As stated above, if the outer race is not restrained the forces on the crank WILL cause it to move.

With loctite be aware you can add quite a bit of crush when the case is hot and bearing cold. When you get it right it will go together. I put one in and it was a tight fit (without Loctite), I ran it for 10 years or more and the D/s was worn a bit when I pulled it down and the needle bearing was exactly the same, a little tight. I replaced it with a C3 which has been in there for years. It has more internal clearance.

What that outrigger plate looks like.



The bearing I use NKIS 30. But the id and od will be different. What John suggests is probably the go.

Might be worth adding, the RNA 4907 is an ideal direct replacement for the NKiB bearing used by SRM/Devimead, all dimentions are same apart from the width, the needle roller part of the bearing is identical but it doesn't have the extra width needed for locking the endfloat.

another solution is to shim the crank as the DS but use the NKIA bearing inner, the endfloat is limited to one side. ie it only has one shoulder for the ball race to work against. you would have your .003" or what ever of crank float as set by the shims but you wouldn't need to worry about needing a thrust face on the timing side.... Or use a DS ball bearing. I've had a ball bearing on my DS for about 7000 miles, the engine hasn't seen an easy life from my riding but the ball bearing is surviving fine, I do also have the outrigger bearing due to my crank not being shimmed and without it the bearing will walk.

From memory the current bearing you have is NKIB 5907 (the other would be NKIA 5907)

I erred, you would have to open the cases and change the drive side bearing to use the thrust collar type bearing.
The thrust collar goes on the drive side bearing. You have the NF type which has the rollers contained in the inner race and the outer race has a single lip to keep the rollers off the case bore.
The NUP as a separate inner race lip that captures the rollers on both sides, taking care of the crank float.
Another arrangement is the NH type which has a separate thrust collar. This would require the distance piece between the bearing and sprocket be narrowed to accommodate it.

Needle bearings do not like angular misalignment. The flexing of the crank makes the shaft rotate in roughly a cone shape so the rollers next to the crank web are loaded and the rollers 180 degrees around are loaded near the pump end. This causes the outer race to walk. Without something to retain the race in the case to will likely move. The triple's drive and timing side rollers are retained by retaining rings in the cases. They do not have nearly the angular misalignment of the twin with only two bearings holding the crank.

Many thanks for all the info. Much appreciated. Going off-line for a few days....

No, loctite suffers when it used with a steel bearing in an alloy case, each heat cycle expands the alloy more than the steel and the loctite is pulled apart and ends up in the sump, tried it twice 30 years ago on the same engine with same results each time. Third time used a new alloy case with the proper interference fit and no loctite and no problems.

see this thread https://www.britbike.com/forums/ubbthreads.php/topics/477561/new-engine#Post477561

I did two needle rollerconversions,
one, in 1984 and this bottom end lasted to 2012
one, a couple of years ago

in both conversions I used the standard axial crankshaft setup, just like ducati and a lot of other enginebuilders do.
the standard solution with a pressure plate and shims is adequate for normal use. and will stand a lot of abuse

regards Anne

The crank on my GM speedway engine (in a home built frame), my Jawa speedway engine (in a Rickman mk3) and my JAP grasstrack engine (in an OK Supreme) all have similar crank bearings.
The bearings are caged needle rollers directly on hardened mainshafts, the outer races are "top hat" shaped providing a thrust face either side of the crank. End float is 0.4mm to 0.6mm (0.016" to 0.024")

Where I am going with this is If a plain needle roller bearing is used, a hardened thrust washer could be fitted between the case and the crank to locate the crank centrally and shims to set the end float between the crank and drive side main bearing

Hi All,
Just to add, On A10's with the SRM conversion an individual to that engine collar is supplied to fit between the ball bearing outside split race
and the timing gear that keeps the two halves of the split race together once all is tightened up

John

+1 to not using Loctite to "hold" the outer bearing in the crankcase.
The differential expansion between the steel bearing and the alloy crankcase produces very high forces that just tears the Loctite apart.
Going back more years than I care to remember-- it is no coincidence that the Manx Norton motors used a mechanical clamp to hold the outer bearing in the crankcase.
Food for thought?

Dave's suggestion with different bearings types is nice BUT they
are only available in metric sizes and involve either machining the
bearing outer or the case.
I still think that just putting a spacer on the drive side not the shim cup
but a parted off spacer allowing about 5 thou or so is a good way, it
provides a little 'belt and braces' type fix.
The flex on the crank will always put stress on the needle rollers if
the fit is too tight and using a wide needle bearing makes it worse.
A C3 type is preferable in this application IMHO.

Hi

+1 on that fix.

John

With respect to the timing side needle roller only bearing (i.e. no ball bearings) is the timing side thrust washer required? I presume so.

JJC

Yes, you still have to keep the crank from running against the case (which is what the timing side ball/needle or drive side ball would do).

https://swf-ignitionsystems.nl/wp-content/uploads/2021/08/P1050079.jpg

https://swf-ignitionsystems.nl/wp-content/uploads/2021/08/P1050081.jpg

https://swf-ignitionsystems.nl/wp-content/uploads/2021/08/BSA-Axiaal-lager-20170108-01641-1.jpg

works fine

JCC with a straight needle roller, a bearing outside the alternator locates it. That's what I do. It needs setting up central when putting it together while you can still see either the T/side or rods in the bores. Spacers position the crank. Other people may use a thrust washer but the bearing plate works for me as it also limits crank flex and chain misalignment, where a bit of flexing and the wide triplex can knock side plates off. So it's extra control to add to one behind the clutch.

You can see the G/box mainshaft flex without its support by feeling the primary and putting your foot on the rear chain and pushing. Big Laverdas use the same chain, the chain is strong but can come undone on a hot BSA pulling the sprocket out of line. Possibly some cheaper ones are more fragile or not well riveted. Regina seem good and originals.

Just looked, neat setup Lemans. No side load on the bearing.

That's a very good point Mark, we found when racing that a duplex primary
chain lasted about twice as long as a triplex. Due to the flex on the crank,
it would pull the chain out of shape in a strange way. BUT that was on an
outfit with a 900 engine and a lunatic riding it, i don't think it would be so
much of a problem on the road.

So is the duplex just the same size. Your bigger front sprocket changes that length a bit, but can you just run them on the std sprockets on the inner runs? I can see as you go wider the more critical alignment would be. You would think it would be 33% stronger but not like it is. I've seen them come apart and punch through the case.

I used an alloy basket for years, it's ok but breathing through that chaincase it was putting fine alloy and anodizing in the engine oil. It would be fine in the Firebird not breathing through that case. I couldn't tell a difference with the alloy basket.

I wonder if an outside support on a Triumph crank would preserve them, it wouldn't be flexing as much. I read the Triumph guys machined alternator rotors so they didn't chew up stators, a bearing could stop that, and maybe the drive is what snaps them.

I'm going to get technical. If you put a length of wood across two bricks with overhang one end and push in the middle it flexes down as the end of the overhang goes up, but if that overhang is held in line it reduces the flex in the middle as well. And pull on that unsupported overhang could add flex.

You could modify a BSA 68-729 stator mount to go inside a Triumph (! that would be a real mongrel) to stiffen the stator support. There is about 28mm from the stator holes in the Triumph case to the cover joint and 22mm from the joint to the 68-729 stator seat. With the stator in the primary cover there is about 15mm from the back of the stator to the cover joint so the stator mount has to be shortened.
I would make a custom mount, since the outrigger bearing mount should locate on the rim of the stator laminations rather than the through bolts which have clearance.
Aren't we all lunatics for riding these old crates?

Mark, the chain is/was the same pitch etc, just duplex 82 links rather than triplex.
I did have an outrigger on my outfit at one stage but when i machined the case
for the 33 tooth sprocket. i couldn't fit it back on. This is an early set of cases so
there are just 3 stud holes into which we've made up an outrigger setup for the
32t sprocket. We could run a triplex chain on here if we replaced the sprocket
but getting one may be awkward.

I like your technical analysis of the crank but you must remember that stiffening
things up too much makes them break like carrots. Shafts must flex around the
bearing areas if they are just made from EN16 etc. Maybe EN40B nitriding steel
can be totally rigid but that's not the case with this old gear.

Thanks, indeed no side load. and the solution is identical to the BSA-layout. this is the expensive one, used a brass-plate as a bearing material. Recently I modified a crankshaft to the same needle-roller and used an aluminium plate. The side load isn't a load at all and most japanese camshafts run directly on the cylinderhead-aluminium.

lemans, the brass plate takes the thrust to the timing side, how did you take the thrust to the drive side, a washer outside the needle bearing?

no it is just the std BSA setup. so the drive-side rollerbearing has the shimms for axial-clearance and the timing side thrust washer (brass or alu plate) positions the crankshaft in the middle of the cylinder bores.

as you can see from the second picture I've left a 3mm ridge (in the timingside casing) to prevent the needleroller from moving out. so the crankshaft is free to move in axial direction, limited by plate and shims. just like BSA did it with the timing-side bush

and the needleroller cage can't move outwards (ridge) and can't move inwards, thrustwasher and crankshaft
but there's no side load on the needles/cage, so the bearing has no reason to move. and a light interference does help.

I was reading about Nigel Rollason's A65 outfit. He said they were getting over 90hp. Running a 3 bearing crank and a 3 bearing G/box main shaft. They put a bearing where the alternator would go which also controlled end float, As with my set up except mine is outside the alternator. I'm rebuilding one at the moment, a 90degree 744 that the bottom end has not been apart since about 1998 when it was built. It was running fine with good oil pressure but the alloy rods though looking in good condition are getting old and without a remote filter we were wondering about sludge in the crank.

Before I had a plain spacer between bearing and threads on the crank and the threads under it were a bit worn. So I made this spacer to use a Norton spigot nut under the bearing. So it slides into the bearing so the hex section acts as a washer between it and the alternator rotor. Because a loose washer in there is very hard to assemble. Then the spigot nut threads under it.



Nigel said with these extra bearings primary chains gave no problem at all. As they would stay aligned. As late as 1986 the Hogg brothers on an A65 were taking wins in modern racing, causing headaches for Yamaha TZ700-750 outfits and 900 Z1s.

The outfits really tested drive lines and they showed just how potent A65s could be.

Hi Mark,I took a slightly different approach. I made the spacer with a step pocket. The pocket is deep enough so the end of the nut does not bottom on it and the spacer locates on the stud. The bearing is clamped between the hut flange and spacer.