Layshaft BALL Bearing Upgrade - "in situ" - Techniques And Tips Help?

Kip said:

Let's shift gears a bit (sorry for the pun) and discuss the best way to remove the old layshaft bearing from the case...IN SITU. Would a little heat and a blind hole bearing extractor make the most sense? I've seen some with a slide hammer and some with a two (or three) legged threaded "puller". Is there room in the box for that type, or is it better to just warm the box and slide hammer the old bearing out? Is there a viable alternative to either type of extractor?

Click to expand...

It might depend of if it is the original bearing and if the case has cracked between the bearing bores as they can do and or if some expert has been in there before and for whatever reason used Loctite.

Here is pic of the clusters, the shift forks engaged in the selector plate is a whole other work around. (Why it is easy to do it on the bench)
You still need to get down to the layshaft and even with it only in the case there will still be two gears between the shaft shoulder (along with the output shaft sleeve gear most likely) and the layshaft bearing but sensible heat to the area of the bearing bore on the case will normally allow it to be puled free. (No hammers, pullers or otherwise)
Remember 80 % of the demise of the classic British motorcycle was home mechanic's so we all have the obligation to show mechanical empathy.



In that pic and remember you are dealing with the shift forks on both shafts also, you most likely still have the two layshaft gears next to the bearing along with sleeve gear. If the bearing comes free from the case the layshaft comes out, if the bearing stays in the case only the shaft comes free then a blind puller comes into play.

If you are digging into it that far, the primary is easy to remove (and check) and then for the sake of three cradle bolts , the top cradle bolt can be loosened and the other two removed. The engine can then be rotated a little so the gearbox can be removed.

If it is an 850 the R/H lower section of the cut out can be checked for cracking, it has less metal at that location than the 750 cradle.

In the Hemmings DVD, he inverts the removed gb and heats it til spittle flash boils on the alloy. A quick couple of drops from a few inches off the work bench and the bearing just falls out.
I did it in situ by heating thoroughly with propane torch. Checking for flash boiling etc. Took 10-15 min with torch moving all around inside of gb and some outside close to bearing area. Came out easy with just a little screw driver levering on inside diameter.

Kip said:

Let's shift gears a bit (sorry for the pun) and discuss the best way to remove the old layshaft bearing from the case...IN SITU.

Click to expand...

If the case is heated sufficiently then the bearing (if the original ball) should come out with the layshaft as shown in the Norton Companion picture, therefore, if the layshaft does not easily pull out with the case 'cold' then don't attempt to remove it.

...And then, more likely than not, have to pull/lever the bearing off the layshaft although the bearing extractor shown shouldn't be necessary.

Kip said:

Let's shift gears a bit (sorry for the pun) and discuss the best way to remove the old layshaft bearing from the case...IN SITU. Would a little heat and a blind hole bearing extractor make the most sense? I've seen some with a slide hammer and some with a two (or three) legged threaded "puller". Is there room in the box for that type, or is it better to just warm the box and slide hammer the old bearing out? Is there a viable alternative to either type of extractor?

Click to expand...

Slide hammer type works.

My (distant) recollection of changing out a superblend for a TB layshaft bearing with the gearbox in situ: With a heat gun applied to the back of the gearbox shell the bearing pretty much fell out on its own. No cracks in the gearbox shell either.
I was unaware of the Hemmings advice for a sliding fit between the layshaft and bearing, and I always thought that the main advantage of the TB bearing was the fact that it provided a positive location for the layshaft, so I was a bit baffled when I saw that piece of advice.
Mick sold me the TB bearing and never mentioned it.

B+Bogus said:

I was unaware of the Hemmings advice for a sliding fit between the layshaft and bearing, and I always thought that the main advantage of the TB bearing was the fact that it provided a positive location for the layshaft, so I was a bit baffled when I saw that piece of advice.

Click to expand...

Mick and John Hudson both recommended this, however, there seems to be a logical reason for not doing so.

comnoz said:

I always thought it was kind of strange that people looked at the ball bearing as a way to control layshaft endplay and then sanded the shaft down so it would move in the inner race. I would suggest a tight fit because a fit loose enough to move will only get looser . Jim

Click to expand...

L.A.B. said:

Mick and John Hudson both recommended this, however, there seems to be a logical reason for not doing so.

Another layshaft/Bearing question

Hi Guys, I took the plunge into the gearbox for bike No2 the other day, seeing that I'm still waiting for engine bits for engine No1 to return. I'm led to believe that the layshaft should be 'sliding fit' into the case bearing, so having more of less survived a major dose of gearbox purchase...

www.accessnorton.com

Click to expand...

...and this is why we heat the bearing

My failed layshaft bearing left its outer in the shell, balls and cage in the sump and the inner on the layshaft. I took all the innards out other than the mainshaft, heated the shell and from the primary side place a long piece of hardwood on the shell outer next to the layshaft bearing and hammered the hardwood. The inner was knocked out to be replaced by a roller bearing.

I used a rawlbolt with a piece of studding instead of the bolt to pull the outer of a roller bearing from my gearbox
After heating with a heat gun
I've used this method many times to pull bearings from blind holes including needle roller bearings

IMHO interference fit bearings in alloy cases must be put in a heated case and removed from a heated case. Pulling/beating a bearing in/to of a case does at least a little damage to the alloy and in the case of the gearbox shell can easily crack the case between the bearing if not already cracked by having a layshaft ball bearing and a worn bushing at the other end or just the stupidity of designing the case with so little metal between them.

baz said:

I used a rawlbolt with a piece of studding instead of the bolt to pull the outer of a roller bearing from my gearbox
After heating with a heat gun
I've used this method many times to pull bearings from blind holes including needle roller bearings

Click to expand...

That is an awesome use for a rawlbolt, thanks for sharing that

Given the "weak link" issue with the so-called "Portuguese" bearing (and apparently other steel-caged layshaft bearings) - has there been much discussion about the root cause? IMO, there's evidence that the problem may well originate in mainshaft flex. Peter William's outrigger bearing would appear to confirm the mass of an essentially unbalanced clutch hanging off the end and the primary and drive chains whipping along with constantly varying loads... ain't good for the stuff downstream. It's been mentioned the box was designed for 30hp singles and hit the limits behind 60hp Commandos...the torque twisting or bending the mainshaft along its length. Roll-off throttle transitions reverse those loads rather abruptly as well. Even steady throttle riding, especially with maladjusted chains, possibly set up a harmonic vibe in the box that could shatter the bearing cage. Needle bearings in the sleeve gear, might also be better than bushings since the bushings are known to move around under these conditions.

To me, an improved ball bearing is enough of a "fix" for a sanely ridden, 50-year-old, stock machine. On the other hand, a resto-modded hopped-up bike, ridden in anger, with a totally unforgiving belt drive primary - better have a roller bearing on the layshaft and probably should have an outrigger bearing in the primary.

Your thoughts?

Kip said:

Given the "weak link" issue with the so-called "Portuguese" bearing (and apparently other steel-caged layshaft bearings) - has there been much discussion about the root cause? IMO, there's evidence that the problem may well originate in mainshaft flex. Peter William's outrigger bearing would appear to confirm the mass of an essentially unbalanced clutch hanging off the end and the primary and drive chains whipping along with constantly varying loads... ain't good for the stuff downstream. It's been mentioned the box was designed for 30hp singles and hit the limits behind 60hp Commandos...the torque twisting or bending the mainshaft along its length. Roll-off throttle transitions reverse those loads rather abruptly as well. Even steady throttle riding, especially with maladjusted chains, possibly set up a harmonic vibe in the box that could shatter the bearing cage. Needle bearings in the sleeve gear, might also be better than bushings since the bushings are known to move around under these conditions.

To me, an improved ball bearing is enough of a "fix" for a sanely ridden, 50-year-old, stock machine. On the other hand, a resto-modded hopped-up bike, ridden in anger, with a totally unforgiving belt drive primary - better have a roller bearing on the layshaft and probably should have an outrigger bearing in the primary.

Your thoughts?

Click to expand...

Define "sanely ridden"

Kip said:

Given the "weak link" issue with the so-called "Portuguese" bearing (and apparently other steel-caged layshaft bearings) - has there been much discussion about the root cause? IMO, there's evidence that the problem may well originate in mainshaft flex. Peter William's outrigger bearing would appear to confirm the mass of an essentially unbalanced clutch hanging off the end and the primary and drive chains whipping along with constantly varying loads... ain't good for the stuff downstream. It's been mentioned the box was designed for 30hp singles and hit the limits behind 60hp Commandos...the torque twisting or bending the mainshaft along its length. Roll-off throttle transitions reverse those loads rather abruptly as well. Even steady throttle riding, especially with maladjusted chains, possibly set up a harmonic vibe in the box that could shatter the bearing cage. Needle bearings in the sleeve gear, might also be better than bushings since the bushings are known to move around under these conditions.

To me, an improved ball bearing is enough of a "fix" for a sanely ridden, 50-year-old, stock machine. On the other hand, a resto-modded hopped-up bike, ridden in anger, with a totally unforgiving belt drive primary - better have a roller bearing on the layshaft and probably should have an outrigger bearing in the primary.

Your thoughts?

Click to expand...

Yes flex is an issue.

Hence the benefits of a lighter belt drive and clutch hanging on the end of that long, thin, shaft.

Its not all about hot rodding though, lugging along at too low rpms (easy to do on an 850) is arguably the most harmful.

I’m intrigued by your ‘unforgiving belt drive’ comment, in what way is a belt drive more unforgiving that a chain ?

Power transmission is slightly less efficient because of higher friction in the belt/pulley interface. Belts need to run cool, offer zero shock absorption, and have no option for tension adjustment. They transmit momentary loads to bearings and shafts that may not be up to increases in stress and can yank a long flexible mainshaft off-center forwards under hard acceleration because the drive chain has its slack taken up abruptly. Deceleration/braking does the same thing in reverse. (IE: the tension in the top run of the drive chain instantly goes slack and the bottom run gets tense.) Meanwhile, that same lack of shock absorption is affecting an already flexible crankshaft as well. It's not always easy on clutches either. If you want to find weak links in the drive train, belt primary drives can help you do that.

They have plenty of advantages but are not perfect.

BTW - you are absolutely right about 'lugging along'.

Kip said:

Power transmission is slightly less efficient because of higher friction in the belt/pulley interface. Belts need to run cool, offer zero shock absorption, and have no option for tension adjustment. They transmit momentary loads to bearings and shafts that may not be up to increases in stress and can yank a long flexible mainshaft off-center forwards under hard acceleration because the drive chain has its slack taken up abruptly. Deceleration/braking does the same thing in reverse. (IE: the tension in the top run of the drive chain instantly goes slack and the bottom run gets tense.) Meanwhile, that same lack of shock absorption is affecting an already flexible crankshaft as well. It's not always easy on clutches either. If you want to find weak links in the drive train, belt primary drives can help you do that.

They have plenty of advantages but are not perfect.

BTW - you are absolutely right about 'lugging along'.

Click to expand...

With respect, I wasn’t looking for (another) generic belt vs chain debate, I am simply interested in why you think a belt is more unforgiving than a chain and has zero shock absorption compared to a chain. Can you explain that please, I do not consider chains as having much shock absorption personally.

Was talking to Andy Molnar via email a while back and he said none of his race bikes use cush rear hubs with belt driven clutches. His race bikes get ridden a lot harder than any street bike ever does. The lack of shock absorption with a belt is over exaggerated on the interweb. That said yes indeed it is a lot easier to adjust the chain tension on a chain driven clutch for just about anybody. A chain is more forgiving with imperfect slack adjustment. A belt not so much.

I have the roller bearing on the layshaft. I don't listen to advice though, so there is that.

Fast Eddie said:

With respect, I wasn’t looking for (another) generic belt vs chain debate, I am simply interested in why you think a belt is more unforgiving than a chain and has zero shock absorption compared to a chain. Can you explain that please, I do not consider chains as having much shock absorption personally.

Click to expand...

I think it was commoz that once pointed out that a chain drive is lubricated by oil
With a film of oil on each pin and roller of the chain it provides a cushion compared to a belt drive

baz said:

I think it was commoz that once pointed out that a chain drive is lubricated by oil
With a film of oil on each pin and roller of the chain it provides a cushion compared to a belt drive

Click to expand...

I can see that, but I‘d say it’s rather minor.

The argument for rubber belts is that the teeth can ‘give’. Which I’d also suggest is minor.

So it’s the poly belts that probably have the least give. In theory.

I think it’s all academic personally though. I’m running belts on 5 of my 6 old Brits. 3 rubber and 2 poly.