Layshaft bearing

Status
Not open for further replies.

MichaelB

"Sons of Arthritus"
VIP MEMBER
Joined
Jan 14, 2004
Messages
2,447
Country flag
I have owned several Nortons over the years and more than once on a resale the topic of the Layshaft bearing has come up. I have heard many horror stories of this bearing failing and causing rear wheel lock up.
The ONLY time this topic has ever come up is when someone is interested in purchasing a bike from me.

I have maybe been under the false impression that a properly maintained gear box should be OK as long as it shifts and the kicker doesn't drop with revs.

I have been asked so many times now I am wondering how dangerous is this stock bearing? My only trans failure was from total abuse when I yanked the mainshaft out of the case with one too many burnouts.

The Tech Digest talks about the poor 'Portugese' Mark III bearing and lists a heavy duty replacement but doesn't list this as a must do repair for the others.
Andover Norton still sells the stock bearing. Norvil sells the stock along with a Superblend and another 'Racing' bearing.

So whats the concessus? Is this a must do repair? Or with proper maintenace and correct oil is this not a problem to be concerned with?
Are there other signs of imminent failure?

Up for slow, down for go.
 
The type of gearbox fitted to Commandos was first used from 1956 and was never originally designed to handle the increased power output produced by the Commando engine, and as a result the original layshaft 6203 ball bearing (part 04-0100) has certainly been known to fail, especially if a 22 tooth gearbox sprocket is used (the later gearbox casings also being strengthened).
UK bikes (certainly later ones) seem to have been fitted with higher gearing as standard so this may be more of a problem that affected UK bikes rather than US ones as I believe the standard gearbox sprocket for the US market Commandos was usually around 19 or 20T?

The NOC Commando Service notes say this:

John Hudson:
<<I believe layshaft ball bearing failures are more frequent when:-

1. The 22T gearbox sprocket is used as on 850s and
2. Every failure I have seen has been a Portuguese SKF bearing>>

(note John is referring to the BALL bearing not the later roller type)

Further on it says:
(text probably by Tim Stevens)
<<The layshaft ball bearing has a self destruct mode in the order of 10,000 miles mainly due to higher torque through [the] gearbox, caused primarily by the 22T gearbox sprocket giving too high a ratio.....>>

From 1975 the layshaft bearing was changed to the NJ203C3 roller bearing (part 06-7710 I think this bearing is sometimes referred to as a 'superblend' although I'm not sure it actually is?) and this bearing I think then became the standard replacement part, although the earlier ball bearing is considered adequate when used in a Dominator or single gearbox, so any original (pre '75) layshaft bearing may have been replaced with a later NJ203 type bearing by now (hopefully)?.

Under racing conditions it has been suggested that the NJ203 roller can hold the layshaft too rigidly causing the layshaft to snap! And as a result of this some racers prefer to use a high specification ball bearing that allows the layshaft to flex slightly.
This bearing is the 06-9366 (FAG6203TB) type and some experts (certainly Mick Hemmings) recommend it both for road and race use.
 
this discussion is perfect timing for me as I'm just about to do a tranny rebuild and have new main and layshaft bearings (plus various springs, seals, etc....) from british cycle supply.

If you don' t have a handy-dandy bearing puller for the layshaft bearing, will it pop out with some heat to the case? I was figuring that I could get it hot with a heat gun and 'finesse' it out, then reinsatll the new bearing while still hot.

any problem with this plan?

thanks

Karl Hoyt
 
The normal method of bearing removal is to heat up the gearbox casing, the gasket face of the case should then be brought down sharply on a wooden block, this should shock the bearings out if the casing is hot enough.
It's sometimes advisable to remove the studs from the gasket face before doing this.

Have the new bearings handy so they can be dropped in before the casing cools.
 
Basketcase,

I recommend heating the gearbox shell in the oven to about 350 degrees F. Then as LAB pointed out, slam the gasket face down hard on a wooden workbench. The bearing will plop right out.

Next, I would let the shell cool enough to enable you to clean out all the spooge that was hidden behind that old bearing. Finally, heat the shell back up to 350 degrees F and install the new bearing; it should drop right in.

I think it's easier to get the correct temperature with an oven versus a gas torch, or similar device. Also, the oven will heat the part more evenly.

Jason
 
MichaelB said:
The ONLY time this topic has ever come up is when someone is interested in purchasing a bike from me.

these prospective buyers can be a real PITA huh? ;)

On my '73 850, the OEM bearing failed at about 20K miles. The cage broke up and then the balls all became free agents. That gave the layshaft a rather large amount of radial play ;) Gearbox locked up at 60 mph without warning. I hear sometimes you'll get some early warning in the form of the kickstart lever moving around "by itself" as you accelerate in the lower gears.

My '71 ebay special had a ball bearing which was still in excellent condition when I opened the gearbox up. Someone had been in before though so I suspect it had been replaced. I read somewhere that you can get a special "deep groove" ball bearing to put in there which holds up better than an ordinary one. I put a roller bearing in my gearbox.

Was the OEM bearing of unusually poor quality? Were some years worse than others? Don't know but would like to!

Debby
 
Basketcase,

I recommend heating the gearbox shell in the oven to about 350 degrees F. Then as LAB pointed out, slam the gasket face down hard on a wooden workbench. The bearing will plop right out.

Next, I would let the shell cool enough to enable you to clean out all the spooge that was hidden behind that old bearing. Finally, heat the shell back up to 350 degrees F and install the new bearing; it should drop right in.

I think it's easier to get the correct temperature with an oven versus a gas torch, or similar device. Also, the oven will heat the part more evenly.

Jason
 
Was the OEM bearing of unusually poor quality?

Debby,
As the late John Hudson said it was the Portuguese bearing that was likely to fail recommending the Hoffman (later RHP) bearing but I think it was more to do with the fact that a standard ball isn't up to the job, (especially if a 22T gearbox sprocket is used) that's probably why the change over to the roller that is now the standard replacement part?

Mick Hemmings still recommends the FAG racing ball bearing which, according to him is 'the answer'. I bought one of these bearings from Mick some while ago and I haven't fitted it-yet (amusingly this FAG bearing is made in Portugal!).

Mick Hemmings = broken link removed
 
Last edited by a moderator:
thanks to all for your advice, and let me apologize if I hijacked the thread.

Now to figure out a way to use the oven without my wife catching on ! maybe I'll cook up something with onions and garlic immediately after to kill off any oil smell that may have been produced

:oops:
Karl Hoyt
 
No worrries, you didn't highjack anything. We are all here for the same reason, information.

I guess the summary is , could be a problem, when it goes it's messy and dangerous. Be aware!!

I too am going to cook up some trans casserole.
Ah Honey, that's not oil drippings, it's spillage from the streudel.
 
I too am going to cook up some trans casserole.
Ah Honey, that's not oil drippings, it's spillage from the streudel.


LOL....................I was thinking linguica and onions (portuguese sausage, a bit odiferous) might hide the smell a bit... and since my tranny case is already cleaned out, the actual 'meal ' will be pretty good for a hungry weekend warrior mechanic!
 
Danger Will Robinson! Danger! Danger!!!

When heating parts in in your primary food oven, always turn available vent fans on high and open the closest outside doors and windows (whilst closing portals to the remainder of the domicile).

Male... female... household chef... or not... the smell of cooked industrial lubricants lingers - and does household harmony great harm!!! Really. Honest. I swear by my dead dad.

Spend the few dollars (quid, loonies, franks, euros, and etc...) to have your parts hot tanked and bead blasted before subjecting them to the family oven. Might even be worth a second blast of carb cleaner (nasty, carcinogenic MEK - avoid if you value your liver).....

(If you're a hermit/hermitess, feel free to ignore this advice. No one will care untill they come to read your deranged ramblings. (Err, umm , sorry - "magnificent manifesto".....).
 
Could someone explain why a larger sprocket causes bearing failure please?
I have a 24 tooth gearbox and a 48 rear ( long story , don't ask) so it has standard gearing.

However, is it the fact that the sprocket has a larger diameter and loads the bearing somehow, or is it simply that higher gearing gives greater loading?
 
The NOC Commando Service Notes say this:

(John Hudson)
"It seems to me very unfortunate that they [Norton] did not change Commando gear ratios by using alternative engine sprockets, because with a larger engine sprocket and retaining say, a 19T on the gearbox (as was standard on all Norton Nortons from Model 50 to Atlas and Manx) the speed of the box would be increased without such a heavy load on the bearings and tooth loading on the pinions themselves."

I would guess that John's reference to "Norton Nortons" being pre-Norton Villiers and NVT products?


I think that it's the increase of the secondary gearing that is the cause rather than the actual size of the gearbox sprocket itself, changing the gearbox sprocket being the normal way to alter the gearing as (I believe?) there wasn't any alternative rear sprocket sizes available?
 
The taller gearing (larger front sprocket) places more stress on the gearbox components, especially when launching the bike off the line.

This is similar to riding a bicycle in too high a gear at low speeds; it's punishing to the knees.

Jason
 
L.A.B. said:
The NOC Commando Service Notes say this:

(John Hudson)
"It seems to me very unfortunate that they [Norton] did not change Commando gear ratios by using alternative engine sprockets, because with a larger engine sprocket and retaining say, a 19T on the gearbox (as was standard on all Norton Nortons from Model 50 to Atlas and Manx) the speed of the box would be increased without such a heavy load on the bearings and tooth loading on the pinions themselves."


The late John Hudson certainly knew more about Nortons than I ever will but it seems to me by changing the engine sprocket the load would then be transferred to the crank / bearing, would it not? This is already a weak point.

Everything affects everything.
 
You're right L.A.B, the rear sprocket is cast into the hub.
I had mine modified....the teeth machined off and a ring welded on behind which is drilled and tapped so I can bolt on different sprockets. It's for racing so I can change gearing without dismantling the primary drive.
 
My layshaft bearing went while on a sat shopping trip with the missus on the back, the kickstart went to the deck and gearchanges went stiff until only 4th was left. Left the missus to catch the bus while I clutch slipped through Birmingham city center back home in 4th. When I stripped the gearbox in the bottom was several half ball bearings, the bearing was in several pieces, put a roller bearing replacement in and its never been open since. The steel clutch plates were turned blue by the clutch slipping but are still working fine.

22 teeth 850 UK bike
 
but it seems to me by changing the engine sprocket the load would then be transferred to the crank / bearing, would it not? This is already a weak point.
Possibly?
But if there's a problem with the crank or (superblend) main bearings caused by raising of the primary drive ratio then I've not heard of it, also belt kits tend to gear up the primary drive ratio, generally to around 2:1.
 
belt kits tend to gear up the primary drive ratio, generally to around 2:1.[/quote]

Good point, However I believe Hudson is talking about the development of the Commando whereas superblends, thicker case weren't evolved yet.
If one is adding belt drive with modified gearing, one would hope that the crank / cases / bearings would be addressed also.

Norton had extensive experience with the Manx and maybe someone realized the Dominator / Commando twin wouldn't take the stress of tall 'race' gearing.

The Norton 'Heavy Twin' was never intended as a race motor as the Manx was. So with that come the limitations.

The more I learn about gearing the more I realize how competent the original ratios, gear spread work with the original design. It is when we get outside the original design that problems creep in.

Is a 21/22 tooth sprocket outside the original design, I think when talking about the original 68 Fastback it is. These taller ratios came later.

Has anyone out there had a failure with 19/20 tooth sprocket?

I have read, maybe from John Hudson, that one of the threories on Combat failures was the taller gearing. Supposedly the Proddy racer
was the forerunner to the Combat with the same roller bearing but without the disastrous results.
Supposedly the 21 tooth combined with the high compression at street speeds hammered the ball bearing to death. I don't know, I just read it somewhere.
Anyone care to open a thread on the 'Illfated" Combat?
 
Status
Not open for further replies.
Back
Top