Layshaft bearing failure frequency

I have a '73 850 Interstate, 10,000 miles on odometer, Hayward belt drive, and am using Redline Heavy Shockproof gear lube. I am putting bike away for winter, but shifting has been great throughout the previous riding season. No issues at all. I am quite sure the laysaft bearing has not been replaced by previous owners. I was just wondering if there is less frequency in layshaft bearing failures if primary drive is belt-driven, there being less shaft stress due to light weight belt, as opposed to the heavy 3-row primary chain.

I'd be considering the fact that the layshaft bearing is a known issue, and if not addressed can result in catastrophic failure.
The stresses transmitted via the clutch are managed by the mainshaft bearings, so the layshaft torsional stresses - to my mind at least, are pretty much independant of the clutch.
I am a subscriber to the philosophy that the belt drive will reduce the shock loads transmitted, but my understanding is that the layshaft bearing failure mode is due to either a fundamentally not-fit-for-purpose bearing - which needs to be changed, or due to flexing of the layshaft - which will destroy the bearing anyway.
Whether layshaft flex is a result of mainshaft flex or not is one to ponder - I don't know the answer to that one :roll:

My bike has done a similar mileage to yours...I decided the risk wasn't worth taking.

850mkiia-rebuild-underway-t5086.html

On the other hand, you may get away with it....

Mine went at about 12K miles, risk is too high not to have it done. Its not a failure you want on the road, I was lucky it happened in a city max speed 40 mph, I lost all gears bar 4th and got the 10 miles back home in 4th, not stopping and with loads of clutch slipping. In the debris in the gearbox was a number of balls from the bearing with some of them in half.

I ran for luck with mine for all too long. Made me very nervous at speed - the thought of locking up the back wheel in highway traffic.

Changed in recently, pulled out a ball bearing with, yes, "Portugal" stamped on it, and replaced with a roller bearing. Took the opportunity to replace seals, bushings, gaskets, and springs while I was in there. About $300 in parts, from Old Britts. You could probably buy them a bit cheaper, and certainly you're unlikely to actually NEED all new parts; I just figured, while I'm in there...

Worth doing. Yours might never go out on you. But if it does, you and/or the bike might not survive the experience.

Best - BrianK

I'd like to get a better idea of the layshaft bearing problem.

I've looked at a lot of threads using "layshaft bearing" as the search words.
Since I'm seeing comments that indicate the problem was all the way up through the MRK III era is it true that from 1969-74 the Commando's
had this problem?

By the time 74 came around the bikes had been made for 4-5 years and the trans had been used in earlier models for a lot longer than that. During this time were higher mileage bikes having failures yet up into 74 the factory was still installing lousy bearings? Or was it a problem with the later bikes due to a bearing change or using 24 tooth transmission sprockets and low revving torque or maybe the higher horsepower of the later models was just too much for the old design.

I have a low mileage 72 and the trans runs fine and although I know to be absolutely sure of avoiding the problem I should spend the time and money to go through the trans I'd like to better understand the odds of this being a problem in mine. This is a weekend putt around locally bike and I'd be surprised if I put more than 2,000 miles per year on it.

Thanks
Bob

It is certainly believed that the larger the gearbox sprocket, the greater the chance of failure and that it was compounded in the later years by the use of cheaper (sub-standard) bearings.

The 850 Mk111 is the most notorious. They were frequently fitted with a 22t sprocket and I think all had the dodgy bearings.

The problem after all this time is that it would be imprudent advice to say '"don't worry about it" The best advice to anyone worried enough to ask the question is to change the bearing. Generally there are warning signs before failure but the consequences of sudden catastrophic failure are so great that the only sensible option is to replace sooner rather than later.

To be honest, I think that pretty well any forty year old gearbox ought to be looked at before being used hard and we're lucky that on the Commando this can be done without dismantling the entire power unit.

If I think back thirty-odd years, this can be done quite satisfactorily by someone with no prior experience using the factory manual (and Mum's oven ! ) The reassurance of knowing that everything is in order is well worth it.

When I bought my first Norton I didn't even know something like a layshaft bearing existed .
3 years and some 80 000 km later , when the kickstart began hitting my leg , I didn't recognise the symtoms .Guess what happened next ..
Altough , by that time I had learned about blowing headgaskets without flame rings , disintegrating slit skirt pistons , burnt valves , loose valve guides , slipping clutches , the Prince of Darkness , soft kickstart pawls , broken gearbox teeth , cracked oil tanks , parts vibrating themselfes to liberty , massive oil leaks and many more of the finer aspects of British engineering .
So , yes , I would say : don't worry too much about it !
( mm ,on second toughts : change the bearing )

BrianK said:

I ran for luck with mine for all too long. Made me very nervous at speed - the thought of locking up the back wheel in highway traffic.

Changed in recently, pulled out a ball bearing with, yes, "Portugal" stamped on it, and replaced with a roller bearing. Took the opportunity to replace seals, bushings, gaskets, and springs while I was in there. About $300 in parts, from Old Britts. You could probably buy them a bit cheaper, and certainly you're unlikely to actually NEED all new parts; I just figured, while I'm in there...

Worth doing. Yours might never go out on you. But if it does, you and/or the bike might not survive the experience.

Best - BrianK

Click to expand...

Brian:

Since you mention Portugal stamped on the bearing it sounds like you think they are the ones to worry about. Did they change to Portugese bearings in a certain year?

From Captain Nortons
Peter Aslan writes:
>The 'layshaft bearing in the case', is the layshaft bearing fitted to a blind hole at the
>back of the gearbox case, the opposite end of the layshaft to the end that fits in the
>inner cover. this means you will have to remove and strip the box to change it. When
>you do, wait till the wifes out and put the gearbox case in the oven to heat it up
>before drifting a new bearing in.

This is not true in the vast majority of cases. I have replaced numerous layshaft bearings and have *never* had to pull the gearbox out of the bike - a rather big pain in the ass as you must not only disassemble the Primary drive but also get into that rather inaccesable rear engine mount. All thats necessary is to heat the gearbox casing around the bearing (say witha propane torch) and stick a cold damp rag into the bearing itself then after a few minutes use a puller and out it comes... simple.

If I gotta go there I'd like to keep it as simple as possible. Any comments about this method of replacing bearing while leaving case in place?


Thanks
Bob

My understanding is the Portugese bearings used in this application were notorious. The one I took this out of was a 74 gearbox. Took one out of my 73 that did not look quite as bad (had more of a "cage" on the balls, and did not say "Portugal"). But I was happy to have it out of there.

BTW, in my 73 gearbox (the 74 is a spare I bought, and thank heaven, because....) two sets of gears had significant spalling.

I ended up using the original 73 box to keep numbers matching on my bike, and mixed and matched the best parts from the two boxes and all new bearings/bushings/seals/gaskets/springs to assemble one fully rebuilt box (got a lotta spares too, although as I say, some are fairly worn...). Knock wood, so good so far.

I guess it could be done in situ. Would be pretty hard I would imagine. And I would think you'd still have to remove the primary in order to get a torch on the back side of the gearbox. I would (and did) bit the bullet and pull the gearbox, allowing you do to the work on a bench.

I contacted a knowledgeable fellow concerning the layshaft question and he replied with some interesting information I thought I would share.
Bob


Many years ago I was thumbing through a few old N.O.C. Roadholder magazines when I noticed a letter from someone telling the tale of their gearbox failure whilst touring Spain or somewhere. It did not take long to find another such letter. Thinking 'the dead dont wrote letters' I phoned a couple of ex Norton/NVT friends asking if they were aware of Commandos as having a gearbox problem. Both said they were and both suggested 3 reasons for the problem.
1. The original Commando motor was the Atrlas moI contacted a fellow familar with the problems with Norton transmissions and asked him what he could tell me about them.
I've copied his reply below. tor. The Atlas if we assume it produced a max torque of 48 ft lb with its 2 to 1 primary ratio could, assuming incorrectly no power losses in the chain, apply 48 x2 = 96 ft lb of torque at the clutch / gearbox input. The Commando employed a 26-57 primary ratio couldapply 48 x 57/26 = 105 ft lb to the clutch / gearbox input....Straws break camels backs.
2 The Atlas clutch weighed statically a lot less than the Commando clutch but more importantly most of the weight increase was ar the periphery where it increased rotating weight a lot more thus putting yet more strain on the poor old gearbox especially as the lump called 'the clutch' was probably unbalanced. Personally I refer to Commando clutches as 'Unbalanced gearbox breaking flywheels'.
3 The lack of a transmission shock absorber.

Any of the 3 could well be responsible for the problem but bung all three together with young riders trying to pretend they were racers....as we all did at one time........
Then I remembered just how many Commando owners were regularly coming into Motor Cycle Shop in London to buy the new stronger gearbox shells they would have made or ask if they had a good second hand std shell they could buy. Then I thought about the speed at which the Aluminium alloy Commando baskets would be bought along with theallloy clutch centres by owners clearly aware that gearbox mounted clutches should possess the LIGHTEST rotating weight reasonably possible rather than the heaviest according to such people as Phil Irving, the staff of The Motor Cycle etc. It is my belief that Norton won outright the inter factory(BSA Triumph Norton) prize for the heaviest rotating weight 'clutch' ever fitted to a British motor cycle with that reduculous bronze plated lump.
I have also investigated our so called shock absorbers and have yet to find ANY evidence of any of them being designed let alone tested to determine whether they work correctly. In fact I have asked numerous people such as Bert Hopwood, Doug Hele and others over the decades for their thoughts on the following statement in the hope that they could tell me I was wrong 'In my opinion you spent more on feeding the drawing office cat than you ever did on clutch and transmission shock absorber development'. To date NO ONE has corrected me or offered any details of actual development taking place. The Triumph shock absorber was removed from the crank and placed in the clutch just after WW2. It was, according to a Gentleman working directly under Edward Turner, a copy of one that had been around in the 1920s and that he and others in the Design Office felt it worked less well than the original crank mounted one. In his design note book for the 7R and Porqupine AJS motors written in the early 1950s Mr Jack Williams wrote that the gearbox clutch shock absorber did not have enough mechanical movement within it to work correctly and that he had requested a car type crank mounted one running within an oil bath chain case so the chain was better lubricated!! He also states that at high chain speed the drip fed primary chain system was about 90% efficient and lower still if lubrication was poor!!

An olde Wakefield oil book of mine shows a sectioned drawing of the old Burman 4 speed box and states that it was designed to be safe up to 50 hp at 2500 rpm at the input =105 ft lb of torque. Now thats what in theory an early Commando could shove unto the box but when I design a clutch for such an application I use a x2 safety factor and the clutch would have a torque capacity of 210 ft lb to ensure it did not slip when fully engaged and 240 ft lb for an 820 motor.The bos was well past its sell by date and in my opinion was not suitable for the application. But the Commando (or Atlas Mk3 as shown on the drawings) was only a 2 tear MAX stop gap model to replace the non selling vibrating Atlas to get some money into AMC whilst a new model was designed tested and put into production. I remember seeing the AMC 800 cc OHC twin motor in a road bike a few times at Plumstead and according to a friend serving his timeat AMC it was under performing and had serious engine oiling problems. The next time I saw it was in a Suffolk old motor cycle garage that had closed in the mid 60s.... the owner removed a pile of sacksand carpets telling me I would never of seen one of them before... I disappointed him in telling him I had heard its only use was to get the fish and chip suppers from down Plumstead High Street for the workforce. I think its in the National Motor Cycle Museum these days.

However if your box hasnt thrown you down the road yet then I suspect you ride the bike without screwing the nuts off of it so the box should not be a problem BUT NEVER buy non genuine Norton spares for it especiallly gears and shafts.Some idiots make and flog Norton gears leaving stress raisers at the roots of the engaging dogs so the break off knackering the whole box ....for example. A friend in Birmingham has one on a shelf in his workshop as an example to show customers..

That bearing..... I suspect the Portugese bearing failed because they were cheapo bearings which met the minimum required performance but not that required when seriously over loaded in our gearboxes. I remember the days when Hoffmans lost the contract to supply the drive end dynamo bearings for Lucas to a Jap company. Hoffman obtained some Jap bearings and were most upset to find on test that they outlasted their product by several times but even more upset when they found the Jap bearings were costing Lucas LESS than Hoffmans could produce them for!!! Mind you visiting the Hoffman works was like going back in time to early Victorian times wheras the Jap factory was probably very modern with the latest state of the ark production machinery.
In my youth the bearing used to replace that layshaft 6203 bearing was a higher load capacity ball race bearing with one extra ball in it. Its designation was M203. Ordered some a few yearsago but received std 7 ball 6203 bearings. Spoke to the NTN bearing designer who said they no longer produced the higher capacity bearing as the metalurgy and machining had improved so much the std 6203 bearing was the equivelent of their old M206 bearing.
As for using the N / NU / NF / NJ / 203 roller bearing it certainly gives an increase in load carrying capacity. My old RHP book lists the load as 6203 ...1660 lb and the roller 203 ....2400 lb. However the internal clearance within the ball race beaing isapprox half that found within the roller bearing although whether 0.001 inch or less differewnce matters I know not. Personally I have used the roller bearing many times BUT I buy, if its available, the NJ203 version which has two lips retaining the rollers in the outer raceway so it is easier to remove at a later date..the inner raceway on the shaft only having one lip. One only drills ones gearbox shell so one can knock out the bearing once in this life and then get the holes welded up.. In fact looking at a couple of new Fag roller bearings I bought a few years back Ithey are marked NJ203E, the E denoting extra high capacity.

Me thinks that for a Commando I would go for the roller version.For an ES2 or Dommy......whatever I had around.

Remember the days when the works race Commandos were not completing a lap of the Isle of Man before suffering gearbox failures?? Riders would pray the failure occured when going in a straight line rather than between the rock faces and dry stone wall bendy sections AND within walking distance of a pub!! Peter Williams found that the cause was gearbox mainshaft deflection that was causing the gears to not mesh correctly. He overcame the problem by putting a main shaft support bearing behind the clutch, correctly mounted so it did NOT move fore and aft like those that people flog that are mounted on the top and bottom fixing bolts,
I make such things for unit BSA and Triumph but they mount in the existing oil seal holder hole in the rear chain case BUTare actually correctly aligned by fitting dowels through the plate as the hole/flange BSA etc put in was approx 0.005 inch plus non central to the mainshaft which is why the oil sealsarealways worn on one side than the other. Have thought about doing it on a commando with a belt system but there is a problem or two . It would have to be fixed and doweled to the cradle.....all belts come with a length tolerance of up to=+/- 0.3mm so if you set up the belt correctly and its one of say 879.7 mm length and replace it with one of say 880.3 mm it will be slack and correct belt tension is fairly tight so the whole support would have to be repositioned which is beyond most people.these days now we no longer produce time served Engineers or 5 year time served fitters. The same applies to the Triumphs and BSAs but I put 10 small guide holes of 4mm in around the edge and supply dowels of 4mm, 5mm, 6mm so the friends can if required reposition by simply redrilling the existing holes out bigger and use the bigger dowels. With a belt expected to last at least 30,000 miles I cannot see many Brit bike owners changing them very often..... Unfortunately I have better things to do with my life such as build a few copies of the works Domirace frame and build the short stroke 500 twin motor with 90 dgree crank so the handle bars are not 8 inches in diameter along the Mountain Mile etc at 8000 plus as they were with the parallel twin version of a few years ago.
Of course having just had my heart ripped out by a heathen Scottish gent who did a replumb job..for the second time in my life,,,,I I think I might have to get a move on........

Trust this helps . Have fun with your STOP GAP Norton!!!!
PS If ever you are thinking to buy a belt system please contact me BEFORE ordering it so I can give you a few facts rather than the bull shit some system makers put out.

Was this Fellow BDM? in England?

Hortons Norton said:

Was this Fellow BDM? in England?

Click to expand...

Yes it's him.
Very helpful on a number of items I've discussed with him.

The bearing can be replaced with the primary side intact and the shell in postion, the mainshaft stays in place too. I heated the shell with a gas burner and then from the opposite side underneath the bike placed a long piece of wood on the layshaft housing, a belt with a hammer and what was left of the bearing popped out.

How long should a layshaft bearing last?

pelican said:

How long should a layshaft bearing last?

Click to expand...

My layshaft bearing (74 model, VIN 318XXX) went out at almost exactly 10K miles which seems about average. Luckily for me it went out at 30 mph and not 70. Watch for the dancing kickstart lever. That might give you some warning. Sure enough, I had the famous Portuguese bearing. I thought I remember reading somewhere that it was mainly used on 74 models but I am not sure. Probably worth replacing with a roller regardless.

Opinions seem about split on whether it's better to replace this bearing (and possibly the others at the same time) with the gearbox in or out of the bike. I prefer the latter because I can heat the gearbox case uniformly on my gas grill to remove and install the bearings. The old ones just drop out and the new ones goes right back in without any drifting at all.

With the orginal ball bearing the mode of failure is more concerning than the mileage, the cage breaks up leaving the balls free to bunch up on one side and then single ones drop out on the far side. A bearing application should be good for 50K miles and the failyre should be an increase of clearances not complete collapse.

My bike has 20k miles. Don't know if it's ever been changed. Now you guys have me freaking out a little. :shock:

pelican said:

My bike has 20k miles. Don't know if it's ever been changed. Now you guys have me freaking out a little. :shock:

Click to expand...

Didn't you need an excuse to powdercoat the frame?

HAhAhahHA no way I'm gonna go back that far