Britbike forum Forums British motorcycles in general Bsa gold star forum Loose timing side main bearing

I have a worn GS timing side main bearing housing. The TS bearing is an MRJA7/8 7/8" x 2 1/4" x 11/16", with a lip

I have read that Phil Pearson had used a different bearing & either a sleeve, or grinding to achieve fit. This was AFAIK done to account for the unavailability of good quality TS bearings.
Does anyone have experience of this & know whether it also solves, or could solve the wear problem?
There are 7/8" x 2" x 9/16" bearings available. On the face of it, sleeving the case housing & using one of these *might* work.
One of the issues that arises is I have only seen these as un-lipped bearings. Are lipped versions available?
It would *seem* quite possible to run an un-lipped one, as the whole crank is pulled up to the drive side, with safe clearances for the lipped bearing on that side as per good practice. The other thing that comes to mind is that if using an un-lipped outer, that could move inwards towards the flywheel face if not sufficiently well clamped (interference fitted) in the sleeved housing.
The crank should never be allowed to float so that it moves over to the timing side. That's not to say that it doesn't happen if the primary lock-nut comes loose but to some extent, the rod prevents it moving over too far & driving the TS bearing into the timing chest!

Anyway, back to sleeving the TS case & suitable bearing. Any pointers?

Thanks

For what it's worth, had Chaterlea25 not pointed out the existence of a metric roller bearing that could be modified for the purpose, I very likely would have installed another ball bearing with a Speedi-Sleeve to deal with the drive-side bearing of my Ariel that had come adrift on the Cannonball. I had purchased another ball bearing and the necessary Speedi-Sleeve to deal with the worn case, but the reason I decided on the roller bearing instead was concern that crankshaft flex might have been responsible for worrying the original ball bearing out of the case. The type of roller bearing suggested by Chaterlea25 allows for a fair amount of angular misalignment whereas the ball bearing doesn't allow for any.

Anyway, I haven't repaired a worn case using a Speedi-Sleeve and determined there were no long-term problems, so take my advice with a grain of salt. But, knowing what I know now (which might not be all that I need to know...) I would use something like a Speedi-Sleeve if I needed to deal with a worn timing-side case.

Another issue you raised was a lipped-vs. a non-lipped roller bearing. As you rightly point out, the crankshaft will be locked to the drive side so there will be no axial load on the timing-side bearing. However, without a lip the ends of the rollers will be free to make contact with the pinion.



Without trying to find a pinion in the garage to see what its inner face looks like, wear of the roller ends due to a non-machined surface (if it is non-machined) would be of potential concern. As for 9/16" vs. 11/16" width, what are the respective load ratings?

Thanks for your thoughts MM
I've just come out of the loft where I replaced a cranky 3 port valve in the heating system, the boiler also having failed completely this week, just in time for a promised cold spell. Figured I should get it done before the boilerman arrived & while it wasn't too cold up there!

So back to bikes; Your query about potential for timing side roller to contact the main pinion might be valid, although I don't think the actual rollers can but the centre could? I have never measured the clearance between the bearing & the back of the pinion / oil pump drive scroll. I guess I can make a trial assembly & attempt to measure this.
Assuming there are only un-lipped bearings in the 2" size (& load ratings were suitable) I wonder if a hardened washer would act as the lip, behind the bearing? The 2" narrower than std. so there is room for one.
The housing centre opening in the case is I think, only 1" in dia. I don't think the bearing inner can get past this but it could move & wear the case...
I did wonder about a spherical roller bearing as they allow for some misalignment - I assume this is what Chaterlea25 meant (I cant find any thread with his comments on this subject in it) .I would be interested to know exactly what the bearing he suggested is.
I would be reluctant to machine the housing out by very much. I know sleeves exist but have not looked up specs etc. thickness, suitability etc. (let alone chafing their way out under load as you mention) At present the bearing can more or less just fall out.
I have a Pearson crank on which the run-out is very small so hopefully the TS shaft is not flailing around in the cases like an ancient BSA crank might!

If you have the stomach for it, the discussion that resulted in me modifying and installing a metric bearing in my Ariel starts here and continues for the following two weeks.

Making a two-week story short, the drive-side bearing now in my Ariel is a FAG 21305-E1-XL-TVPB spherical roller bearing with ID 25 mm (honed to 0.9978" for a snug fit on the nominal 1" shaft), OD 62 mm, thickness 17 mm, and with a fiber reinforced polyamide cage. It can accommodate angular misalignment up to 1.5° for loads up to 1169 lbs. and has an upper speed limit that limits my Ariel to a maximum of 13,000 rpm.

MM, Ha! I did get rather immersed in the in & outs of your Ariel problem/s!
Quite a lot of educated guesswork going on prior to stripping the motor down to find the real cause! Some funny asides in there. Having seen your attention to detail re. measuring, I do wonder how I managed to keep my Goldie racer running well for 10 years with only the most rudimentary of engineering tools & know-how!

Anyway thanks for the lead. I did read up a little on speedi-sleeves, which I was previously only vaguely aware of. If I understand correctly, they really seem designed to solve a different problem - that of repairing damaged surfaces that oil / fluid seals are intended to run against. I note they have particular surface qualities (e.g. oil retaining) to assist the movement of seals. However, I can see that such a sleeve could be close fitted to the outer bearing ring as discussed in the Ariel thread & interference fitted in the case. The case bored (if absolutely necessary) to accommodate a fractional oversize.
I do wonder about the oil holding surface & the tendency for main bearings to spin but TBH, despite that qualm & assuming I have understood the sleeve's capabilities, I could see this as being a perfectly acceptable solution.
But I like to explore all the options.
I will look further into the specs of the 2" bearings & the spherical bearing possibility. Funnily enough a friend just mentioned Vintage Bearings for just the reason that Chaterlea25 said, as they seem capable of supplying specials or modifying bearings, which sounds interesting.
Ultimately it might have to be a case of "needs must" as although in an ideal world, my preference would be to keep the original bearing size if possible, the bike has been off the road for far too long & I'd really just like to ride the damn thing again!

An oil retaining surface will also retain locktite bearing retainer

You don't "need" precision measurements to keep a bike running. That is, until you need precision measurements to keep it running longer, or to make it run again. See below.

I suspect information about the surface was written by the marketing department. However, if had doubts about this, one of them would go under the metallurgical microscope to examine the surface at high magnification, as well as the surface roughness tester to compare with that of the outer surface of a bearing. If those two precision instruments detected no difference, I would know that there was no difference.

As you research your options, I'd immediately cross Locktite off your list because it simply isn't "designed" to do with what you need done. Instead, I'd suggest looking into metal-filled epoxies, some of which have compressive strengths nearly that of Al. The purpose of the epoxy wouldn't be to grip the bearing with its adhesive qualities before it set, but to replace the missing ~0.003" of Al.

The possible options that come to mind to investigate, in no particular order, seem to be:

-- Plate the necessary ~0.0015" on the outer surface of a standard bearing to increase its OD by ~0.003".
-- Machine an oversize metric bearing to have the necessary ID and ~0.003" initerference on the OD.
-- High-compressive-strength metal-filled epoxy.
-- Speedi-Sleeve.

There is a difference, but unfortunately my surface roughness tester was badly damaged when changing the drive belt on the mill. The finish is a little rougher than a bearing outer race, but I don’t see any need for concern. It was a very qualified machinist (who has built some successful LSR engines) that originally told me about Speedy Sleeves. Personally I wouldn’t have any worries about using it in your situation. Plating would be preferred, but if the bearing bore shows any signs of wear, the sleeve gives you the ability to clean it up or even relocate it if the line bore isn’t perfect.

Ps… with care and skill, one could polish the finish, but don’t think I’d bother.


1E8B7921-94C5-4461-97D8-D158903BFA6C by First Last, on Flickr

HI Flowboy,
Well worth asking Vintage Bearings about an oversize option,, a couple or 4 years ago I was able to get a +0.005 OD MC22 bearing
from them.. It was for the outer drive side main bearing on a 1937 Rudge
The outer bearing housing on the Rudge is quite thin and the bearing is held in by a threaded ring, so sleeving the case is not the way to go
it cost about £90 at the time.
I also bought a modified drive side bearing from them to fit a DBD crank into a small bearing crankcase :-)

John

All pertinent stuff, thanks to you all.
I have used various "loctite-ish" fluids over the years. Recently in an "emergency" repair on a rear wheel bearing (alu' hub) which fell out again when I next removed the wheel (a week later). I know I had rushed the curing time. On repeat & given some warmth & more time it "seems" to have succeeded, although I know it also will need a proper repair this winter.
The bearing retainer stuff has been ok on occasion elsewhere but I agree that in this circumstance it would be a waste of time. I haven't used metal filled epoxy in bearing fits previously. Although it is tempting, it does feel slightly like a bodge that I will still have to come back to, no doubt sooner rather than later. (I appreciate your sensitivity MM to my natural bodging tendencies! :laugh)
Of course someone will now chime in & say they JB Welded a main bearing in 20 years ago & it is still going strong!

The Speedi-Sleeve does seem to have a fair bit going for it, as some of you seem to have used them in exactly this way, although having now communed with Vintage bearings they said it was not what they would use on the bearing OD/housing, saying they are specifically for shafts & seals.
I take your points on the sleeve finish being adequate for holding bearing retainer if needed. Of course ultimately we hope the interference fit will be doing the work, rather than any "glue"! Polishing anything is something I very rarely indulge in!

Vintage Brgs do say that whilst they *could* supply a special oversize outer ring, it might be a 10 week wait, with a wild guesstimate cost of £150. No doubt plus post & VAT. Plus the donor bearing? So that could be a ballpark £200 depending on quality of bearing chosen. Plus the time & cost for my local precision engineer to open up the case sufficiently to match. So I would be well into next spring...
It felt like they were trying to warn me off, but probably just being frank. It's not an instant off the shelf option & future replacements would have the same problem.

I also looked at an NSK bearing catalogue online (mainly as it was convenient to navigate)
LRJA7/8" which is of course a smaller OD at 2" so a 1/8" sleeve would be fitted to the case, does have a lip on the outer & is available elsewhere as C3. Also the LRJ7/8" specs look quite close to MRJA7/8",
Load ratings being C; 25,100 + C/or 21,100 with rpms to 17500 for the MRxxx
& for the LRxxx C; 24,600 + C/or 20,800 with rpms to 19500
which on the face of it seems more than acceptable given the limitations I'm working with.
I note that NSK do not specifically list an LRJA7/8" but take it that other manuf. bearings of this type will be similar; Obv. TBC.

The other option which has been mentioned is the spherical roller bearing
The NSK load ratings for a 25 x 52 x 15 are way higher than MRxxx, at 37.500 & 37,000 although rpm is much lower at 9000. Obviously a sleeve to the TS shaft would be required but might be quite easily managed (Speedi-Sleeve again?)
Is the lower rpm limit too low for a 7000 rpm plus missed gears Goldie?
Also these brgs require (are recommended) to have a 2.5 - 5mm wide x ? deep oil groove machined in the housing (sleeve in my case) plus oil feed drillings to it, as supply to the centre of the bearing is somewhat compromised by nature of the design. As the outer is already narrower than MRJA, then that makes even less support for the brg interference fit - effectively 2 x 5mm strips.

Other than that, the advantage of the LRJA or Spherical brg. paths are they are off-the-shelf, repeatable replacements.

I very much appreciate all your inputs & apologise for my overly long posts - I'm self isolating at the moment so it's easy to spend a lot of time on the 'net!
Whether we will all arrive at a consensus on the best course of action I doubt but the thread is interesting to me at least

BTW, I forgot to mention plating. I have yet to speak to my engineer, who can take an accurate measurement re actual existing "clearance". Then I will know whether plating will work or that I am forced down the other paths.

I'm not pushing Speedi-Sleeves (or Timken's Redi-Sleeves) for this, but I would take that comment from someone at Vintage Bearings with a grain of salt.

There are two good reasons for them to have told you that: 1) they don't want to be blamed when someone whose skills they can't evaluate over the telephone screws up the installation, or 2) they don't have the technical expertise to realize a 0.010" steel band with an interference fit on a bearing, and in turn an interference fit in the case, will have the same holding power as if the bearing itself had a 0.020" larger OD.

The crankshaft is held against the drive side so there is no side thrust on the timing-side bearing. It only is subjected to radial force, so the only duty a Speedi-Sleeve would have would be to resist that radial force. It's annealed stainless steel so is softer than the steel in a ring that might be machined for the purpose, but still should have a compressive strength at least 25% higher than the cast Al it would be replacing.

Again, I'm not pushing this as the solution, only that you shouldn't eliminate it from the possibilities to explore just because some guy of unknown expertise told you so on the telephone.

I have mentioned Speedi Sleeves elsewhere…. was immediately declared a heretic and the suggestion was written off as a bodge. The fact that they were designed for repairing shafts to correct leaking seals seems to cause brain block in some folk. They aren’t for every occasion, but are a good option in some situations. I would have been happy to use one recently in a Norton gearbox, but there is just a thin web between the 2 bearing bores, so shaving it down by .010 is probably just kicking the can down the road.

The problem is this particular application, a steel bearing fitted into an Aluminium housing that goes through heat cycles. The bearing and the housing have 2 different expansion rates, every time you run the engine both the bearing and the housing are heated and the loctite is required to expand to fill the bigger gap and then when the engine cools shrink back to original size. Well after 2 attempts (I did it again after the first failure assuming a mistake on my part, second time I used much more aggressive cleaning fluids) many decades ago, both only lasting 1000 miles before failure showed as section of loctite in the sump. I gave up, the loctite under these conditions does not work. It will work in an unheated alloy hub with a steel bearing as the conditions are different. I then bought some new crankcases for a few pennies, living in Birmingham just a few years after the factory closed had its advantages.

MM, I like to think my natural tendency is to withhold judgement - or take things with a tiny pinch of salt; I thought I should mention the comments as they came up. As you say, I don't know if VB are very experienced engineers (one would like to think so) or just salesmen.
If using the sleeves in LSR engines works as suggested by Cyborg, then perhaps that should be good enough for me. Although it might be the case that such engines are stripped down & inspected far more often than I would ever wish to do with the Goldie.

So, until I am allowed out, can see the engineer & assess the plating possibility with MRJA7/8, I am slightly moving towards the LRJA7/8" with a sleeve in the housing, as UK made RHP versions are available.
If the supply of MRJA7/8 is as poor, quality-wise, as I have read elsewhere & UK RHP versions or similar quality brgs are unavailable, or virtually so, then the slightly smaller brg may be worth a punt given the closeness of specs.

Of course the bike is a big Goldie, it has some value, why would anyone scrimp & put cheap bearings in it? I wonder if this is behind some of the failures - people buying the cheapest version they can source, likely as not on Ebay, or unknown brgs from autojumbles. That said, Phil Pearson also became concerned enough to try something else. I imagine he would only be fitting what looked to be a good alternative to the UK made RHP. At the very least this suggests to me some bad quality control from certain manuf's.

Kommando, handy to be so close to the factory in the hour of need!
Re. Loctite-ish fluids, I was told that some used it on metal to metal cylinder-head to barrel joints, around the edge of the cyl. liner flange into the head. I once tried this. On removing the head some of the fluid had bonded very well (& was tedious to remove) but some had kind of flaked or chafed loose. Or the heat cycles had got to it. To me this meant that rather than sealing any tendency to blowing at the joint, it could actually make it worse by leaving gaps where it flaked & could just blow out. A straight metal joint was perfectly ok without added gunk.

I think I need to do a trial assembly & measure the clearances on the timing side - I am now intrigued as to what space there is at the bearing to timing chest interface.
I also have a rigid gearbox with bronze bush bearings - wondering if needle rollers could be fitted. But that is for another thread methinks!

It takes very little technical knowledge to, say, order a new 67-3065 (gearbox sprocket) or 65-2487 (tappet) to replace worn or bent ones. However, it requires more than a parts book and credit card when entering the realm of reclaiming worn or damaged pieces.

The web has lots of suggestions for repairing loose bearings (sleeving, Speedi-Sleeve, knurling, welding, center punching, Loctite, JB Weld, ...). As a relevant aside, although many uses of JB Weld are bodges, that doesn't mean all uses of appropriate engineering epoxies for specific applications are bodges. Unfortunately, much of the information to be found that might seem to apply to a loose Gold Star timing-side bearing is either wrong or inappropriate. So, if falls on the individual to either have the technical knowledge themself to separate the wheat from the chaff, as it were, or to be able to decide who might be offering correct advice, and who is probably blowing smoke.

Deciding not to use a Speedi-Sleeve to sleeve a Gold Star bearing because they were "designed for repairing shafts" would be the same as deciding not to use a, say, cut down Ford 351 valve because it was "designed for a Ford." What matters in each case is whether the technical specifications are appropriate for the intended use, not what they were "designed for."

I don’t know if he has used one of those sleeves in the LSR engines, just that he has an excellent reputation and he installed one for me. Unfortunately it hasn’t been torture tested, so can’t speak to longevity…. but I would wager it is no worse than the stock setup when it was new. At the time, I had just set up the mill and that sort of work was outside of my wheelhouse. Could be that it still is, but that sleeve in the photo is destined for a Norton Inter drive side main bearing where the bearing bore was staked and the bearing only contacts the nubs created by the punch….. its ugly. I have no concerns about using it and makes sense for my particular application.

I think Loctite gets a bad reputation sometimes because folks use if in places where they shouldn’t, or they use the wrong type. Sometimes out of ignorance or sometimes just don’t want to spend the money for the correct stuff. Using the proper cleaner or primer if required makes a big difference. Even thread locker if it’s a critical area….clean the threads and when you are sure they are clean….clean them again.

The very thin aluminum (appx .035") between the cam spindle bores and the main bearing housing is a concern on these engines. This already skimpy situation leaves very little for boring oversize for a speedi-sleeve type fix. A slightly oversize bearing OD sounds really attractive to me if one is available.

Good Thinking Stuart

John

The *weakness* of these castings, particularly in the cam spindle area next to the bearing housing has been noted historically. That said, I have seen very few cases actually broken in this area, unless something else has seriously let go in the motor. Worn main bearing & cam spindle housings, definitely! The age of the castings mean that bearings will have been replaced many times, sometimes more crudely than others, cranks have run out of true with loose shafts, the bearings have had a hard time & have spun.

Later cam spindles are ground away where they face the main bearing bore (when correctly aligned). I believe this was done to relieve stress in this area. I assume BSA noticed some distortion or problems with the main bearing ring seating correctly.

In considering the use of a Speedi-Sleeve or similar to take up housing wear; IF my calculations are correct...
The SKF Speedi-sleeves are 0.28mm 0.011" thick, which would require the housing to be bored out to 57.63mm / 2.269" dia. for a .076mm / .003" press fit with the MRJA7/8.
This reduces the material between the housing & the spindle bores to .024" - as Monty Python said, "wafer theen" :-) TBH I have never measured this but the bores are Very close.
Speedi-sleeve no. 99225 is for dia.'s 57.12 - 57.28mm, in the range of the 2.25" (57.15mm) MRJA7/8 brg. The inner sleeve dia is shown as 57.15, Exactly the same as the bearing. At 19,84mm wide to the "snap" groove, it gives enough depth for the bearing. As discussed in MM's Ariel thread, it does req. some width to be trimmed from the sleeve to fit flush in the housing; as it is so thin it would have to be mounted on a support tube (old bearing?) & parted off. I could easily imagine instantly reducing a very thin £35 sleeve to scrap if badly done,,,

If there is concern about over-boring the case for an 11 thou. sleeve; with a more delicate skim, a nominal 1/8" alu. (or steel?) custom sleeve in the housing would enable the use of the 2" LRJA7/8 bearing.
Either way, the housing if worn by a spinning bearing, will have to be checked for truth & corrected if necessary. There is not much to play with, so it seems the very faintest of skims will be reqd. to avoid making the housing too slack again...
It would be nice to think that a spinning bearing has worn the housing evenly but... It may not be, & just fitting a slightly oversize custom bearing could be less than ideal.

The fact that engines have survived with loose housings "staked" in emergency, with a centre punch, shows how much abuse an otherwise healthy engine can take, or how much tolerance there might be for a misaligned bearing.
As the crank is pulled over to the drive side & quite firmly located there, the timing side is "merely" in a supporting role...
I will update as & when I get to see the engineer & see what options that leaves me with. :-)

On the subject of trimming a Speedy sleeve, I would mount it on the bearing that I am going to use. Then mount the race on a mandrel and use a tool post grinder to trim it to size.

As Stuart points out…. you’ve got a bit of a dog’s breakfast there.

Ps… while the measurements suggest the the sleeve will more or less slip over the bearing, it does take some force. It would be good practice to make sure both the outer race and sleeve are clean and use the appropriate Loctite.

Makes sense Cyborg.
In my pondering I assumed that if the bearing ring was cooled sufficiently & the sleeve warmed then that would likely be enough to slip it on. I guess that at a later brg replacement time, the sleeve would have to be re-done as well as I can't see it being easily reused.

As you say a dogs breakfast, or I was going to say, Hobson's choice but I think it's actually Sophie's, whoever she is

I am quite torn between the 2 main options; to bore out the loose housing several more thou to take a std bearing with Speedi sleeve, which seems entirely feasible, or go for a thicker more permanent custom sleeve & the 2" bearing.

The reputed shortage of a reliable std bearing, plus the wafer thin wall between the bearing & spindle bores is slightly pushing me more towards the thick sleeve & 2" brg as quality versions are apparently available. This would also mean the bearing could be replaced with an off the shelf bearing, rather than having to order an oversize special or get another bearing plated, or faff with the sleeve.
I hope to see the engineer tomorrow to discuss the clearance I currently have.
No-one has questioned the spec' differences between 2 1/4" & 2" brgs. I mentioned earlier in thread. I trust that everyone thinks its a fair trade off!?

Have you spoken with a metric bearing supplier who may have a bearing that would fit your needs . Think i would prefer to shim the shaft if I could get the right size to fit the existing aluminium casings

The timing-side roller bearing is a Hoffmann RM9L, R&M MRJA⅞, SKF CRM7, or FAG RFM9.

The SKF catalog shows the following

CRM7A
⅞" ID
2¼" OD
11/16" width
Static capacity: 3750 lbs
Dynamic capacity: 6400 lbs
Max. speed: 10,000 rpm
(BSA used a CRM7 but the catalog I have notes the 'A' suffix is a replacement design with increased capacity and a pressed steel cage)

Their lighter-duty equivalent to this is the CRL7A:
⅞" ID
2" OD
9/16" width
Static capacity: 3600 lbs
Dynamic capacity: 6200 lbs
Max. speed: 13,000 rpm

Substituting the smaller, marginally-lighter-duty CRL bearing for the correct CRM would only cost ~4% in load bearing capacity, which seems to me to be negligible for all but possibly Isle of Man use. The smaller bearing would work in your present bearing cavity with a bespoke sleeve (and a ⅛" thrust washer) having ~⅛" wall thickness.

A quick look at possible metric alternatives didn't turn up an obvious solution, but I only spent a few minutes looking.

The metric equivalent seem to be exactly the same size as the imperial . In the U.K.
Brand: RHP.
Inside Diameter: 22.23mm
Outside Diameter: 57.15mm
Width: 17.46mm
Seals or Shields: Rubber Seals
Clearance: Standard

Availability:



Our Price: £59.51
(Exc. 20% VAT)
(£71.41 Inc. VAT)

That's because it is a nominally imperial bearing but with the dimensions shown in metric.

When I worked in a UK factory the buildings were labelled as Metric or Imperial. In a metric building all the measuring equipment was in metric only and all the drawings were dimensioned in metric only. In an imperial building all the measuring equipment was imperial and so were the drawings.

But both buildings could make the same parts if needed and the customer would be none the wiser, you just ended up with 2 drawings, one dimensioned in Inches the other in mm depending on which building was chosen.

When you changed a building from imperial to metric you did it over a weekend, pulled and replaced all the measuring equipment and drawings so on monday the building started up with all metric measurements and equipment to measure.

At the end of the day 1" = 25.4mm is all you need to know, I work 80% on imperially dimensioned parts on my lathe and mill but both DRO's are set to mm.