Replacing Hollowtech II Bearings

For discussions about bikes and equipment.
Brucey
Posts: 44516
Joined: 4 Jan 2012, 6:25pm

Re: Replacing Hollowtech II Bearings

Post by Brucey »

well here is another fact for you; every assembly with half decent seals and correct lubrication I have used on cars, trucks, industrial plant, motorcycles, bicycles etc has lasted very well in foul conditions despite not having stainless steel bearings. In some cases, because of that.....

I'm sure the Hope BB is a half decent piece of kit but you can buy about five shimano ones for the same money and if you just want stainless bearings they can be fitted to a shimano BB for peanuts.

As I mentioned previously the load rating of stainless bearings is usually LESS than that of standard chrome steel bearings, so where stainless bearings outlast standard ones you can be sure that the sealing/lubrication are simply not working for you, and there are some easy gains to be made there.

If anyone thinks this is 'psuedo-science' then I suggest that they go and read a few engineering texts and bearing specifications. :roll:

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~
MisterTea
Posts: 41
Joined: 6 Sep 2013, 1:30pm

Re: Replacing Hollowtech II Bearings

Post by MisterTea »

Fyi. Hope BB are approx twice the price of XT, not 5 times.

Personally I believe in buying once, longevity and reliability are key to me. (This is a principle instilled in me long before my days studying Engineering at Brunel Uni, where I also learnt that Engineering texts are useful guides, that they are always being updated with latest findings and that observations of real current world complex situations are invaluable)

The quality of the steel and the design of the seals used is vital, and the quality of Hope bearings and their sealing methods are clearly far in excess of others. The fact they use stainless bearings is simply to ensure that any inevitable minuscule water ingress does not degrade the integrity of the bearings. They are custom made for Hope by INA

One other observation made is that the reason left hand (non drive side) bearings fail first is due to the fact that many folk over tighten the knurled adjuster that takes up the play on Shimano cranks. This places unequal pressure/too much pressure on the left hand bearing. Users of different designs such as srams GXP system don't get this problem. This is a well known weakness of the Shimano system and not helped by the poor bearings, poor seals and poor/lack of grease in Shimano BBs.
Brucey
Posts: 44516
Joined: 4 Jan 2012, 6:25pm

Re: Replacing Hollowtech II Bearings

Post by Brucey »

MisterTea wrote:Fyi. Hope BB are approx twice the price of XT, not 5 times.
fair enough, but I wouldn't buy XT ones since there appears to be little difference between those and the cheaper ones. When I've opened them up (unless I've missed something....) the bearings appear to be the same in most shimano units.

Personally I believe in buying once, longevity and reliability are key to me. (This is a principle instilled in me long before my days studying Engineering at Brunel Uni, where I also learnt that Engineering texts are useful guides, that they are always being updated with latest findings and that observations of real current world complex situations are invaluable)

The quality of the steel and the design of the seals used is vital, and the quality of Hope bearings and their sealing methods are clearly far in excess of others. The fact they use stainless bearings is simply to ensure that any inevitable minuscule water ingress does not degrade the integrity of the bearings. They are custom made for Hope by INA


they might be custom made but they will have at their heart standard bearing technology that is available elsewhere. Unless they have more, larger balls in they will have a lower load rating than the equivalent chrome steel bearing. Users report that they might get a couple of years out of the Hope BB in rough conditions; does this tally with your experience? If so, this is suggestive to me that the water still gets in and increases the wear rate, just not as much as the same amount of water would in a standard bearing with standard grease in it. Note also that any failed bearing I've seen has clearly worn seal edges; the failure often goes; seal lip not wetted with lube, seal wears, water gets in, the end. A properly specified lubricant, (with corrosion inhibitors, EP additives etc) and plenty of it, inside the assy, able to bathe the bearings from the inside would prolong the life of any bearings (including stainless ones), by keeping the seal edges wetted, and inhibiting corrosion should any water get in. Even reducing the airspace inside the assy is beneficial since this will discourage air (and water...) 'exhange'...-which is more like one way traffic in the case of water.... :roll: If they don't corrode then they won't wear out at anything like the rate, and by dilution even the same amount of water inside will suddenly be confronted with x100 the lubricant and x1000 (or more) the corrosion inhibitors.

One other observation made is that the reason left hand (non drive side) bearings fail first is due to the fact that many folk over tighten the knurled adjuster that takes up the play on Shimano cranks. This places unequal pressure/too much pressure on the left hand bearing. Users of different designs such as srams GXP system don't get this problem.


I agree that plenty of the BBs die through excess preload, but I'm not sure the argument that it knocks out left bearing preferentially holds water; the load is reacted by an equal and opposite load through the right bearing.

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~
User avatar
geomannie
Posts: 1093
Joined: 13 May 2009, 6:07pm

Re: Replacing Hollowtech II Bearings

Post by geomannie »

Hi

To return this thread to my original question, I sourced some off-the-shelf replacement 6805 bearings (dimensions 25x37x7mm) which I was lead to believe would fit my Hollowtech II BB shells. I should have checked the original bearings more closely. They are not 25x37x7mm but 25x37x6mm. The ones I bought are 1mm too thick.

If I search for 25x37x6mm bearings (e.g http://www.ebay.co.uk/sch/i.html?_trksi ... &_from=R40) I find almost nothing available except very expensive ceramic bearings. Am I missing something? Is my Hollowtech II (Ultegra) non-standard. Am I looking for bearings in the wrong place?

Luckily I only bought cheap 7mm replacement bearings so not much money wasted. Thanks for any suggestions.

Cheers
geomannie
Brucey
Posts: 44516
Joined: 4 Jan 2012, 6:25pm

Re: Replacing Hollowtech II Bearings

Post by Brucey »

The Ultegra cups have the same size bearings in as all other shimano units that I have seen. The difference between a 'road' and MTN BB unit is the thickness of the cups. The road ones are thicker by 1mm at 11.25mm vs MTN cups at 10.25mm. You can of course fit MTN or road BBs in most cases using appropriate spacers; the only combination that may be a struggle is road cups on a 73mm BB shell; I've not tried that.

With the bearings you have, there are two solutions;

1) Quick and dirty. Knock the original bearings out of the cups. You will find that there is a 7.6mm depth recess in the cup; deep enough for the new 7mm bearing, but not deep enough for that and the seal ring. This is OK, the seal isn't needed; the bearing has its own seals. You probably only need the outer seal, so remove the inner one if you like, esp if you are going to add extra grease. File the spigot off the 25mm diameter on the plastic bearing cover/adaptors, and refit those. The bearing assy will now be 2mm wider than it was so some mucking about with spacers is required. The bearing covers will show a bigger gap than before; you can protect them using a couple of sleeves made from old inner tube, and/or keep that area well greased/waxoyled to discourage water ingress.

2) Proper job. You need to start with road cups, not MTN cups. Machine these 1mm deeper (you need a mate with a lathe...) so that the cups have an 8.6mm recess in them. Increase the centre bore of the cup while you are at it (so that the bearings can be removed more easily next time round). Fit the new bearings with the inner seal removed. Fit the original shimano seal and modified (as in 1 above) bearing cover. You now have two seals on the outside.... Job's a good'un.

In either case it is a good idea to pack plenty of grease in the centre cavity so that the bearings stay happy.

If you buy alternative bearings with a 24mm bore (e.g. enduro bearings) they are stronger and will run directly on the BB axle but you may find that the BB shell needs facing before they can be fitted; it requires more accuracy to the facing with this arrangement.

hth

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~
User avatar
geomannie
Posts: 1093
Joined: 13 May 2009, 6:07pm

Re: Replacing Hollowtech II Bearings

Post by geomannie »

Hi Brucey

While I very much appreciate your reply, I am left wondering why my cups had 6mm wide bearings and yet all the bearing that are for sale (and are said to be for BB's) are 7mm. The simplest fix for me would be to source 6mm bearings. I am just puzzled by why these do not seem to be available. Can anyone comment on whether the standard Shimano Hollowtech II BB bearing is 7mm or 6mm?

Thanks
geomannie
Brucey
Posts: 44516
Joined: 4 Jan 2012, 6:25pm

Re: Replacing Hollowtech II Bearings

Post by Brucey »

the standard shimano bearing is 25x37x6mm.

The 'off the shelf' industrial bearing 6805 2RS is 25x37x7mm

Hope (and some other BBs) use a 24x37x7mm bearing. These will fit other cups but mean that they will bear directly onto the axle and also present outer seal issues.

'MisterTea' is right in that Hope's bearings are not available in that quality and size elsewhere. Because the bore is 1mm smaller diameter, they might have larger balls in them; anyone care to measure up?

One of the few alternatives in that size are the enduromax bearings. These bearings have a higher load rating in that size because they are a full complement bearing but they are chrome steel, so... if corrosion is at the heart of the problem these bearings won't be an improvement.

The standard shimano bearings are identical to a 6805 2RS except that one side of the bearing is lacking a seal, and can be made narrower, so that is what they did. Campagnolo UT bearings are (or were) the same 25x37x6 size and are simply 6805 or 61805 bearings with one side ground off. Some campag bearings even come with seals marked 61805.

Standard 6805 bearings typically have a static load rating of ~600-660lbs. Enduromax bearings are claimed to be about 730lbs rated. I've not seen the spec for the Hope bearings.

BTW bearing quality, seals etc are important but so are tolerances; no bearing will meet its load rating if the clearances, preloads and and fits are wrong; plenty of BBs fall down on those things big time.

[edit; if you want a cheap solution, get a tiagra BB (about £15) or similar, reverse the bearings, and use plenty of the right grease between the cups.]

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~
Cyckelgalen
Posts: 221
Joined: 21 Sep 2018, 11:29am

Re: Replacing Hollowtech II Bearings

Post by Cyckelgalen »

I resurrect an old thread to avoid proliferation of similar threads

Replacing HT bearings is uneconomical given the prices of new BB's, the new type XT BB-MT800 goes for around 15£, much cheaper than any top quality replacement bearings. It would only make sense if you want to construct a truly durable BB with SKF bearings and you could get 6 mm wide bearings. Unfortunate the standard width in that size is 7 mm and wouldn't feel comfortable machining the cups to skim off 1 mm in a critical aluminium part that has to withstand the whole weight of the cyclist.

Most cyclist can live with replacing the whole BB for 15£ every year or two, so really, going through the faff and expense of fitting SKF quality bearings to a HT BB makes sense only to the long term world tourer that needs durability at all costs.

A few years down the line, Shimano has produced a new incarnation of the often despised HT II, with smaller bearings and better sealing, they claim. After blowing my fair bit of BB52's I just fitted a MT800 XT. Are they really better that the previous generation?
I am surprised to see these BB's touted as being better because of the smaller bearings(?). Larger bearing will always be stronger and roll better over imperfections, just like larger bike wheels roll better on uneven terrain.

I have not opened one of these new type BB's and cannot comment on the sealing. I truly hope they are an improvement on the older XT BB70, which shouldn't be difficult as I believe they were identical to the Deore BB52. I opened both of them side by side and couldn't tell the difference, same inadequate sealing, identical bearing from same manufacturer, Sunhill. So shimano were charging a higher price only for the privilege of having BB70 instead of BB52 printed on the outer shell.

I would appreciate some views on the newer small bearing HT BB's and how they compare to the older type, particularly in terms of durability for long term touring.

Thanks.
Last edited by Cyckelgalen on 3 Mar 2020, 1:48pm, edited 1 time in total.
Woodtourer
Posts: 354
Joined: 23 Jan 2018, 1:51pm

Re: Replacing Hollowtech II Bearings

Post by Woodtourer »

Just came across this thread and I thought I would mention our experience with Hollow Tech. I work in a bike shop and built our touring bikes with these BB. They have performed flawlessly on 2 tours across North America. Then just for the sake of preventative maintenance I replace them. 4 European tours and they are still going strong!
Also because they are so light I carry , I don't know why!!, an extra set in our kit because they don't require a special wrench,in a pinch, if they did need replacing.
Cyckelgalen
Posts: 221
Joined: 21 Sep 2018, 11:29am

Re: Replacing Hollowtech II Bearings

Post by Cyckelgalen »

Thanks Woodtourer,

By "these BB" you mean the new type, small bearing cup XT's ? That is, MT800, MT8000, etc .
bgnukem
Posts: 694
Joined: 20 Dec 2010, 5:21pm

Re: Replacing Hollowtech II Bearings

Post by bgnukem »

If the bearings are anything like those in the square taper brackets I use then they would benefit from being filled with a decent waterproof grease. If the bearing shields can be pried off with a seal pick or thin-bladed craft knife then grease can be pushed in from one side.

A bearing I recently opened up in a new square taper bracket was almost dry inside. As Brucey has pointed out previously, lack of grease/oil at the seal lip will result in the seal wearing out rapidly allowing contamination of the (minimal) grease in the bearing.

Given the amount of flooding at the moment I guess the ability to extract and re-grease both bearings is an advantage of Hollowtech vs. square taper where typically only one (or neither) bearing is accessible.
Woodtourer
Posts: 354
Joined: 23 Jan 2018, 1:51pm

Re: Replacing Hollowtech II Bearings

Post by Woodtourer »

Not sure what bearings are in them I just use the Deore not the XT. BTW I thought about replacing just the bearings for preventative maintenance but the $ was not worth it.
Cyckelgalen
Posts: 221
Joined: 21 Sep 2018, 11:29am

Re: Replacing Hollowtech II Bearings

Post by Cyckelgalen »

The basic Deore is the BB52 and you can get it for a tenner. You shouldn't even consider bearing replacement.
The bearings are identical to the older type XT BB70, I believe, so god value really. But on a long tour I'd rather have something truely dependable that I don't need to replace regularly or carry a spare.

They are really easy to clean and lube and that will extend their life. You just need to pry out the cap and outer seal and the bearings are fully exposed, as the cartridge bearing has its own seal only on the inner side. It could be done on tour even without taking the BB off the bike. I replaced a BB52 because one bearing was going rough, left one as usual, and on closer examination after taking it completely appart, there was no real damage or wear on the races or balls. Thoroughly clean and lubeb it could have carried on for a long time. The way the bearing is constructed, you can actually take it appart and put back the bearing balls in the races without any special equipment, quite unusual for a cartridge bearing. The balls are held by a plastic cage, flexible enough pry out and remove and replace the individual balls.
So I presume that my rough bearing was a contamination rather than a wear issue and it could have been revived if needed in a touring situation.


Still, a basic question remains. The older Deore and XT (BB52 and BB70) were identical. What about the new small bearing XT, MT800-MT8000, is it really better? It is supposed to have better sealing, tha is an improvement but the smaller bearing diameter isn't. Well, if no one comes forward saying that they have blown one of the newer XT's, that may be a good sign...
Brucey
Posts: 44516
Joined: 4 Jan 2012, 6:25pm

Re: Replacing Hollowtech II Bearings

Post by Brucey »

the unknown factors in any bearing installation are often

1) preload and
2) load sharing

Re. preload. This is comprised of those loads imposed on the bearing by its installation (e.g. push fit into a housing) plus any lateral preload (eg from an adjustable bearing cap). Quite often the first signs of trouble in a bicycle bearing installation are when a cartridge bearing starts to feel rough. Very often the bearing, once removed from its housing (eg BB cup) , is mysteriously smooth-running. In this case you can be pretty sure that the interference between the bearing and the housing is 'too much'. By too much I mean that it is more than enough to remove any few microns of free play there might be in the construction of the bearing. When a bearing starts to feel rough in situ, it is certain that there is some damage to the balls/raceways; however the damage may be trivial, and the preload on the installed bearing not. It is very easy to end up with a preload on the bearing that vastly exceeds the usual service loads.

Re load sharing; it is important that the load is shared between several of the ball bearings present, if they are small in size. [In old BB assemblies, each 1/4" ball was capable of accepting the entire (normal) service load, more or less, which is an inherently more tolerant design.] Factors which influence load sharing are
- free play in the bearing
- the stiffness of the raceways and housings
- the preload on the bearing

Most cartridge bearings are designed (or based on designs) for high speed operation. In these designs it is normally vital that there is no free play in the bearing. In fact it is desirable that there is enough preload on the bearing that the unloaded side of the bearing doesn't go slack even when the peak service load is encountered. This means that the preload is usually equal to or in excess of the anticipated service loads. If the balls are allowed to run slack at high speeds, they soon start to scuff and this soon wrecks the bearing. However at low speeds (all bicycle use) this is not quite the same issue; the usual consequence of slack running is that the balls scuff more than normal against the clip or bearing retainer. If the retainer is steel this eventually causes enough wear that the balls will fall out of an (open-sided) clip, and obviously the bearing will be contaminated with wear debris well before that. If the retainer is plastic (usually acetal (eg Delrin)) then it will normally withstand prolonged scuffing at low speeds without appreciable damage either to the clip or the bearing itself.
A further alternative is that there is no clip, just balls present. This is called a 'full complement' bearing and will normally have an increased load bearing capacity and a much reduced speed rating. During slack running the balls scuff against one another, which (at low speeds, well lubricated) is usually relatively harmless. If there is more slack then the balls may scuff against the raceways as they start to see load (bad). Most traditional bicycle bearings are (effectively) full complement angular contact bearings. Note that keeping the bearings adjusted properly always improves load sharing and it also reduces any possible wear arising from scuffing. IME very many traditional bicycle bearings ought to last 'for ever' if they are kept properly lubricated and adjusted.

It is possible to make a deep groove bearing which is full complement, but it is more expensive to do this (you need loading slots in the raceways) and there may be changes in the way the bearing ought to be used. If such bearings were sold in similar numbers to standard bearings, they might only be x2 the cost; however they are not at all common and normally cost ~x10 as much as a consequence.

Clips are used in bearings for reasons of practicality (eg deep groove cartridge bearing or simply to ease assembly) or just to reduce cost; the clip often costs less than a single high precision ball to manufacture. Thus fewer balls and a clip costs less to start with and can make the thing easier to assemble.

It is instructive to look at what works in practice; for example most SA hubs use 1/4" balls retained in steel clips and furthermore they are meant to be adjusted for slack running. Yet they usually give no trouble whatsoever; if you pull apart a SA hub and the balls fall out of the clip (because the clip is worn), you can be sure that the hub has been run very slack, for a long time and/or badly lubricated and/or under very high loads. In recent years SA has changed the design of their ring bearing from 'full complement' to fewer balls in an acetal retainer. It remains a reliable assembly. However clipped balls in some installations (eg the left side of many coaster brakes) are living on borrowed time from new IMHO; they are practically doomed to fail at some point, it is more a question of when rather than if.

So returning to the question at hand; the newer HTII BB assemblies use a slightly smaller bearing than previously. But the bearing stiffness and preload will also be different, and thus the load sharing will be too. If this is done well, there could be a significant improvement in the life of the assembly, even though the bearing on the face of it 'looks weaker' . If the seals are also better (and I am pretty sure they are) then the usual life expectancy will be much greater, even if the bearing is in fact weaker; the usual failures in HT-II BBs are related to excess lateral preload (sometimes) and/or (more commonly) water ingress.

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~
Cyckelgalen
Posts: 221
Joined: 21 Sep 2018, 11:29am

Re: Replacing Hollowtech II Bearings

Post by Cyckelgalen »

Thanks Brucey for your thorough reply.

I have actually experienced rough BB cartridge bearings running smooth once out of their assembly/housing. I thought it might be a random effect of my unorthodox method of knocking them out, but it never occurred to me that press fitting them into place might compress the outer raceway enough to add a certain preload. That means that they are etremely sensitive to even the slightest preload and that, in most cases, no further preload should be added with the adjustment cap. The exact point where the lateral play of the spindle disappears should then be the correct adjustment.

On the new XT bottom brackets, certainly spreading the load share to a higher number of balls would result in a stronger bearing. The old HT BB's have 15 bearings, 1/4" size. I have no idea how many bearings the new HT have, as a non serviceable Shimano part, no spares or data are available. But it is logical to assume that they have quite a few more if the ball size has been reduced while the inner race ring remains the same, 25 mm diameter.

Still, I have my doubts that shimano's intention was really to produce a more resilient bearing with an improved load sharing. This is probably an unintended effect of their quest to improve things by simply making them a few grams lighter.
Post Reply