The "Trinity Bicycle" guide is quite good. The crucial part is in picture number two: Keep the plug assembly steady while turning the lock nut. That is to avoid breaking the internal wire.
Particularly enlightening the figures in microns of cone movement. A 50th of a turn is just over one degree, and that translates into a very relevant tightening of the cone! Impressive. That explains why the effect of a quick release on the adjustment is so massive. All hubs are too tight out of the factory, maybe they think it would look even worse if they sent out new hubs with a bit of play to allow for the quick release compression. Plus not everyone tightens the quick release the same. Difficult issue for the manufacturer.
I insert a set of steel washers on either side to make up for the dropouts and tighten the quick release to check the proper adjustment without having to mount the wheel on the frame. I get a more precise feel of the adjustment with the axel in my hands, But that is with regular hubs, dyno hubs are hopeless.
I agree that most likely my RHS cone is also damaged. I didn't take any of tne precautions you recommend to prevent the balls from dislodging. I oh ad already opened left cone and flipped the wheel around in all directions. I had read you before saying that sometimes you can move the right cone out a tiny bit, enough to squeeze some lube into tne bearings without undoing the connector. I tried hard to press on tne axel from the left side but it wouldn't move. I was trying to get at some of the grease on that side to get an indication of possible damage as the grease on the left bearing was like glitter, full of metal flakes. But in the case of the T785 at least, there is no movement at all. I should think that there is hardly any risk of the bearings dislodging because Shimano sell the whole internal assembly as a spare, ready to thread onto the shell and it must be constructed in a way that the bearing balls stay put, for obvious reasons.
Thus, I cannot understand how I could have a feel at the RHS bearing following your tip 3. It may not be possible with this hub. But it may be a a good idea to take out the whole internal assembly undoing the 32 mm cap nut. I would probably notice a badly failing bearing without the magnet drag at play, or at least get at some of the grease to check for "glitter". Is it a normal thread that loosens counterclockwise?
As I said, the axel would not move out even applying force on it, but it would wiggle around as far as the shell allows. No idea if that is normal with a healthy bearing, only way to know is having a fresh internal assembly to compare.
One Option would be buying two cones and attempting to replace both, assuming they are identical. The brittle wiring and the delicate "painted" insulation are a source of concern. If I fail I would have to get an internal assembly, so the entire innards would be new. The cup race looked OK and I believe they have a hardening treatment far superior than the cones.
Last edited by Cyckelgalen on 1 May 2020, 12:13pm, edited 1 time in total.
the reason the bearings cannot be felt as smooth or draggy normally is because of the magnets. With the centre out (which does indeed unscrew on a RH thread), there is no magnetic resistance and the RH bearing is complete and any roughness can be felt quite easily. You will soon see once you have the centre out.
There is normally at least 1mm movement possible of the axle rightwards (upwards if you have the wheel positioned as suggested), once the LH cone is removed. There is always some clearance rather than none before the stator hits the bearing plate. The balls indeed ought to be captive in a spare centre assembly but I can tell you that in many assembled hubs they are not, and can get out of place. If you back off the RH cone one turn (eg to allow lube to be introduced more easily) then the balls will probably be able to come unseated, even if they couldn't before.
FWIW I am working on a modification to Shimano hubs which will
a) be possible to carry out without taking the hub to bits and b) will allow the bearings to be regreased in situ and c) will not interfere with the normal operation of the hub.
Obviously I shall carry out some trials on one of my hubs first. to make sure it is 'a goer'.
The above* seems to stretch the definition of serviceable a bit......? If they wanted to sell it as a servicable item to the masses surely they would not have made the electrical connections so vulnerable?
*edit I meant the hassle of lubimg an ummodded shimano dynamo hub not Bruceys proposed improvement.
Last edited by SA_SA_SA on 31 Mar 2021, 1:56pm, edited 1 time in total.
------------You may not use this post in Cycle or other magazine ------
SA_SA_SA wrote:The above seems to stretch the definition of serviceable a bit......? If they wanted to sell it as a servicable item to the masses surely they would not have made the electrical connections so vulnerable?
shimano (unlike freehub bodies where they started one way and then changed their mind later) never have claimed that the RH bearing in their hub generators is serviceable, hence they don't list the RH cone as a separate part, only the entire centre, connector block and the LH bearing parts are listed as spare parts.
If you want a hub generator where you can buy the spare parts and rebuild the connector assembly yourself, I think that Sturmey Archer is your only option; the connector assembly is available as part HSE001
Exactly, the connector/right side cone is not serviceable and not listed. That is why I wondered if it is identical to the left one, wich you can get as spare y2yj98020. Still some people service it and with luck, succeed. If the connector side bearing is bad and the whole unit has to be discarded you can just as well have a try. If you fail the new cone and balls can be kept as a spare.
On the issue of balls dislodging, most likely the balls on the connector unit are caged. I have seen a few of the newer shimano hubs that have a cage on one side and loose balls on the other. The XT dyno hubs belong to the new breed with aluminium axels. Locknuts should be well tight to retain adjustment but I always fear ruining the obviously softer aluminium thread. Another unwelcomed development is that the ball size in this hubs has been reduced from 1/4 to 3/16, that is the standard size of front hub balls is now used at the rear. Bigger balls should make the bearing more durable, that is why they,were used in the hubs that carries most of tne weight.
By the way, I see no data on Shimano's original ball hardness and grade. Anyone knows? Are there different qualities for various product ranges?
Those modifications you are working on Brucey, are going to drill a hole in the shell to allow easy refilling with grease?
Brucey wrote:....shimano (unlike freehub bodies where they started one way and then changed their mind later) never have claimed that the RH bearing in their hub generators is serviceable, ...
If that is case would not (hopefully large) cartridge bearings not require less manual labour in production?
------------You may not use this post in Cycle or other magazine ------
All dynamo hubs that I know appart from Shimano have indeed cartridge bearings. But there we enter the realm of the never ending controversies in cycle touring particularly , cup and cone-cartridge, disc-rim brake, steel-aluminium...
Brucey wrote:....FWIW I am working on a modification to Shimano hubs which will a) be possible to carry out without taking the hub to bits and b) will allow the bearings to be regreased in situ and c) will not interfere with the normal operation of the hub. Obviously I shall carry out some trials on one of my hubs first. to make sure it is 'a goer'....
Excellent (hopefully) . I didn't notice this earlier.
------------You may not use this post in Cycle or other magazine ------
Brucey wrote: FWIW I am working on a modification to Shimano hubs which will
a) be possible to carry out without taking the hub to bits and b) will allow the bearings to be regreased in situ and c) will not interfere with the normal operation of the hub.
Obviously I shall carry out some trials on one of my hubs first. to make sure it is 'a goer'.
An update on servicing the DH-T785 dynamo hub. That is the disc brake version of the T780 and I presume they are identical, apart from the disc spline. Subsequent incarnations like the current XT hub T-8000 are almost identical.
One thing to notice is that the 32 mm nut that you need to undo to dislodge the Internal assembly on tne RH of tne hub is octagonal, so you can't use any regular hexagonal spanner. Shimano making things easy again, and probably making you buy their special spanner for the job.
The "non-serviceable" internal assembly can easily be disassembled and the cables that run through a groove in the axel seem sturdy enough to endure some bending to remove the connector bits and the cone. So it would be easily serviceable if it wasn't for the fact the both cones are quite different, the internal assembly cone (RH cone) is not sold as a spare and, by the look of it, I am almost certain it doesn't match any other Shimano cone that you can source.
To rebuild the hub you would need the LH cone, 13 bearing balls and an internal assembly. Given the prices, I opted for buying a new hub on eBay and substituting all the innards of my hub to avoid having to lace the wheel again. After the job I only have the LH race of the old hub, which seems to be OK despite the cone being slightly pitted.
Cyckelgalen wrote: One thing to notice is that the 32 mm nut that you need to undo to dislodge the Internal assembly on tne RH of tne hub is octagonal...
that has been the case for all shimano hub dynamos. There are 32mm and 36mm octagons. There is also a splined tool that is used on some of the less pricey models.
The "non-serviceable" internal assembly can easily be disassembled and the cables that run through a groove in the axel seem sturdy enough to endure some bending to remove the connector bits and the cone. So it would be easily serviceable if it wasn't for the fact the both cones are quite different, the internal assembly cone (RH cone) is not sold as a spare and, by the look of it, I am almost certain it doesn't match any other Shimano cone that you can source.
different cones? This is new to me.
I wonder how many of the hub generators which have twin-wire only connectivity (rather than earth-return compatibility which is how all shimano generators were for many years) share the 'odd cones' approach? Note that in some cases the same model (eg DH-3N72) exists in both types.
Q; are the cones both M14 threaded? If so there are several other shimano M14 threaded cones which might accept the (large diameter) seal from the RH cone?
.