Sa ag/fg/gh6 dynohubs: why no modern magnet ring spares

For discussions about bikes and equipment.
User avatar
6.5_lives_left
Posts: 38
Joined: 9 Oct 2020, 9:27pm

Re: Sa ag/fg/gh6 dynohubs: why no modern magnet ring spares

Postby 6.5_lives_left » 10 Jan 2021, 6:21pm

I am surprised that no one has pointed out that all you need is for one of those replacement magnets to come loose, jam the rotor and stator together and the rider is pitched over the handle bar.

Actually with one of those neodymium magnets, being so brittle it might just be rendered into dust, but I wouldn't want to count on that.

Maybe you could pot everything with epoxy to stop any bits becoming loose, except for the air gap of course.

Brucey
Posts: 42939
Joined: 4 Jan 2012, 6:25pm

Re: Sa ag/fg/gh6 dynohubs: why no modern magnet ring spares

Postby Brucey » 10 Jan 2021, 8:59pm

I have seen both electrical motors (many of which are built in a similar way, magnet-wise) and hub generators where the magnets have started to come unstuck; in both instances the thing typically just becomes noisy and draggy to turn, rather than jams solid. The reason for this seems to be that the magnet wants to be stuck to one or the other part. In the case of the most recent hub generator the thing worked perfectly again once I simply used some fresh glue on the magnet.

BTW the hub generator in question was a shimano DH-3N20 model, which uses ferrite magnets. These magnets come in a set of three, and they must be assembled in the correct sequence inside the hub, because they are magnetised (as a set) during manufacturer such that the alternating N-S-N-S field pattern (which matches the pole pieces on the stator) is specific to the magnet arrangement.

Incidentally I have reason to suppose that some cleverness has been used in the manufacture of this hub generator, in that the magnetic field in magnet appears not to be in parallel-sided rectangular areas which are an exact match for the pole pieces, but rather have a slightly different shape, such that the transition from one pole to another is made less abrupt. In any event this particular hub generator has the least obvious 'cogging' of any similar (inexpensive shimano) hub that I have seen.

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

User avatar
ConRAD
Posts: 720
Joined: 20 May 2010, 10:55am

Re: Sa ag/fg/gh6 dynohubs: why no modern magnet ring spares

Postby ConRAD » 11 Jan 2021, 8:30am

Well, and why not instead of an iron-copper plated back ring with eg 26 magnets glued on it, we do not think about a single cast molded piece of neodymium, properly shaped and magnetized in a N-S 26 poles sequence. Minimizing in this way overall magnetic circuit reluctance, strengthening the field and improving so far overall hub efficiency??
Image

Brucey
Posts: 42939
Joined: 4 Jan 2012, 6:25pm

Re: Sa ag/fg/gh6 dynohubs: why no modern magnet ring spares

Postby Brucey » 11 Jan 2021, 9:20am

there are twenty poles in an old SA generator. There might well be something to be gained in theory by using a one piece magnet -but in more modern materials- provided it can be made at a cost (large shaped magnets tend to be expensive) and it doesn't turn into a structural problem; it needs to be mounted without risk of cracking for example. However the constant field reversal in the claw poles means that this is where magnetic hysteresis losses are most likely to affect the generator's performance.

That the most efficient generators presently available (even expensive ones) use separate magnets and backing rings, (and appear to use cold-formed claw poles too) suggests that clever design can overcome deficiencies in materials to some extent.

One of the things I think is very interesting (hence the other thread about reactive loads) is the synergy between the load and the generator. Increasingly it seems that generators are built so that they satisfy the Stvzo test, but are more likely to be used with a non-Stvzo load attached, and not all pairings respond in the same way.

Probably there is something to be said for designing both generators and the lamps they drive as a set, if the best match between generator and load, and best overall efficiency is to be obtained.

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

SA_SA_SA
Posts: 2034
Joined: 31 Oct 2009, 1:46pm

Re: Sa ag/fg/gh6 dynohubs: why no modern magnet ring spares

Postby SA_SA_SA » 11 Jan 2021, 11:52am

Brucey wrote:...Probably there is something to be said for designing both generators and the lamps they drive as a set, if the best match between generator and load, and best overall efficiency is to be obtained...

Or have a matching specification* for allowed electronic front lamp load behaviour, so that pick and mix from different Led lamp suppiers can continue. If the rear lamp is always required to be wired to terminals on such a front lamp, then no change in specification would be needed for rear lamps, so backwards compatibility with all dynamo rear lamps is retained.

*a bit like Emc rules for mains PowerSupplies.
------------You may not use this post in Cycle or other magazine ------ 8)