Theory for dynamos

Stronger magnets (or a shorter distance between the magnets and the inductor) result in bigger peaks in attraction/repulsion (to get from one stable position to next to you need to overcome the repulsion) and perhaps a slightly more efficient generator, but that all depends on the particular construction. Compare this with car alternators, in which the magnetic field is altered giving an efficiency of ca. 60% over a large range, while dynamo hubs get to about 60% peak. From what I read the lightspin dynamo (which is no longer being made) used the same principle and was supposed to have much higher efficiency of 85% (but tests with that dynamo didn't show such high efficiencies).

In a wheel with a SON 28 you can feel the number of poles by rotating the wheel by hand. There are 26 positions in which the wheel remains in the same position, if you want to turn the wheel forward/rearward a little, that takes effort. With a wheel built with a Shimano dynamo hub, it is much harder to find the steady positions and thus count them in this way. This means the magnets in the Shimano hub aren't as strong (or the distance between them and coils or rather the metal parts are bigger). Using a loose dynamo hub the number of poles can be determined much more easily. A Shimano dyohub has 28 poles, which is likely part of the reason why its vibrations are less noticeable, similar to what the tests with the SP hubs in 2011 showed (in the SP hubs there is more, there is an asymmetry in the force curve, i.e. going one way is different than the reverse, this may help...).

The speeds at which I feel the vibration are esp. multiples of about 7.5 km/h. This means, when one considers the wheel's circumference is a bit more than 2 m (ca. 2.12m for a 37-622 tyre), and 26 poles for the dynamo, that the vibrations are multiples of ca. 26 Hz (7.5 km/h is just over 2 m/s), and at 22 and 30 km/h it's about 75 Hz and 100 Hz vibrations. As I said it depends on resonating frequencies, wheel weight etc. on how much you will notice such vibrations and what the exact vibration frequency (and thus speed at which you notice it) is on another bike.

So the vibration frequency will always be low (small number of poles in a hub generator and slow rotation frequency of the wheel), and conversely the strength of each jolt high, and the best way around this would be to increase the number of poles in the dynamo. S-Ram uses 36 instead of 26 and 28 for Schmidt and Shimano respectively, but a geared hub generator is another way to get there. The only contender for such a system is the Renak Enparlite, which has noise from the gearing. I've been meaning to test the Renak Enparlite but haven't got around to it yet.

What I noted since the inception of these pages: The movement in the front fork, which is of course not 100% stiff, makes resonance possible. A rear wheel dynamo hub would therefore be better, because then almost the whole bicycle will be a sort of counterweight for each jolt.

Another thing that influences the resonance is the electronics in the headlamp. It may interest you to know that the Edelux is one of the worst headlamps in this respect... I always test dynamo hubs with the same headlamp so that the experiences between hubs are comparable, and that headlamp is the Edelux. I also always test dynamo hubs on the same bike that I've tested all other dynamo hubs with, using the same size/mass tyre (ca. 32-622). Actually I test dynamo hubs with other headlamps/tyres/bikes too, but my conclusion is really based on the one with the Edelux in my standard bike and nearly the same rim/tyre (changes in those don't matter all that much in my tests) because it shows the vibrations most.

Addition 2011-3-14: Tightening the skewer of the SON28 as was suggested to me does not remove the vibrations. This simply has nothing at all to do with the problem. Note that if the skewer wasn't tight enough, so that the axle could move up/down a bit, the SON28 would rotate around its axis because of the resistance caused by the magnets and by the induction, and start pulling the cable to the headlamp taut, which I would have noticed! (well, theoretically there is a small chance in a specific case, see for more on this below) The skewer was tightened quite a bit in all tests; the SON28 simply has larger vibrations than the other dynamo hubs. The vibrations also don't disappear in those other hubs by tightening the skewer there. What can be an issue in noticing dynamo hub vibrations is the headset: If it's loose vibrations will become more noticeable. Note that the SON28 tests were all done with a properly tight headset so that wasn't an issue.

Btw., the guy who suggested this tightening the skewer is Peter White, who almost certainly did it because my page gives him problems explaining away the vibrations to people whom he sells SON28 dynamo hubs to. Nearly 2 months later he has not responded to my dismissal of this solution in which I explained the issues (10 months later still nothing, but then that was to be expected...). I have named him in a posting but now also here not just because of not responding, but also because suggesting this obviously non working solution to me implies I don't do any proper testing (a friend of mine got angry about that, I wasn't bothered at first, but I thought about it and had to agree).

Schmidt themselves acknowlegde the vibration issue: http://www.nabendynamo.de/service/pdf/Service_SON_e.pdf. In the fault finding table they write:

Noise (rattling noise) - Are there any vibrating parts on the bike? -
vibrations
                         may be reinforced by a SON hub dynamo.
                       - Is the hub axle fastened correctly?

This "Is the hub axle fastened correctly?" is of course not a solution to the general problem of vibration in the handlebar. Perhaps others believe this is a factor because Schmidt says it can be. Well, it isn't!

Addition 2011-10-22: After an email about this subject, where a reader mentioned to me that the vibrations got bigger one ride, and when tensioning the skewer, they were normal again, I want to add the following which you can deduce from the rest I wrote, but I suppose being completely explicit is better: The only case where the vibrations can be increased is when the hub is not secured tightly enough, such that there is movement possible (hub compared to fork I mean). Then the hub can move forward/backward with the repulsion/attraction from the magnets, but, then the hub will start to rotate (slowly). This is because the hub in general is prevented from rotating by friction, not by a notch fitting into a hole (unless in an unlikely case where a front fork has a pattern of ribs exactly fitting the pattern on the axle ends of the dynamo). So if you leave the hub with less tightened than required skewer, after a while you will very probably notice the connector for the cable to the headlamp is in a different position, and after that at some point it will stop rotating, when it has pulled the cable to the headlamp taut... I have experienced that phenomenon of stronger vibrations + rotating hub a few times (e.g. with an SP PD-8 too) but when testing for vibrations I make sure they are fixed properly.

I have put part of my response in the section further on, which gives a lot of explanation on all the issues involved. I will edit and clarify that soon.

Addition 2011-6-27: In the review of the SP switchable dynamo hub PD-7 (=HB015) I've put some information showing that the people from SP are well aware of the issue of vibrations from dynamo hubs and they take this seriously. This isn't something that is eliminated by 'tightening of a skewer'.

A solution to the vibration issue? As I noted in a few places, the lamp's electronics is of significant influence on the vibrations. With a different type of electronics, that chops up the output from the dynamo and transforms it as in a DC-DC convertor you get not only the maximal power output from the dynamo, but also the advantage of the low frequency vibrations (possibly) disappearing for a large part. More about this on my page about this new dynamo driver where I will test how well this works.

The SON28 will not become usable for me with that system because with the SON28, slight vibrations can still be felt with light (power) off...

More thoughts on vibrations

Here's a response I sent someone who experienced more vibrations from a SON 28 with a somewhat loose skewer than with a properly tightened one. He tightened it after that ride, and this indicates vibrations can get stronger with a too loose skewer, however, this is the same situation I mentioned, where the dynamo's axle will start to rotate:

There is a possibility of a hub moving a little bit, while at the same time not being able to rotate, only if the hub's notches correspond with those on the fork's ends such that you get a situation as with gears, where one fits into the other, and one cannot move unless the other moves even if there is play.

However, notches in the forks are not exact fits for those on hubs, so what you invariably have is friction keeping the hubs from rotating. If there is movement, then the vibrations can be felt stronger because slight rotation + counter rotation is possible which means stronger swings back/forward are possible. Note that the hub pushes and pulls so will give rotation movements 2 ways. The total will however go one way, just a tiny bit at a time, because as power is generated with light on there is more friction one way (because of the power being generated). So, as there is movement between fork and dynamo hub axle, which you feel by way of the stronger vibrations, the hub will start to rotate.

The experience given does not contradict anything I say on my website, because I say "Tightening the skewer as was suggested to me does not remove the vibrations". This is a fact. This was in response to Peter White who gave me a bullshit story of: "tighten your skewers and the vibrations will be gone". This is not the case, not for any dynamo hub.

If the skewer was left as it was for a little longer, the dynamo plugs would have been in a different position because the hub's axle has been rotating (slowly).

I also say on my web page:

This "Is the hub axle fastened correctly?" is of course not a solution to the general problem of vibration in the handlebar.

So, I don't say it can't be an issue, but I say that it is not a solution to vibration in the handlebar in general. What I did not explicitly state, but readers should be able to determine this from the rest of what I write, is that it can only be an issue when the hub will start to rotate, i.e. as I said, I would have known if this was the case in my tests. So Schmidt could (and should) have said something like "If you have stronger vibrations than usual, you should tighten the skewer because the axle will start to rotate".

Why choose a schmidt hub? Are they not the best?

Lots of people recommend Schmidt hubs, lots of people who ride brevets use them, does that not show they are the hubs to get?:

Schmidt hubs are used a lot by long distance riders, but perhaps largely because they are recommended by others who don't care about this vibration issue, or who don't experience it, or who value the long warrantee, or who feel the need to defend their own purchase (this is a common phenomenon, especially with expensive products; such people will not like what I say as I say Schmidt hubs are not worth it). Or perhaps the people who recommend it still think other dynamo hubs are far inferior to Schmidt's hubs. This is not the case, not in efficiency, and mechanically there is only a small advantage a Schmidt hub has over other dynamo hubs, be they Shimano, Sanyo, but especially SP, where Schmidt only has a little mechanical advantage, but no advantage in efficiency nor in weight... The SON28 has only been getting real competition regarding quality, weight and efficiency since about the time the Shimano DH-3N80 came out, and I think many people still don't know anything other than "If you want a good dynamo hub, get a SON28". But that's 3 years out of date...

Fact is that the SON 28 vibrates more than any other dynamo hub I tried, and people who care about this issue, should not choose that hub.

The only sidenotes I can place with this is that the vibrations depend on bike, fork, tyres etc. But that's exactly what I say on my website. However, as one doesn't know this before mounting a wheel with a specific dynamo hub in one's bike, the only logical course of action seems to me to buy a dynamo hub known to have the least amount of vibrations, to prevent this even becoming an issue (unless you have an option of trying out a hub, or if you ride mostly on roads other than very good asphalt).

Others who may already have experience with dynamo hubs and don't mind this vibration, can use the information I give as a way to determine if that will be an issue for them when they buy a new type of dynamo hub (for the same bike). For example, people who have a Shimano hub and don't like the vibrations shouldn't buy a SON28!

So, what about just using the SONdelux, as that should have less issues with vibrations?

Further regarding SON20R/SONdelux or other 2.4W dynamo hubs used in a big wheel: As I have I noticed with the SP SD-8 (see the review) I couldn't feel vibrations whereas with the PD-8 which is identical in build type, but with 3.0W power output (both 26 poles too), I could. So, this makes a significant difference. Just like the reduction of poles in the SP series 8 from 28 to 26 compared to the PD-7 (HB015) made it significantly more noticeable than the PD-7 (HB015) switchable hub.

But using 2.4W hubs solves one problem at the cost of another: Power at low speed is crap... And the same power you get at 30 km/h with a 3.0W dynamo hub, you may get at 35-40 km/h with a 2.4W hub. This is important for coming use for USB devices or multiple LEDs, as I've tested. So I consider using the SONdelux and other 2.4W hubs in a 622mm wheel a dead-end solution.

To email me go to the email page