kyselad

I'm wondering whether anyone has seen drag stats for the Sturmey Archer X-FDD?? The only good numbers I've managed to find online (in this VBQ article) don't include the X-FDD and show *much* greater drag with the lights off for the older (HB-NX*) vs newer (DH-3N*) Shimano models. In fact, on the older Shimanos, the drag barely changes regardless of whether this light is switched on or off (!). I read a post on another forum suggesting the lights-off drag of the X-FDD might be comparable to the older Shimano models, but no solid data to support this comparison. Supposedly, Shimano and SA hubs use the same Sanyo internals, but I can't tell which Shimano internals (the not-so-great Nexus or the newer DH-3N*) are supposedly comparable.

Anybody happen to know for sure about the drag and/or the internals? I really like the features on the SA hub, but if it really does drag quite a bit at high speeds when disengaged, that might be a deal-breaker.

kyselad

Another thought -- how are drag measurements performed? I know the Shimano hubs saw an improvement to "Ultegra-level" parts, but wouldn't that performance upgrade mostly result from better bearings? Is it possible that the high drag encountered in the older Shimano (and possibly the SA X-FDD) hubs when turned off has to do with shoddy bearings, or would this not have such a noticeable impact? If the drag in practice turned out to be more than I could handle, replacing them seems like no big deal, particularly since Sunrace/SA provides detailed schematics.

Fenway

I use that hub on a daily basis and honestly don't notice any drag even with the lights on. Mind you this is with 38mm tires which already have some rolling resistance on an upright hi-ten bicycle weighing 35lbs+, so your experience may vary.

It is my understanding that the generator part of the hub in both the Archer and Shimano Hubs are made by Sanyo, so the only real difference aside from the addition of a drum brake mechanism, may in fact be the quality of the bearings. Perhaps a mechanic has opened a hub to do a comparison at some point?

kyselad

Good to know. I plan to put it on a more performance-oriented build, but nothing race-ready -- it's a 650B bike that I hope will serve double-duty as a commuter and a moderately nimble weekend ride. It will certainly come in a few pound lighter, so we'll see if the drag proves noticeable. If so, I'll try a bearing swap and report back.

tatfiend 

Per my understanding the decreased drag on the newer higher end Shimano hubs is due to different dynamo internals, not bearings. Bearings with that much drag would self destruct in almost no time.

kyselad

Hm, if the dynamo internals do indeed differ, then the recurring (but seemingly unsubstantiated) rumors of equivalent Sanyo internals in SA and Shimano hubs doesn't hold water. Perhaps this was true on the older, higher drag Shimanos but not the newer models?

tatfiend 

I do not know but the owner/wheelbuilder at Longleaf Bicycles carries the Sanyo dynamo hub that Peter White Cycles imports and reports that based on feel while building wheels with it, it has similar drag to the Novatec or older Shimano NX30 hubs, considerably more than the current Shimano or SON high end dynamo hubs.

JimBeans83

This is a photo from eurobike past year , i too find it hard to believe the 7 watts resistance unloaded for the x-fdd. It spins free in the wheel when unloaded - how could it be using 7 watts? Where is that heat or energy going?

JimBeans83

ps - photo from SON booth , what's strange about their bar graphs is that the color for x-fdd changes from graph to graph, maybe this is an error on their part?

Any more info found from others on the hub?

davidad

Ignore the idea of bearings causing drag. They are not a factor in what we do.

JimBeans83

I'm not so sure that assumption can be made - yes it's published that other hubs have a ~0.2 W loss for bearings, but the fact that the charts discussing SA X-fdd show same resistance regardless of the presence of electrical load or not forces one to also consider all non-electrical components of the system.

It would be interesting to pull of the drum pads and repeat the test, but of course it is not so easy to do a loaded bearing proper test for these. Schmidt has a complicated apparatus to test this.

davidad

A 700 bike wheel at 15mph is turning 168rpm. High quality bronze bushings would be good enough for us.

JimBeans83

But "good enough" means what, quantitatively?

7W loss at 30km/h is good enough for me- I don't care, but the reports are comparing numbers. And at the top of the efficiency list the differences are very small, and very difficult to compare (1/2% is the difference between 50 and 200 Euros).

davidad

Read "Bicycle Science" and get back to us with an estimate of bearing loss.

JimBeans83

Sorry, the burden of substantiation for those seeking credibility rests on the one with "gut feelings" or subjectivity. That would be you here.

By the way,

https://www.enhydralutris.de/Fahrrad/...leuchtung.html

implies short circuiting the X-FDD instead of leaving it open will bring it within 1/2 watt of other competitors for "off" resistance. (source : link above / SON published efficiencies)

davidad

Sorry, I violated rule number 2.