Brompton Gears Relative (In)Efficiency

[simonhill]

I have recently bought a 6 speed Brompton. I love it, but have a quick question.

Are any of the gears less efficient than others? I know with Rolhoff that certain gears are less mechanically efficient, so I just wondered if any of the 3 in the SA hub are?

(NB. This is only about mechanical efficiency, I do understand gear tables, etc)

[mjr]

On most three speeds, including the SA AW so I expect the BWR too, middle gear is direct so most efficient, while the others are 2-3% less.

[Brucey]

On most three speeds, including the SA AW so I expect the BWR too, middle gear is direct so most efficient, while the others are 2-3% less.

At best.... In fact it may be a fair bit worse than that on a Brommie with a new, modern AW/BWR (NIG) hub;

a) they are grease lubricated not oil lubricated, which is a bit draggier
b) the NIG design has more built-in parasitic drag; the shift spring bears (via the driver) against the toggle key all the time; to turn the driver (at all) uses 1-2W (I estimate) regardless of the gear that you are in. This ought to get less as the hub wears in, and can be reduced by using better lube and/or a better washer (eg brass, specially made) where the spring bears against the driver. In some hubs the spring rotates with the driver (sliding where the spring collar bears against the end of the cone) and occasionally a hub will fail because the spring snags on the axle and starts to wind/unwind.
c) the BWR hub uses unusually small planet pinions which results in some noise and less efficiency, especially when the hub is new and hasn't been run in yet.
d) the planet pinion bushes in current AW hubs are an incredibly slack fit; the pins are about 6.0mm and the holes in the pinions are nearer 1/4" in size. I think this makes for a draggier bushing, under load.
e) the small wheels won't roll quite as well as larger ones
f) most brommies come with tyres that are not especially quick
g) the small wheels mean the hub is turning faster and this increases certain types of frictional loss.

SA published some test results in about 1908 (or something ) which claimed that the losses were of the order of 1.5% in gear one and a bit less than that in gear 3. But that was when they had ball bearings on the planet pinions, and non-one has measured such low losses since then.

cheers

[NetworkMan]

I looked in Bicycling Science 3rd edition but it looks as if it's a Sachs not an SA since it says "source: Fichtel & Sachs 1987, Keller 1983"

For what it's worth. e.g. at 100 watts
Single speed: 97.3%, Low: 92.8%, Direct 95.7%, High 90.9%

Did Sachs make theirs under license to SA, Brucey?

[Brucey]

I looked in Bicycling Science 3rd edition but it looks as if it's a Sachs not an SA since it says "source: Fichtel & Sachs 1987, Keller 1983"

For what it's worth. e.g. at 100 watts
Single speed: 97.3%, Low: 92.8%, Direct 95.7%, High 90.9%

Did Sachs make theirs under license to SA, Brucey?

Fichtel and Sachs started making hub gears not long after SA did. They have held many patents and (I think) were first to patent or produce several types of hub gear. They did however also produce some hubs under licence from SA.

In 1983 or 1987 they would have been making 3s hubs to their well-proven 'torpedo' design. In essence this is like any other 3s planetary hub gear but it differs in detail; in a SA hub the same pawls drive in high as middle gear but in a F&S torpedo the same pawls drive in low and middle gears. Actual losses vary with parasitic drag (eg whether a coaster brake is fitted or not) and other things like how well made the gears are made; often the gears are more efficient running one way vs the other (i.e. high gear vs low gear) because they mesh better when driving in one direction. [ In a similar way a car gearbox often is quiet when driving forwards but may well whine on the overrun when the loads are reversed.]

IIRC in the IHPVA journal the losses in (well used but freshly oiled) SA and F&S 3s hubs were measured and there wasn't much to choose between them. They were also more efficient (on average) than almost any other type of transmission, barring singlespeeds of course. See

http://ihpva.org/HParchive/PDF/hp52-2001.pdf

Note that the efficiency values quoted therein are not absolute; they are (for various reasons) about 2% down on absolute values; thus the best that could be measured in their apparatus would be around 96%; 2% loss offset due to the equipment, and ~2% loss in the chain alone (as per a singlespeed drive). The direct drive gears in several IGHs measured around 94% or 95%.

Pre NIG SA 3s hubs are liable to be slightly more efficient than NIG ones, BTW.

cheers

[Pneumant]

Broms go best with the tyres pumped up hard (Marathon's 75+psi) and held in 2nd (direct) gear where you can. Reducing the hub sprocket by a couple of teeth also helps customise the direct gear to suit your riding style / terrain. Compared to regular bikes Broms are more effort to ride and definitely don't 'coast' as well. Stop pedalling and you will soon slow down! However they really are good fun to ride and the small wheels accelerate well. I love the mechanical' sound(s) of the SA hub too.

[drossall]

Bromptons are mainly commuting bikes and the like. I have missed the odd train narrowly, but I have never put it down to mechanical losses in the hub on the way to the station.

I do ride my P6R home from work occasionally (37 miles), when the fancy takes me. It's not as quick as if I use a "proper" bike, but if I did that I'd have to ride a few miles extra to get the same amount of time in enjoyment. And I don't think I'd then get up the hills quicker by choosing a more mechanically-efficient gear over one that suits the hill.

[Brucey]

re sturmey archer planetary gear losses. When they only made hubs with ball bearings on the planet pinions they made some efforts to convince punters that "this was the way" vs plain bearings. Later, when they dropped the ball bearings (which were expensive to make) and went to cheaper plain bushings they kept schtum about the change in efficiency.

In 1908 sturmey archer efficiency/losses were measured:
http://www.sturmey-archerheritage.com/files/view-14.pdf
see p7

they claim (BTW the losses are additional to those in the chain anyway (which are around 1.5 to 2% typically if the chain has a straight run and the sprockets are a sensible size));

1.8% and 1.4% losses in gear 3 and gear 1 respectively (with ball bearing pinions) vs
5.1% and 4.5% losses in gear 3 and gear 1 respectively (with plain bushings)

The reality is that under light loads and with the correct lubrication, well-made plain bushings can sustain fairly low losses (as you might expect; they would soon wear if there was metal to metal contact all the time) but they are not as efficient as ball (or roller) bearings are. So I'd suggest that

a) there are parasitic losses in some hubs (due to drag springs etc that give small losses whenever the hub turns, independent of loading (that might be around 1-2W in a typical NIG 3s hub and
b) actual (load dependant) frictional losses in the gear train will vary from about 1.5% to about 5% in a typical, adequately lubricated three speed gear depending on exactly how it is made and the conditions of use.

So if you push out 150W (say) into a NIG hub, you might lose 3W in the chain and sprockets, another 1-2W in parasitic losses, and a further 3% of the remainder, thus total losses of ~9W or about 6% of the total.


More complex IGHs are subject to additional losses because there are (at least) cantilever loadings on stepped planet gears (and larger losses accordingly) and in most cases there are multiple gear trains in series. Each stage might have load dependant losses of (say) 5% and three in a row makes for an efficiency of 0.95x0.95x0.95 = ~0.85 i.e. ~15% load dependant losses in total.

cheers

[simonhill]

Thanks for replies. Second is best, but as ever I'll be using the gear that feels right for the circumstances, while bearing this in mind.

Thanks too for the info about the (frightening) amount of power loss due to all the bits and pieces inside the hub.Nice to have something to blame. Ho, ho.

[Brucey]

Thanks for replies. Second is best, but as ever I'll be using the gear that feels right for the circumstances, while bearing this in mind.

Thanks too for the info about the (frightening) amount of power loss due to all the bits and pieces inside the hub.Nice to have something to blame. Ho, ho.

The tensioning pulleys also eat about 1W and the chain is rendered less efficient by the use of

-small sprockets and
- bad chainlines.

The Brompton 6s system suffers some additional losses that arise because of these things but such losses are considerably smaller than those which arise in full-blown derailleur systems.

cheers