Rohloff makes the case for Cyber Nexus

On Apr 28, 11:22*am, wrote:
Clive George wrote:
Of course it calculates effort, to compare it to the effort level
chosen as a target, but it measures effort indirectly.
In which way?
There's a speed sensor in the dedicated hub dynamo of Shimano's
Cyber Nexus gruppo, and an inclination sensor somewhere.
Is there really an inclination sensor in there? I'd be interested to
see mention of it on a Shimano manual/website.

Deriving inclination on a vehicle moving at varying speed is an
illusive goal, gravitational and vehicle acceleration not being
separable. *That is why assessing the gradient of a road is not
readily measured as users of altimeters that do that by averaging
climb per sampled distance. *It doesn't work well, but better than
attempting it with an inclinometer that acts more like a cadence
pendulum than an inclinometer.

Jobst Brandt

But hill inclination and vehicle accelerations are quite separable, by
filtering in software. Vehicle accelerations have no DC component to
speak of. Today's solid state accelerometers are stable enough to make
it easy to tell these signals apart. It is part of the strategy used
by the iBike power meter, for example.

-pm

On 4/28/2009 11:10 AM Andre Jute wrote:

On Apr 28, 6:31�pm, Helmut Springer wrote:
Andre Jute wrote:
Speed, inclination and gear together give effort, indirectly as I
said.
...only if you neglect wind resistance.

--
MfG/Best regards
helmut springer � � � � � � � � � � � � � � � � � � � � � panta rhei

If there is a headwind, the cyclist slows down and the CPU puts him in
a lower gear so that his effort is the optimum he selected. If there
is a tailwind the cyclist travels faster and CPU puts him in a higher
gear and his effort again is the optimum he targets. This is exactly
the same concept as a racing cyclist changing gears to maintain his
cadence, which is his optimum energy expenditure.

Well, regardless of how it gets its smarts, it sounds like it might be fun
to ride a bike with one of these things.

--
Mike "Rocket J Squirrel"
Bend, Oregon

On Apr 28, 10:22*am, Andre Jute wrote:
On Apr 28, 2:29*pm, Bernhard Agthe wrote:

Hi,

Andre Jute wrote:
Well, no, it's a pretty obvious conclusion. The human will always be
seconds slower on each gearchange than electronics can be.

You know, I can even shift beforehand - that is, when I approach a hill,
I can shift (manually) just before I hit the grade and be in the perfect
gear right away ;-)

That's even better than automatic shifting ;-)

Sorry to disappoint you, Bernard, but if you change "beforehand" there
will be several meters where you and the bike do not operate at
maximum efficiency. *No human an react as fast as electronics.

The electronic system can *only* react, and whatsmore, it only reacts
to the rider's change in cadence (i.e. it reacts to the rider's
reaction to the terrain, making it necessarily slower than an
attentive rider). A human rider sees what is coming down the road and
can anticipate shifting at the optimal moment, accounting for the
mechanical time it takes to shift, so that the shift is completed
before an electronic system with its superhuman reactions has anything
in its sensors to react to.

A human can also do things like upshift at the moment he decides to go
from sitting pedaling to standing, so that the first standing stroke
is in the appropriate gear -- while the electronics can't know when
the rider chooses to switch pedaling styles, and have no capaiblity to
distinguish sitting from standing anyway, so will choose an
inappropriately high cadence for standing.

So I don't see autoshifting taking hold in the racing world. Everyday
utilitarian cycling, perhaps, and I think Shimano are following the
lessons learned from indexed shifting -- if you introduce the tech at
the utilitarian end, as Positron was, it is perceived as low end. Make
something for the high end -- electronically assisted manual shifting,
in this case -- and automatic/manual electronics might sell down the
range.

-pm

Andre Jute wrote:

Sorry to disappoint you, Bernard, but if you change "beforehand" there
will be several meters where you and the bike do not operate at
maximum efficiency. No human an react as fast as electronics.

You don't mention the need for the human to react to the
electronics, which can be considerably slower than mechanicals
reacting to human command in a manual system.

More than anything else, what I found off-putting about my
automatic Nexus hub was its changing gear in traffic when I
didn't want it to.

A common situation involves crossing an intersection, where one
must climb up the camber of the road being crossed. And then
accelerating briefly down the other side once over that crest.

In that situation, I often want to reliably apply a carefully
controlled burst of power. But the Nexus frequently insisted on
its two gear changes (as I slowed then sped up over the crest).
The resulting cadence change, forced and unwanted, detracted
from my ability to cross the intersection with maximum control
and safety.

John

On Apr 28, 9:19*pm, pm wrote:

But hill inclination and vehicle accelerations are quite separable, by
filtering in software. Vehicle accelerations have no DC component to
speak of. Today's solid state accelerometers are stable enough to make
it easy to tell these signals apart. It is part of the strategy used
by the iBike power meter, for example.

-pm

Are you saying that inclination is derived from speed by the CPU?

If that is true, if there is no direct inclination sensor, that
software must work amazingly well because my Cyber Nexus locks out the
suspension with utmost reliability, always on hills, never when
travelling at the same speed on the level. If it does all that merely
from the speed, all I can say is Wow! Arthur C Clarke said that any
sufficiently advanced technology looks like magic...

Andre Jute
Visit Andre's recipes:
http://www.audio-talk.co.uk/fiultra/FOOD.html

On Apr 28, 9:22*pm, Mike Rocket J Squirrel
wrote:
On 4/28/2009 11:10 AM Andre Jute wrote:



On Apr 28, 6:31 pm, Helmut Springer wrote:
Andre Jute wrote:
Speed, inclination and gear together give effort, indirectly as I
said.
...only if you neglect wind resistance.

--
MfG/Best regards
helmut springer panta rhei

If there is a headwind, the cyclist slows down and the CPU puts him in
a lower gear so that his effort is the optimum he selected. If there
is a tailwind the cyclist travels faster and CPU puts him in a higher
gear and his effort again is the optimum he targets. This is exactly
the same concept as a racing cyclist changing gears to maintain his
cadence, which is his optimum energy expenditure.

Well, regardless of how it gets its smarts, it sounds like it might be fun
to ride a bike with one of these things.

I hated derailleurs: obstructive, dirty, fragile, nasty, grinding
leftovers from a more brutal age. Hub gears are the coming thing, and
the best hub gears are definitely automatic. (I haven't tried
Fallbrook's NuVinci CVT.)

But it isn't about fun, really. After a while you don't notice that
you don't change gears. You just treat the bike like a singlespeed.
You get on and ride, and the electronics do the rest perfectly
unobtrusively. The girlfriend of one the local fixie boys said to me,
"You must have hamstrings from hell to ride a one-gear bike up that
hill." She was quite disappointed to hear I had the assistance of Mr
Shimano's autobox.

Andre Jute
Visit Jute on Bicycles at
http://www.audio-talk.co.uk/fiultra/...20CYCLING.html

On Apr 28, 9:26*pm, pm wrote:
On Apr 28, 10:22*am, Andre Jute wrote:



On Apr 28, 2:29*pm, Bernhard Agthe wrote:

Hi,

Andre Jute wrote:
Well, no, it's a pretty obvious conclusion. The human will always be
seconds slower on each gearchange than electronics can be.

You know, I can even shift beforehand - that is, when I approach a hill,
I can shift (manually) just before I hit the grade and be in the perfect
gear right away ;-)

That's even better than automatic shifting ;-)

Sorry to disappoint you, Bernard, but if you change "beforehand" there
will be several meters where you and the bike do not operate at
maximum efficiency. *No human an react as fast as electronics.

The electronic system can *only* react, and whatsmore, it only reacts
to the rider's change in cadence (i.e. it reacts to the rider's
reaction to the terrain, making it necessarily slower than an
attentive rider). A human rider sees what is coming down the road and
can anticipate shifting at the optimal moment, accounting for the
mechanical time it takes to shift, so that the shift is completed
before an electronic system with its superhuman reactions has anything
in its sensors to react to.

A human can also do things like upshift at the moment he decides to go
from sitting pedaling to standing, so that the first standing stroke
is in the appropriate gear -- while the electronics can't know when
the rider chooses to switch pedaling styles, and have no capaiblity to
distinguish sitting from standing anyway, so will choose an
inappropriately high cadence for standing.

Okay, maybe it only works faster than I do, and you will beat it every
time. But I ride with a wide variety of people, and on my Cyber Nexus
the automatic gearchanges are without fail more suitably implemented
than their manual ones, regardless of who they are or what their
experience or outlook

So I don't see autoshifting taking hold in the racing world. Everyday
utilitarian cycling, perhaps, and I think Shimano are following the
lessons learned from indexed shifting -- if you introduce the tech at
the utilitarian end, as Positron was, it is perceived as low end. Make
something for the high end -- electronically assisted manual shifting,
in this case -- and automatic/manual electronics might sell down the
range.

If that is the case -- and I've been saying the same thing for years
-- Shimano has already dropped the ball by first launching full
automatics for city/commuter/touring bikes years before they launched
the cut down Cyber Nexus for racing bikes. The general market moment
for high-end autoboxes came and went, and left only a small niche
market of gearheads. Even if the Dura-Ace electronically assisted
manuals take off, the general market has looked and shrugged and just
about turned its back on high end automatics. (Not the emphasis on
high-end -- I have no knowledge of the low-end autos in the Lime class
-- maybe they're taking the world by storm?)

Andre Jute
Visit Jute on Bicycles at
http://www.audio-talk.co.uk/fiultra/...20CYCLING.html

On 28 Apr 2009 20:49:31 GMT, John Henderson
wrote:

Andre Jute wrote:

Sorry to disappoint you, Bernard, but if you change "beforehand" there
will be several meters where you and the bike do not operate at
maximum efficiency. No human an react as fast as electronics.

You don't mention the need for the human to react to the
electronics, which can be considerably slower than mechanicals
reacting to human command in a manual system.

More than anything else, what I found off-putting about my
automatic Nexus hub was its changing gear in traffic when I
didn't want it to.

A common situation involves crossing an intersection, where one
must climb up the camber of the road being crossed. And then
accelerating briefly down the other side once over that crest.

In that situation, I often want to reliably apply a carefully
controlled burst of power. But the Nexus frequently insisted on
its two gear changes (as I slowed then sped up over the crest).
The resulting cadence change, forced and unwanted, detracted
from my ability to cross the intersection with maximum control
and safety.

John

Dear John,

That's what Shimano warns users about:

"This system is an automatic gear shifting system in which the
shifting points are determined by the speed of the bicycle (the
rotating speed of the wheels). Because of this, gear shifting can
occur independently of the rider's intentions, and so shocks may
be felt during shifting. Make sure that you understand and
become accustomed to these shifting characteristics before
using the system extensively."

http://cycle.shimano-eu.com/media/te...9830605819.pdf
or http://tinyurl.com/chumtd

Cheers,

Carl Fogel

On Apr 28, 9:49*pm, John Henderson wrote:
Andre Jute wrote:
Sorry to disappoint you, Bernard, but if you change "beforehand" there
will be several meters where you and the bike do not operate at
maximum efficiency. *No human an react as fast as electronics.

You don't mention the need for the human to react to the
electronics, which can be considerably slower than mechanicals
reacting to human command in a manual system.

Never bothered me. I always found my Cyber Nexus in the right gear for
what I wanted to do, and very quickreacting when I changed my mind in
mid-manoeuvre -- a damn sight faster than I could turn the rotary
control on the manual version of the same Nexus 8 speed box I also
have, vastly faster than I could ever change derailleur gears.

More than anything else, what I found off-putting about my
automatic Nexus hub was its changing gear in traffic when I
didn't want it to.

As I say, I just didn't find that. There was never any hesitation in
the operation. Perhaps the 8 speed auto has more sophisticated
electronics than 4 speed auto; it is after all around a decade
younger. In the beginning, at a couple of dangerous, busy crossings, I
would switch on the manual switching and leave it in a high gear to
get quick acceleration off the mark, but I soon found that just
pedalling forcefully would snap the box through the gears in a quarter-
turn of the crank. I helped that initial acceleration along by
generally keeping the box in the sporting mode which excludes the
lowest gear and starts it off in second. (Three modes and manual:
little old lady, normal, sporting, instantly available under your
thumb.)

A common situation involves crossing an intersection, where one
must climb up the camber of the road being crossed. *And then
accelerating briefly down the other side once over that crest.

In that situation, I often want to reliably apply a carefully
controlled burst of power. But the Nexus frequently insisted on
its two gear changes (as I slowed then sped up over the crest).
The resulting cadence change, forced and unwanted, *detracted
from my ability to cross the intersection with maximum control
and safety.

That would terminate the gearbox for me. I feel that way about
derailleurs, that they're just crouching there waiting to miss the
change to expose me to an onrushing truck. But I haven't felt that way
since I changed to hub gearboxes, first manual and then auto by
Shimano, then Rohloff. My current problem with remembering to reset
the Rohloff gearbox between coming home and riding out in traffic in
an ultra-low, ultra-slow gear is a learning situation, something under
my control, not an intrinsic fault of the gearbox.

****
I guess there is also an element of attitude. Did you come to hub
gearboxes from years of riding sporting bikes? I arrived from years of
daily hating derailleurs as inelegant engineering. Bound to influence
one's enthusiasm for hub gears and auto boxes.

Andre Jute
Visit Jute on Bicycles at
http://www.audio-talk.co.uk/fiultra/...20CYCLING.html

aka Jobst Brandt wrote:
Clive George wrote:

Of course it calculates effort, to compare it to the effort level
chosen as a target, but it measures effort indirectly.

In which way?

There's a speed sensor in the dedicated hub dynamo of Shimano's
Cyber Nexus gruppo, and an inclination sensor somewhere.

Is there really an inclination sensor in there? I'd be interested to
see mention of it on a Shimano manual/website.

Deriving inclination on a vehicle moving at varying speed is an
illusive goal, gravitational and vehicle acceleration not being
separable. That is why assessing the gradient of a road is not
readily measured as users of altimeters that do that by averaging
climb per sampled distance. It doesn't work well, but better than
attempting it with an inclinometer that acts more like a cadence
pendulum than an inclinometer.

One could always use a gyroscope to determine inclination - see
"artificial horizon" in aircraft. Not economical or practical for a
bicycle, however.

--
Tom Sherman - 42.435731,-83.985007
LOCAL CACTUS EATS CYCLIST - datakoll