Power Meters

A light bike doesn't replace good fitness.

Moderator: Moderator Team

User avatar
Bruiser
Posts: 1385
Joined: Mon Sep 30, 2002 1:59 am
Location: Sydney, Australia
Contact:

by Bruiser

Very interesting.

Does anyone any experience with one?

They claim to measure:

Gravity (hill climbing)
Wind resistance
Inertia
Rolling resistance


How do they measure rolling resistance and inertia or even gravity?

Can you callibrate the ibike for your bike/setup?

by Weenie


Visit starbike.com Online Retailer for HighEnd cycling components
Great Prices ✓    Broad Selection ✓    Worldwide Delivery ✓

www.starbike.com



redbiker
Posts: 93
Joined: Tue Aug 30, 2005 12:25 pm
Location: UK
Contact:

by redbiker

I got a Powertap SL built onto a Mavic Open Pro for my training bike.

Only had it a couple of weeks, but so far it has been very good. Even rode it in the pouring rain yesterday for 3 hours with no problems.

I think if money is no object then SRM is the way to go, otherwise the Powertap is a very good system. I'll have to wait and see if I experience any of the reported problems.

User avatar
wdbike.com
Posts: 72
Joined: Mon Mar 14, 2005 4:11 pm

by wdbike.com

Bruiser wrote:Very interesting.

Does anyone any experience with one?

They claim to measure:

Gravity (hill climbing)
Wind resistance
Inertia
Rolling resistance


How do they measure rolling resistance and inertia or even gravity?

Can you callibrate the ibike for your bike/setup?


Nobody has any experience yet because only the prototypes have been made ... But their design is very clever. They measure gravity through an accelerometer, which is calibrated simply by facing a bike in one direction, pressing a button, then doing exactly the same thing 180 degrees away. One direction is the uphill, the other down and it can work out the percentage accordingly - and exceptionally accurately. This calibration can be done whenever the mount moves. The accelerometer will also measure inertia, obviously.

Rolling resistance they calibrate using a roll-down. This is done at the beginning of a ride by coasting from 20mph to 5 without braking or pedaling and from it they get a frontal area coefficient and rolling resistance (just by internally fitting a curve). Naturally the drawback of this is that the frontal area measurement will change if you change position, less so if the road surface varies. But in that position (drops for a race, of course) it should be very very accurate. And they claim that it's going to be flexible enough that future firmware updates will be available allowing one to record multiple profiles and manually switch between them while riding. I personally think the latter will be crucial to their success among serious power devotees, but for me just being able to race with power without changing my wheel, crank, or bb to something heavier or more awkward than i would normally ride is going to be fantastic.

John979
Posts: 1046
Joined: Sat Jan 15, 2005 11:29 pm

by John979

The iBike power meter was not very well received by the “wattage” community where is not even viewed as a power meter, rather, its viewed as a sophisticated bicycle computer that uses a variety of measurements to estimate power, which is all it does. The only “wattage weenies” I know of interested in purchasing this product are doing so out of gadget curiosity, not because of its functionality.

While I see on this thread Polar getting dissed again, for many it is an excellent product choice, especially for those already owning a compatible Polar computer.
John979

User avatar
wdbike.com
Posts: 72
Joined: Mon Mar 14, 2005 4:11 pm

by wdbike.com

John979 wrote:The iBike power meter was not very well received by the “wattage” community where is not even viewed as a power meter, rather, its viewed as a sophisticated bicycle computer that uses a variety of measurements to estimate power, which is all it does. The only “wattage weenies” I know of interested in purchasing this product are doing so out of gadget curiosity, not because of its functionality.

While I see on this thread Polar getting dissed again, for many it is an excellent product choice, especially for those already owning a compatible Polar computer.

IMO use of the past tense ("was not very well received") isn't terribly appropriate here, since nobody in the wattage community has actually ridden an iBike - to my knowledge, anyway. The company itself is just finishing its testing, and products are generally not released to *anyone* before something like that happens. So to say the wattage community has a view of it at all seems like a significant overstatement.

I might also point out that there is no such thing as directly measuring power; all power devices estimate it from a series of measurements that correlate more-or-less well to power output (which is just energy output per unit time). Most of those in use today use fitted data from deflection of metal components under stress, which is relatively well-understood in the stress ranges humans are capable of applying.

I *would* say that the wattage communities I've looked at have been skeptical of the iBike, which is healthy. I'm skeptical too, but I like the idea and the theory is just as sound as estimating power through deflection measurements in a piece of metal. In fact, in a certain sense it's a more direct measurement of things, since the iBike is directly measuring the forces a rider is working against. My only real skepticism about the iBike is that it might not be flexible enough for most real-world training situations, ie. when the road surface changes or when position changes.

Anyway, it's cheap enough that I'm going to try it. But then, I'm sure I can sell mine off if it's complete crap and at least make what I paid for it back, which I will freely admit is a luxury most don't have.

Finally, a number of our team riders have the Polar. None of them like it because it's difficult to set up properly and even then relies on too many factors to estimate power. This alone means the errors are huge - never mind the fact that the equations involved with estimating massy string tensions are nonlinear and very sensitive to small measurement errors, bumps in the road, chaos effects at high tensions, etc. Its principal advantage is that it's light enough to use in races, so I was going to borrow one for use in a 40k time trial I did a few months ago. But I was convinced by the owner not to use it because it would be "completely useless for gauging efforts in real time." He said he only trusts it to give him an idea of his relative efforts in a race.

John979
Posts: 1046
Joined: Sat Jan 15, 2005 11:29 pm

by John979

wdbike.com:

Yes, it is true nobody to my knowledge has yet ridden with an iBike or at least any Wattage Forum members, but this does not keep this very learned and credentialed group from asking valid questions and making pertinent comments. Extraordinary claims require extraordinary evidence, something not supplied by the iBike's manufacturer: not even +/- accuracy is provided on its website.

Personally, I was very excited to learn of a new “power meter” and initially viewed the iBike with enthusiasm and an open mind, until I quickly realized it is nothing more than a glorified Ciclosport. The iBike cannot be considered a true power meter, primarily because its approach depends upon a constant rider + bicycle weight (an input), a generalized CdA and a constant barometric pressure.

While you are correct in stating that no power meter directly measures power (SRMs and PowerTaps use stain gauges to measure force) this comment obfuscates the issue at hand, as virtually all measurements are indirect. I find it very interesting that you discredit Polar, which uses an indirect approach (sensing chain frequency and convert to power) that is more similar to iBike’s approach than SRM's or PowerTap's, which use stain gauges to measure force, which is then converted to power in a relatively direct fashion.

I agree Polar has its issues, but a properly installed and calibrated Polar PM is far more useful than you are giving credit. I have both a Powertap and Polar and find the Polar to be consistent over a wide range of applications. Interestingly, one area I find the Polar useful is for gauging a time trial, especially one over moderately varying terrain whereby gearing bias, a documented concern, is avoided. I challenge you or your rider to provide documented data from a properly installed Polar PM supporting contention a Polar PM is useless for gauging a TT effort.
John979

User avatar
wdbike.com
Posts: 72
Joined: Mon Mar 14, 2005 4:11 pm

by wdbike.com

John979 wrote:wdbike.com:

Yes, it is true nobody to my knowledge has yet ridden with an iBike or at least any Wattage Forum members, but this does not keep this very learned and credentialed group from asking valid questions and making pertinent comments. Extraordinary claims require extraordinary evidence, something not supplied by the iBike's manufacturer: not even +/- accuracy is provided on its website.

I agree. But the company hasn't done any testing yet, so how can they post error bars?

As for the wattage community at large, "learned" is a great thing, but there are a great many learned people out there who get things wrong, especially when they make judgements on an entirely new paradigm without the benefit of data. This is not meant to be insulting, it's a fact that anyone who is sufficiently learned to be listened to should readily admit.

Now "credentialed," on the other hand ... I find credentialed to be a highly overrated measure of someone's competence.
Personally, I was very excited to learn of a new “power meter” and initially viewed the iBike with enthusiasm and an open mind, until I quickly realized it is nothing more than a glorified Ciclosport. The iBike cannot be considered a true power meter, primarily because its approach depends upon a constant rider + bicycle weight (an input), a generalized CdA and a constant barometric pressure.

It doesn't rely upon a constant barometric pressure at all. The constant weight and CdA, yes, which are not unreasonable assumptions if one maintains a similar riding position and doesn't pee on one's bike. But the barometric pressure gauge it contains (which I note measures a *non*constant pressure) is for altitude measurements - and the altitude measurements are not used in the power calculation. Slope measurements from the accelerometer are.

The force a rider must apply to offset wind drag is not inferred by measuring the momentum changes in the air at the front of the bike, it's a *linear* function of those changes (after all, it's the air you're pushing against, and whatever resistance is there is the resistance you feel). Linear functions produce less error than nonlinear because of the fractal effect, but usually one has to look pretty carefully to find linear situations. Except for the fitting of the CdA, this is actually the closest anyone's come to a direct measurement of rider resistance. Which is not to say anything about its accuracy, of course! but it is pretty direct.

The source of error I see here that could potentially cause major problems is the issue of turbulent eddies, which could cause different wind resistances to occur on the small scales represented by the iBike's input chamber than those that occur across a rider's frontal area.

If you don't consider the iBike to be a true power meter, then I would honestly like to hear your definition of the term.
While you are correct in stating that no power meter directly measures power (SRMs and PowerTaps use stain gauges to measure force) this comment obfuscates the issue at hand, as virtually all measurements are indirect. I find it very interesting that you discredit Polar, which uses an indirect approach (sensing chain frequency and convert to power) that is more similar to iBike’s approach than SRM's or PowerTap's, which use stain gauges to measure force, which is then converted to power in a relatively direct fashion.

I don't think it obfuscates the situation. Rather I think it's a matter of marketing that people think that anyone actually measures power directly. The measures employed by the major PMs are very much *not* direct, they rely on piezoelectric currents caused by the flexing of materials and convert those signals using curves they have fit. And the curves do vary with temperature and pressure, which is why one has to zero them every so often. I will grant you that the moving target they fit moves less quickly than CdA, but the fact that iBike has recommended frequent calibration indicates to me that they at least acknowledge this and have at least tried to develop a solution.

Polar's monitor relies on something much more difficult to quantify, chain rattle. This is a very nonlinear function of power and is therefore a very error-prone method. Not only that, but their solution for measuring altitude gain (and therefore slope) is based on barometric pressure, which is a horrible method because the atmosphere is almost never "hydrostatic" (which, you may know, is the name for the assumption all of those things make). I once used an altimeter in a MTB race up Pike's Peak and ended up lower than where I started.

Our rider with the Polar does not have a PT, and you will forgive me if I don't set up a controlled experiment for the sake of an argument on a message board! I will say that he is ordering an iBike, but he's getting a pro-deal so it's not too expensive ... I will post results from my iBike (when I get it) against my PT; but until there's actual data to look at, perhaps we would both be better off getting some work done instead of continuing this discussion at this level of detail? :wink:

User avatar
ras11
Posts: 851
Joined: Thu Sep 25, 2003 9:11 pm
Location: SC, USA

by ras11

I'm seconding John979's opinion regarding the Polar and the iBike. Although maybe John979 was going a little overboard with his criticism. I highly doubt iBike would be very accurate. If it were it would be selling at 10x the price. I also think it's a good estimation, but not a direct measurement. I think this would be perfect equipment for Jrs and novice riders who don't need the higher resolution that Polar/PowerTap/SRM give.

FYI, PT is coming out with a new SL; June 2006. It's supposed to be totally wireless and USB with a $300 premium atop of the SL price. It'll also come with a coded HR strap (no more data drops!!!).

If only they would include altitude, temperature, and a little color graphing screen with a GPS map, that would be "the" perfect computer. Oh.. and might as well include a bluetooth headset for the race radio and downloading files. And might as well add in an emergency cell phone... and run off solar power.. and not cost an arm and a leg. Maybe just an arm.
:-) Toys-R-Us

User avatar
Bruiser
Posts: 1385
Joined: Mon Sep 30, 2002 1:59 am
Location: Sydney, Australia
Contact:

by Bruiser

Re: Ibike, I'm concerned about the calibration of rolling restistance, in particular for when the road surface changes in quality during a ride.

User avatar
wdbike.com
Posts: 72
Joined: Mon Mar 14, 2005 4:11 pm

by wdbike.com

I personally doubt that iBike will replace PTs or SRMs as "the standard" in power measurement. I don't see how it could be more accurate than either of them, which have relatively easily quantifiable sources of error and which have established the power market itself. Or Ergomo, for that matter. In that sense, I think iBike's marketing hype is getting in their way.

However, I will make the prediction right now that it will be more accurate (in general) than Polar. I believe its method is more robust on average. In certain specific situations, like where the asphalt changes mid-ride, it could be less accurate; but remember the Polar doesn't do well in the presence of external vibrations either. And unless it's a huge change in the surface (like smooth pavement to dirt), I don't think the error will be that large. I suspect the larger source of error will be the use of different positions.

I personally don't like racing with a PT (too heavy), and an SRM or Ergomo is way too expensive, considering I have spent way too much money on my bike as it is. Plus I don't like the way an SRM looks ... at 58g, the iBike for me will be a racing tool, and I think that's where the company should try for their niche. They may be already, since the engineer at Interbike was telling me about a snorkel system they've designed for time trial bikes to minimize the pooling effect and backflow of air in the chest area.

John979
Posts: 1046
Joined: Sat Jan 15, 2005 11:29 pm

by John979

wdbike.com wrote:I agree. But the company hasn't done any testing yet, so how can they post error bars?


If I were introducing such a product, especially one promoting a "paradigm" shift, I would first fully document its capabilities instead of presenting a glowing press release.

As for the wattage community at large, "learned" is a great thing, but there are a great many learned people out there who get things wrong, especially when they make judgments on an entirely new paradigm without the benefit of data. This is not meant to be insulting, it's a fact that anyone who is sufficiently learned to be listened to should readily admit.
The whole lot? Please. In the Wattage Forum, not a single sports scientist or trainer of note has been impressed with this product. These people are smart, learned and experienced enough to comprehend the underlying concept and are not impressed, other than the iBike Pro simply being a more advanced Cyclosport. Current discussion of this product is NIL. That says it all.

Now "credentialed," on the other hand ... I find credentialed to be a highly overrated measure of someone's competence.
Such commentary does not reflect well of you.

It doesn't rely upon a constant barometric pressure at all. The constant weight and CdA, yes, which are not unreasonable assumptions if one maintains a similar riding position and doesn't pee on one's bike. But the barometric pressure gauge it contains (which I note measures a *non*constant pressure) is for altitude measurements - and the altitude measurements are not used in the power calculation.
The iBike Pro appears to employ a venturi anemometer to measure windspeed. These devices are affected by both barometric pressure and wind direction.

Except for the fitting of the CdA
Except for? This is the underlying problem with the iBike and the reason its not a powermeter. IT IS A POWER ESTIMATOR. Go play with analyticcycling.com and see what "estimating" CdA does...

If you don't consider the iBike to be a true power meter, then I would honestly like to hear your definition of the term.
SRM, Powertap, Ergomo, Polar.
The measures employed by the major PMs are very much *not* direct, they rely on piezoelectric currents
Doesn't the iBike's accelerometer depend upon a piezoelectric (resistive) sensor? Probably the most accurate thing it does too.

Not only that, but their solution for measuring altitude gain (and therefore slope) is based on barometric pressure...
Dissing Polar again. Doesn't iBike use the same method via its barometric pressure port?

Summary:

iBike Pro's ESTIMATION of power is dynamically affected by any of the following:

1) Change in position
2) Change in weight
3) Change in tires or road surface
4) Change in barometric pressure (in vortex chamber) due to a front, elevation, or temperature change.
5) Drafting
6) It does not include a heart rate monitoring function, this must be factored into the value (and weight) equation.

As such, iBike Pro is not only inaccurate, but non-repeatable. Repeatability actually being a more important factor than absolute accuracy.

At this point, I am going to recuse myself from this discussion, as I suspect you have some relationship to the company, given you are its only defender as a powermeter vis-a-vis the aforementioned brands. :roll:
John979

User avatar
wdbike.com
Posts: 72
Joined: Mon Mar 14, 2005 4:11 pm

by wdbike.com

Holy s**t, John, you sure know how to get emotional about a power meter. My apologies, straight-off, for defending something you seem bent on discrediting ... again, I repeat, without data.

For the record I don't work for their company or have any affiliation with them. I may sell them at some future date, but that depends upon a. their testing providing good results and b. my testing providing good results. I don't want to sell an item I don't think is worth buying.

I don't know why disagreements on message boards always have to get personal. If you have some affiliation with Polar and I am thereby insulting you by expressing my dissatisfaction with it, I apologize.

As for the rest:
As for the wattage community at large, "learned" is a great thing, but there are a great many learned people out there who get things wrong, especially when they make judgments on an entirely new paradigm without the benefit of data. This is not meant to be insulting, it's a fact that anyone who is sufficiently learned to be listened to should readily admit.
The whole lot? Please. In the Wattage Forum, not a single sports scientist or trainer of note has been impressed with this product. These people are smart, learned and experienced enough to comprehend the underlying concept and are not impressed, other than the iBike Pro simply being a more advanced Cyclosport. Current discussion of this product is NIL. That says it all.
Now "credentialed," on the other hand ... I find credentialed to be a highly overrated measure of someone's competence.
Such commentary does not reflect well of you.

I stand by my statements, and I find it unfortunate that you can make a judgement of me because I am not impressed by credentials. I find it even more so that you would malign my character because I don't believe in credibility by diploma?!

The first statement above is simply true and is not intended to be denigrating to anyone. A learned person should not make categorical statements without something to back themselves up, though in my work I see it happen very frequently. I review a lot of papers that make this mistake, and it's frustrating to see such things appear in the peer-reviewed literature regardless. Since a wattage forum is anything but peer-reviewed, its results are bound to be that much worse. Sorry. And no, I don't consider a coach qualified to make a judgement on something he cannot compare to anything currently in existence! I wouldn't even trust a scientist to make such a statement.

The second statement has two facets. One: a message board is a convenient place to make up credentials. Who could or would check? And two: I prefer to listen to people who rely more upon their work than their credentials, eg. "I graduated summa cum laude from Harvard" means a lot less to me than "I investigated the relationship between power output and nutritition in 250 athletes who participated in a study for a period of two years."

Since you're so bent on credentials, I might as well tell you I have a PhD in physics and currently work as a research scientist, with a joint appointment at the CIRES Institute at the University of Colorado in Boulder and Colorado Research Associates. My personal webpage is here, if you want something to back that up. You'll note that I'm actually relatively well-qualified to discuss this topic.

Please note that I am only one of two people who post on this account.
The iBike Pro appears to employ a venturi anemometer to measure windspeed. These devices are affected by both barometric pressure and wind direction.

My mistake. But my main point was that the barometric pressure sensor is independent of the anemometer and therefore the iBike does not have to rely upon a constant barometric pressure - it can adjust for it on the fly, at least in principle.

Except for? This is the underlying problem with the iBike and the reason its not a powermeter. IT IS A POWER ESTIMATOR. Go play with analyticcycling.com and see what "estimating" CdA does...

The question you aren't asking is what the error is in the CdA. I know exactly what changing CdA does, what I don't know is what the measurement error associated with this is. And calling it a "power estimator" is pure semantics.
If you don't consider the iBike to be a true power meter, then I would honestly like to hear your definition of the term.
SRM, Powertap, Ergomo, Polar.
Haha so because this is not an SRM, Powertap, Ergomo, or Polar, it is not a power meter. Very good. If this is your definition, I cannot help but agree, since iBike is definitely not any of these.

I would define powermeter somewhat differently. Perhaps this is the source of our semantic difficulties!
The measures employed by the major PMs are very much *not* direct, they rely on piezoelectric currents
Doesn't the iBike's accelerometer depend upon a piezoelectric (resistive) sensor? Probably the most accurate thing it does too.
Not only that, but their solution for measuring altitude gain (and therefore slope) is based on barometric pressure...
Dissing Polar again. Doesn't iBike use the same method via its barometric pressure port?


No. It uses the accelerometer you mentioned above. Polar relies upon barometric pressure for estimating hill slope, and the outcome can vary quite widely.
Summary:

iBike Pro's ESTIMATION of power is dynamically affected by any of the following:

1) Change in position
2) Change in weight
3) Change in tires or road surface
4) Change in barometric pressure (in vortex chamber) due to a front, elevation, or temperature change.
5) Drafting
6) It does not include a heart rate monitoring function, this must be factored into the value (and weight) equation.

Here you have shown why this argument is occurring, and in #5 I suspect I see why iBike is not well-liked on the wattage forum.

#2 is a fairly unusual thing to change by any significant amount during a single ride. Certainly even a water bottle's difference shouldn't change things dramatically.

The reason #4 is not an issue is listed above, so I won't go into it further.

#5 is a misstatement and exactly the sort of misunderstanding that could be iBike's biggest problem: of course iBike's readout is affected by drafting (if you think for a moment, you'll realize a PT is affected in exactly the same way ...)!

That leaves #1, #3, and #6. #1 I will agree is an issue - I just don't know to what degree (and I don't plan to make a judgement prior to getting data on it). #3 could also be an issue, though I suspect not a big one. #6 is a marketing problem, but you can get a cheap HR monitor (~$30 on eBay) separately if it's that important to you.
As such, iBike Pro is not only inaccurate, but non-repeatable. Repeatability actually being a more important factor than absolute accuracy.

This statement cannot be backed up and IMO you shouldn't be making it without data. Replace "is" with "seems to me to be" and we're fine.

If iBike sucks, I will be selling mine off pretty quick. Again, I'll let you know when it comes out.

User avatar
Bruiser
Posts: 1385
Joined: Mon Sep 30, 2002 1:59 am
Location: Sydney, Australia
Contact:

by Bruiser

#3 is a major issue for the roads I ride on. I regularly see my speed drop by 5-8kph because I've made the transition from new hot mix to a cheap blue stone surface. This could be avoided by having a number of calibration settings you could flick though.

Does Ibike have a forum or method to discuss these points?

User avatar
wdbike.com
Posts: 72
Joined: Mon Mar 14, 2005 4:11 pm

by wdbike.com

No idea. I kinda doubt it because it'd probably be on their web site. But they should.

The engineer at Interbike told me there was a future firmware update that was going to allow the user to use a few different calibration settings, but that it wasn't in the prototype model (because I asked him the same question lol). I think it's pretty important they do something like that. It would be even better if you were able to adjust each independently, ie. change from position in the drops to hoods with or without heavy surface changes. But I didn't ask them that, so no idea :?

Where do you ride that you see such huge changes?

User avatar
Bruiser
Posts: 1385
Joined: Mon Sep 30, 2002 1:59 am
Location: Sydney, Australia
Contact:

by Bruiser

I ride around regional Australia.

Our roads are so long the Gov't can't afford to maintain many of them, so the repairs they do are done piece by peice.

I ride in the drops usually but I'd still like to see the power difference of an aero tuck holding your cables.

by Weenie


Visit starbike.com Online Retailer for HighEnd cycling components
Great Prices ✓    Broad Selection ✓    Worldwide Delivery ✓

www.starbike.com



Post Reply