Re: When is that new power-measuring device arriving?
Ryan Cousineau 27 March 2005 12:07:24
In article <f6122af.0503261902.732ea892@posting.google.com>, anon_a_mouse@msn.com (Paul Kossa) wrote:
At last fall's bike show I heard that there is a new power device> coming out -- I thought they said in March, but apparently not --> that, while it only works on indoor trainers, it has the big advantage> of only costing $50 - $60. Anyone know about this?
No, but a power measurer that costs about a tenth of current systems would be huge news. If you hear anything, let us know.
Right now, I think the sweet spot in cycling equipment would be to do a logging HRM-cyclecomputer (like the Polar 520) for less than the couple hundred dollars the Polar costs.
The current best deal in cycle computers that I know of is the Filzer dB4LWc. I can't find any mention of it online yet, note even on Filzer's site, but MEC had this computer, complete with wireless speed and wired cadence sensing, for C$25.
Right now, I think the sweet spot in cycling equipment would be to do a >logging HRM-cyclecomputer (like the Polar 520) for less than the couple >hundred dollars the Polar costs.
The feedback I've seen from power meter users is that the Polar system (which measures chain tension) is rather finicky.
Most seem to favor either hub based systems such as the Powertap or crank based systems such as the SRM.
Unfortunately both are more $$$.
Chris Neary diabloridr@comcast.net
"Science, freedom, beauty, adventure: what more could you ask of life? Bicycling combined all the elements I loved" - Adapted from a quotation by Charles Lindbergh
Ryan Cousineau 28 March 2005 00:13:58 [ permanent link ]
In article <g2qd41l83jdenvaft9bd7d44pnesgp5gvg@4ax.com>, Chris Neary <diabloridr@comcast.net > wrote:
Right now, I think the sweet spot in cycling equipment would be to do a > >logging HRM-cyclecomputer (like the Polar 520) for less than the couple > >hundred dollars the Polar costs.>
The feedback I've seen from power meter users is that the Polar system> (which measures chain tension) is rather finicky.
I sorta changed the subject there: the Polar 7xx is the one with the power meter option. Power measurement is a great option, but you can do a lot with just heart rate logging. It just takes the burden off the rider to try to get a handle on their performance while in the middle of such distractions like racing.
Most seem to favor either hub based systems such as the Powertap or crank> based systems such as the SRM.>
Unfortunately both are more $$$.
What surprises me is that there doesn't seem to be much competition in the logging computer market. Basically, if you want to download data from your bike to your computer, the choices are Polar, SRM, or Powertap. This despite the availability of sub-$100 HR cyclocomputers and the aforementioned $25 cadence computer.
Maybe I'm naive about the costing, but it seems to me that adding a USB interface chip and enough memory to hold a couple of hours of data to a cheap HRM should be possible for $150 or less. I think I'll send an e-mail to Filzer and Cateye....
Ryan Cousineau wrote:> In article <g2qd41l83jdenvaft9bd7d44pnesgp5gvg@4ax.com>,> Chris Neary <diabloridr@comcast.net > wrote:>
Right now, I think the sweet spot in cycling equipment would be to do a >>>logging HRM-cyclecomputer (like the Polar 520) for less than the couple >>>hundred dollars the Polar costs.>>
The feedback I've seen from power meter users is that the Polar system>>(which measures chain tension) is rather finicky.>
I sorta changed the subject there: the Polar 7xx is the one with the > power meter option. Power measurement is a great option, but you can do > a lot with just heart rate logging. It just takes the burden off the > rider to try to get a handle on their performance while in the middle of > such distractions like racing.>
Most seem to favor either hub based systems such as the Powertap or crank>>based systems such as the SRM.>>
Unfortunately both are more $$$.>
What surprises me is that there doesn't seem to be much competition in > the logging computer market. Basically, if you want to download data > from your bike to your computer, the choices are Polar, SRM, or > Powertap. This despite the availability of sub-$100 HR cyclocomputers > and the aforementioned $25 cadence computer.>
Maybe I'm naive about the costing, but it seems to me that adding a USB > interface chip and enough memory to hold a couple of hours of data to a > cheap HRM should be possible for $150 or less. I think I'll send an > e-mail to Filzer and Cateye....>
I haven't seen enough of this to know for sure what has been tried but power is going to work out to 2 things that can be measured. Linear chain speed times tension averaged out over at least 10 seconds, so that it would not matter what gear you were using or cadence. Coast pedaling would show little or no power gain but could track RPM. The 10 second average gets rid of the stroke by stroke nature of pedaling. The math works but measuring the chain tension without power loss is a bit tricky. Memory for logging is not a problem if you have looked at digital camera memory lately. 512MB is under $100.00 so nearly free compared to other such 'goodies'. Bill Baka
At last fall's bike show I heard that there is a new power device> coming out -- I thought they said in March, but apparently not --> that, while it only works on indoor trainers, it has the big> advantage of only costing $50 - $60. Anyone know about this?
Measuring power is a simple concept that is more illusive than it fist appears. All it takes is a torsional strain gauge and tachometer in the rear hub. The trouble is that driving torque varies widely during each pedal stroke and varies in magnitude over such a wide range that a sensor is difficult to build.
Ideally there would be a torsion bar (equipped with a strain gauge) between the gear cluster and body of the hub. Such a torque sensor would need a limit stop because rider output on standing starts can exceed any reasonable torque for which the sensor is designed. Such a mechanism is hard to build inside a hub. Then, because the load varies widely during each wheel rotation, an averaging method must be employed to reduce data to some usable form, for instance average torque per wheel rotation, and then this data must get out of the wheel somehow. That requires a transmitter and power. This hub is getting larger all the time... and expensive. I believe that this is more expensive and difficult than it first appears and will probably stay that way.
Placing it in the BB seems even more difficult, the rotation rate being even slower and forces greater with no more space than in a rear hub.
512MB is under $100.00 so nearly free compared to other such 'goodies'.
This is a very conservative statement. 512MB secure digital flash cards are readily available for under $50.
Chris Neary diabloridr@comcast.net
"Science, freedom, beauty, adventure: what more could you ask of life? Bicycling combined all the elements I loved" - Adapted from a quotation by Charles Lindbergh
My lbs guy Steve writes reviews of gear for a local bike magazine and says the optical sensor in the bb is a clever idea but it only measures torsion in one crank, and attempts to infer the power going into the other crank, leaving the output readings open to question.
The Powertap hub is a bit heavy for use in a race. the Polar system is fiddly and not always reliable. which leaves the SRM system the only reasonably accurate on the bike power measurement at present - and as with all unregulated monopolies, SRM price their system accordingly.
My reason for enquiring about this is that I recently re-built my ergo which uses blades attached to the spokes of the driven wheel and a 7 sp Nexus internally geared hub for resistance - in the re-building I asked Steve about including some form of power measurement and we went through the alternatives, all of which cost a lot for something you can get a rough sense of from a simple cadence counter on the crank and recording how big a gear you can push for an effort at a constant cadence. If you keep to a given cadence, presumably reaching for a bigger gear indicates the training is going well - and vice versa.
From the set up I describe would it be possible to calculate watts delivered to the driven wheel ?
Chris Neary wrote:>>512MB is under $100.00 so nearly free compared to other such 'goodies'.>
This is a very conservative statement. 512MB secure digital flash cards are> readily available for under $50.>
Chris Neary > diabloridr@comcast.net>
"Science, freedom, beauty, adventure: what more could> you ask of life? Bicycling combined all the elements I> loved" - Adapted from a quotation by Charles Lindbergh
For once I am trying not to start a flame war of some sort because some people might be happy to pay 100 and others can get lucky and find one for maybe 25 dollars. Either way that is a lot of data to keep for a bike ride, so you could keep your heart rate from pulse to pulse, speed, power, wind, temp, and whatever else you can think of.
At last fall's bike show I heard that there is a new power device> coming out -- I thought they said in March, but apparently not --> that, while it only works on indoor trainers, it has the big advantage> of only costing $50 - $60. Anyone know about this?
Well, putting it on a trainer would be a lot easier than on a bike. So I'm not surprised if someone has come out with a cheap trainer unit.
If you want to have any hope of measuring power on a bicycle, why not use the accepted method for automobiles--a dynamometer. Even on cars, where the space and weight restrictions are much looser, I have never heard of a power meter that can be read from the driver's seat on a car in a race.
It seems to me that a stationary exercise bike setup could be fairly easily converted to measure power. You need to measure torque and RPM at the roller. By riding a stationary bike/dynamometer setup, you could find a relationship for YOURSELF, of heart rate vs power. This might vary, even for an individual, from day to day, and also for different states of fatigue. Still, that's a lot more than we know now.
If you want to have any hope of measuring power on a bicycle, why> not use the accepted method for automobiles--a dynamometer. Even on> cars, where the space and weight restrictions are much looser, I> have never heard of a power meter that can be read from the driver's> seat on a car in a race.
That's because it does not matter. The engine got tested on a dynamometer before installation in the car and not being humanly frail and emotional, it will develop that same power at the same RPM in a race. The bicyclist, in contrast, has an unpredictable power output that varies with many parameters and durations. Besides, in a car, power is easily measured at the output shaft of the engine during a race if there were any merit to the concept.
It seems to me that a stationary exercise bike setup could be fairly> easily converted to measure power. You need to measure torque and> RPM at the roller. By riding a stationary bike/dynamometer setup,> you could find a relationship for YOURSELF, of heart rate vs power.> This might vary, even for an individual, from day to day, and also> for different states of fatigue. Still, that's a lot more than we> know now.
As I mentioned, that isn't what the rider wants to know. Interesting would be the power output in different gears on a hillclimb, since torque and RPM don't have a fixed relationship for a human, especially with a different duty cycle as occurs in a flat TT or a variable grade hillclimb.
Has this has ever been done?
Although this may reveal at what percentage of full power one is riding, actual power is a derivative and not accurately determined by heart rate.
I haven't seen enough of this to know for sure what has been tried but> power is going to work out to 2 things that can be measured. Linear> chain speed times tension
This is how the Polar S710 power unit measures it.
jobst.brandt@stanfordalumni.org wrote:> Ideally there would be a torsion bar (equipped with a strain gauge)> between the gear cluster and body of the hub. Such a torque sensor> would need a limit stop because rider output on standing starts can> exceed any reasonable torque for which the sensor is designed. Such a> mechanism is hard to build inside a hub. Then, because the load> varies widely during each wheel rotation, an averaging method must be> employed to reduce data to some usable form, for instance average> torque per wheel rotation, and then this data must get out of the> wheel somehow. That requires a transmitter and power. This hub is> getting larger all the time... and expensive. I believe that this is> more expensive and difficult than it first appears and will probably> stay that way.
It seems to me that a stationary exercise bike setup could be fairly> easily converted to measure power. You need to measure torque and RPM> at the roller.
Interesting> would be the power output in different gears on a hillclimb, since> torque and RPM don't have a fixed relationship for a human, especially> with a different duty cycle as occurs in a flat TT or a variable grade> hillclimb.
Ideally there would be a torsion bar (equipped with a strain gauge)>> between the gear cluster and body of the hub. Such a torque sensor>> would need a limit stop because rider output on standing starts can>> exceed any reasonable torque for which the sensor is designed.>> Such a mechanism is hard to build inside a hub. Then, because the>> load varies widely during each wheel rotation, an averaging method>> must be employed to reduce data to some usable form, for instance>> average torque per wheel rotation, and then this data must get out>> of the wheel somehow. That requires a transmitter and power. This>> hub is getting larger all the time... and expensive. I believe>> that this is more expensive and difficult than it first appears and>> will probably stay that way.
This design does not generate much confidence or me. There is no clear explanation how and where the optical sensors are located or protected, although their need for careful protection is mentioned almost as a warning to not use this under anything but the most ideal conditions. As was mentioned, torque from the right crank (that appears on the left in the web blurb) goes directly into the chain and is never "seen" by the sensor.
This design does not generate much confidence or me. There is no> clear explanation how and where the optical sensors are located or> protected, although their need for careful protection is mentioned> almost as a warning to not use this under anything but the most ideal> conditions. As was mentioned, torque from the right crank (that> appears on the left in the web blurb) goes directly into the chain and> is never "seen" by the sensor.>
This design does not generate much confidence for me. There is no>> clear explanation how and where the optical sensors are located or>> protected, although their need for careful protection is mentioned>> almost as a warning to not use this under anything but the most>> ideal conditions. As was mentioned, torque from the right crank>> (that appears on the left in the web blurb) goes directly into the>> chain and is never "seen" by the sensor.
I'm still not convinced the mechanism is reliable and will not get contaminated by dry dirt, aside from rain and road splash. For singe riders, this is a technically manageable measurement task but where it might do more good is on tandems where balancing effort is one of the difficulties in such riding. For tandem riders, quantified power is not so much a problem as is effort. The two riders, not in competition with anyone cannot easily determine how much effort each is exerting.
The measure could easily be determined by riding at maximum output and speed and normalizing that effort as 100% and then having a percentage readout of both riders on each handlebar to help assess what ones fair share of effort should be. I believe this would be a boon to tandem riders who currently have only indirect ways of knowing which of the two is propelling the bicycle. This is a more complicated task than power meters on a single bicycle but it would have a long term practical value.
"Maggie" <lbuset@allsecretarial.com> wrote in message news:1112039134.498322.323960@o13g2000cwo.googlegroups.com...>
jobst.bra...@stanfordalumni.org wrote:
<< a bunch of stuff >>>
Are you a walking encyclopedia of bicycle knowledge?> Come mantenete tutto il questo nella vostra esplosione capa
withoug?> Going to Italia next year...gotta practice.> Maggie>
Jobst is more like a pedaling encyclopedia. For lots more info, see the FAQ's to the bicycle newsgroups which he wrote: http://www.sheldonbrown.com/brandt/
"Fiddly" describing the Polar might be due to the wide scope of bicycle frames which affect its application. On my Cannondale, it has provided great data save the times I have accidently disturbed the ream mounted speed sensor. The last time was during a flat fix and the sensor was too close to the wheel, making the speed read twice actual.
Measuring torque on the rear wheel seems even more "fiddly" to me due to the rotation. What would be better would be reaction torque or the compression of the right chain stay. Bending would have to be accounted but accurate compression measurement would indicate chain tension. Knowing chain speed via a Polar type sensor or crank speed and chainring position is easily done.
Weisse Luft wrote:> "Fiddly" describing the Polar might be due to the wide scope of bicycle> frames which affect its application. On my Cannondale, it has provided> great data save the times I have accidently disturbed the ream mounted> speed sensor. The last time was during a flat fix and the sensor was> too close to the wheel, making the speed read twice actual.
That is most likely a magneto-electric effect causing the sensor to be over stimulated and to have a 'return to zero' rebound and the echo is picked up as a signal. A minimal correction of the programming ( a few lines of code) would fix that but that is just my electronics background talking.>
Measuring torque on the rear wheel seems even more "fiddly" to me due> to the rotation. What would be better would be reaction torque or the> compression of the right chain stay. Bending would have to be> accounted but accurate compression measurement would indicate chain> tension. Knowing chain speed via a Polar type sensor or crank speed> and chainring position is easily done.>
Chain speed is the easy one and can be done optically for non contact. Torque must be corrected to know what gear you are in and frame flex is not the best way to go here what with bump flex, road noise, and many more variables which add up. Just moving your weight from sitting up and going 'no hands' would make a fairly big variable.
I looked at the graph shown and it looks more like a scatter plot than a power plot of any sort. It is sure not something I would base my training on, were I training. All that it is labeled to show up is variance, but from what?
I'm still not convinced the mechanism is reliable and will not get> contaminated by dry dirt, aside from rain and road splash. For singe> riders, this is a technically manageable measurement task but where it> might do more good is on tandems where balancing effort is one of the> difficulties in such riding. For tandem riders, quantified power is> not so much a problem as is effort. The two riders, not in competition> with anyone cannot easily determine how much effort each is exerting.>
The measure could easily be determined by riding at maximum output and> speed and normalizing that effort as 100% and then having a percentage> readout of both riders on each handlebar to help assess what ones fair> share of effort should be. I believe this would be a boon to tandem> riders who currently have only indirect ways of knowing which of the> two is propelling the bicycle. This is a more complicated task than> power meters on a single bicycle but it would have a long term> practical value.>
Jobst.Brandt@stanfordalumni.org
Regarding tandems, the only use I have seen mentioned here is when the stronger spouse wants to take her husband along on a 50 mile ride and she carries the load when he is winded. Organized tandem racing might have an application for the above logic, but in marital rides it might cause more disharmony than intended. Scientific evaluation is good but not healthy to put between a man and wife, especially if the wife shows 70% of the average power for a ride or vice versa. Some of the graphs I have seen do show that a human is good for more than a horsepower but only for a short burst. Happy riding. Bill Baka
Sigh. Nope, the link's not broken; yes, it's for one of those freebie sites that's (ahem) underpowered. Usually, but not always, when you get a "page not found" you can hit the reload current page button and it will appear.
Power at the wheel usually equals speed.>>
No.>>
Really? So you are implying that more power reduces your speed? I think> we were talking about level ground road biking with no wind. More power> equals more speed in this universe. Negative power is possible with the> brakes and that makes heat, which, again, is power.>
Please qualify your blanket 'No.'
In my universe, but perhaps not yours, roads aren't flat, the winds do blow, speeds aren't constant, and position changes. We're talking about on-bike power measuring devices. Why would you need such a thing if power equals speed? All you'd need is a speedometer.
I looked at the graph shown and it looks more like a scatter plot than a> power plot of any sort. It is sure not something I would base my> training on, were I training. All that it is labeled to show up is> variance, but from what?
That page compares measurement variation between the Ergomo and the PT so it addresses Jobst's scepticism about data quality from a BB-based system. The Ergomo and the PT produce pretty similar power numbers so if you think that one system is suitable for power-based training, you might think the other would be, too--at least, from a data quality perspective. These comparisons don't take into account reliability, usability, features, or cost; they only look at comparative data quality.
Sigh. Nope, the link's not broken; yes, it's for one of those freebie> sites that's (ahem) underpowered. Usually, but not always, when you get a> "page not found" you can hit the reload current page button and it will> appear.>
Power at the wheel usually equals speed.>>>
No.>>>
Really? So you are implying that more power reduces your speed? I think>>we were talking about level ground road biking with no wind. More power>>equals more speed in this universe. Negative power is possible with the>>brakes and that makes heat, which, again, is power.>>
Please qualify your blanket 'No.'>
In my universe, but perhaps not yours, roads aren't flat, the winds do> blow, speeds aren't constant, and position changes. We're talking about> on-bike power measuring devices. Why would you need such a thing if power> equals speed? All you'd need is a speedometer.>
I haven't seen a mention that this was just for indoor use and have seen a graph of power over a trip, so we need to be all talking about indoor or outdoor. Indoor is simple to put on a trainer so if that is the direction of the discussion, what is the big deal? How many people train indoors anyway??? Bill Baka
Torque must be corrected to know what gear you are in>
Not if you have wheel speed.>
Even that has to be corrected according to the gear you are in unless you are talking a single speed. This is not as simple as some would like and this thread could go on for a while due to that. Bill Baka
bbaka wrote:> Robert Chung wrote:>> bbaka wrote:>>
Torque must be corrected to know what gear you are in>>
Not if you have wheel speed.>>
Even that has to be corrected according to the gear you are in unless> you are talking a single speed. This is not as simple as some would like> and this thread could go on for a while due to that.
The Power Tap calculates power by using torque and angular velocity at the rear hub. The SRM uses torque and angular velocity at the crank. The Polar uses chainspeed and chain tension. None of these devices needs to know gear ratios.
Robert Chung wrote:> bbaka wrote:>> Robert Chung wrote:>>> bbaka wrote:>>>
Torque must be corrected to know what gear you are in>>>
Not if you have wheel speed.>>>
Even that has to be corrected according to the gear you are in unless>> you are talking a single speed. This is not as simple as some would>> like and this thread could go on for a while due to that.>
The Power Tap calculates power by using torque and angular velocity at> the rear hub. The SRM uses torque and angular velocity at the crank.> The Polar uses chainspeed and chain tension. None of these devices> needs to know gear ratios.
The Ergomo uses torque and angular velocity at the bottom bracket. It doesn't need to know gear ratio either.
Sigh. Nope, the link's not broken; yes, it's for one of those freebie> > sites that's (ahem) underpowered. Usually, but not always, when you get a> > "page not found" you can hit the reload current page button and it will> > appear.> >
Power at the wheel usually equals speed.> >>>
No.> >>>
Really? So you are implying that more power reduces your speed? I think> >>we were talking about level ground road biking with no wind. More power> >>equals more speed in this universe. Negative power is possible with the> >>brakes and that makes heat, which, again, is power.> >>
Please qualify your blanket 'No.'> >
In my universe, but perhaps not yours, roads aren't flat, the winds do> > blow, speeds aren't constant, and position changes. We're talking about> > on-bike power measuring devices. Why would you need such a thing if power> > equals speed? All you'd need is a speedometer.> >
I haven't seen a mention that this was just for indoor use and have seen > a graph of power over a trip, so we need to be all talking about indoor > or outdoor. Indoor is simple to put on a trainer so if that is the > direction of the discussion, what is the big deal? How many people train > indoors anyway???> Bill Baka
Bill, I think you've now argued both sides of the argument. What Robert is saying is that power, while a useful measure of a rider's ability, is by no means the only factor in performance. Remember how we were talking about the importance of aerodynamics? There are many cases where a rider with superior power output over a TT has been beaten by a rider with inferior power and better aerodynamics (through some combo of positioning, equipment, and morphology).
Similarly, power to weight mattes more than power once the road turns uphill, but even there aero matters more than you might think.
Ryan Cousineau wrote:> If we can come up with some plausible numbers for Lance's rate of climb,> distance, time elapsed, and the aero drag, we should be able to> calculate Lance's rough power output.
But the short answer is that Lance claims a peak power output of 600 W:
The grain of salt is that the nature of human power is that sustained> power output drops off dramatically from the peak. On his endurance> training rides, Lance claims an average power output of 245-280 W.
Based on the 2004 Alpe d'Huez ITT, I calculate his hour power to be around 400 W.
Ryan Cousineau wrote:> The best answer to your question is probably available from Dr Michele > Ferrari, who has posted on his site about VAM, a way of indirectly > estimating power and performance from climbing rates:>
If we can come up with some plausible numbers for Lance's rate of climb, > distance, time elapsed, and the aero drag, we should be able to > calculate Lance's rough power output.>
I went here and the peak power of only 600 W seems a bit low considering his sometimes nearly impossible sprints. Those are peak power over maybe 15-20 seconds so may be hard to actually measure.>
The grain of salt is that the nature of human power is that sustained > power output drops off dramatically from the peak. On his endurance > training rides, Lance claims an average power output of 245-280 W.
This one I believe since that is a good fraction of a horsepower which I believe is at 745 watts, so he is at the top of what a human his size could possibly put out. A 6'6" rider would have to put out better than one half horsepower to be competitive, especially on the climbs where raising ones weight to a higher elevation eats the power.
In article <1112299369.136500@news01.syix.com>, bbaka <bbaka@syix.com> wrote:
Ryan Cousineau wrote:> > The best answer to your question is probably available from Dr Michele > > Ferrari, who has posted on his site about VAM, a way of indirectly > > estimating power and performance from climbing rates:> >
If we can come up with some plausible numbers for Lance's rate of climb, > > distance, time elapsed, and the aero drag, we should be able to > > calculate Lance's rough power output.> >
I went here and the peak power of only 600 W seems a bit low considering > his sometimes nearly impossible sprints. Those are peak power over maybe > 15-20 seconds so may be hard to actually measure.
Well, my impression of Lance's hill-climbing prowess is that he isn't sprinting. It's just that when he attacks, he's putting out more power than any other rider can sustain, so he moves away rapidly. Note that small changes in power output (or power-to-weight ratio, really) make big differences in speed on climbs, because the effort required to overcome gravity increases linearly, while the effort required to overcome aerodynamic drag increases as the...square? of velocity.
There is still an aero component at the climbing speeds of pro cyclists, but it is considerably diminished compared to their usual flatland pace.
So, he's not sprinting, he's just that much better.
I'd also bet that Lance has not seriously trained for sprinting power in several years, since it's basically not part of his game plan. He has to do huge work to prepare for the recovery elements and get his aerobic power higher, he works hard on TT performance, and he has to be able to climb like a Spaniard, but the sprinting he leaves for others.
