The Powermeter Thread

[sciguy]

What percentage of power meter owners do you think do that?/QUOTE]

Why should I care about that when I do know that 100% of them have the option available

Not what percentage of your friends but what percentage of all owners./QUOTE]

Well that is 100% even though you didn't want to know that

Sure, I know a PM can be tested for accuracy. I told you how I would do it (compare it to a Velotron) /QUOTE]

Yes you mentioned

by my best estimate from rudimentary efforts to find a compensation factor, about 40% low"/QUOTE] but that doesn't really seem like much of a test of accuracy.

How about comparing ride files between the Velotron and the Icrank? You aught to be able to get a lot closer than about. How about a test at 50, 100, 150 and 200 watts for 15 minutes each. Compare files and see is there is a consistent % difference. Jeepers with 200+ hours on these you should have a mountain of data to compare. Do you have an aversion to collecting and using actual data?

Hugh

 

[FrankDay]

How about comparing ride files between the Velotron and the Icrank? You aught to be able to get a lot closer than about. How about a test at 50, 100, 150 and 200 watts for 15 minutes each. Compare files and see is there is a consistent % difference. Jeepers with 200+ hours on these you should have a mountain of data to compare. Are you adverse to collecting and using actual data?

Hugh

The "problem" is one (the Velotron) measures power at the wheel and the other (the iCranks) measures power at the crank arm. So, to do a calibration requires making some assumptions regarding chain and bearing losses. The fact that the two files are not the same is not evidence that they are not both accurate (unless, of course, the Velotron's power is higher than the iCranks, which it wasn't). Anyhow, I did something like your suggestion but it isn't necessary to do 15 minutes because in ergometer mode the VT maintains a set power quite closely so all one need do is set the iCranks smoothing to something long and see where it settles out. I did this at various crank lengths to see if I could determine a compensation factor for the different crank lengths. I did this down to 145mm (see chart).

The iCranks are supposedly calibrated for 172.5 crank length so it looks like I am seeing about 10W chain/bearing loss in VT at this power. Note the calibration curve is not a straight line because the strain gauges are not located at the hub of the cranks.

My estimate of 40% low for 125-130 is based upon the further drop I saw compared to my average power for my usual training rides that I had been doing at 145 mm crank length (avg power for rides was typically within 2-3 watts day after day) without doing the whole calibration thing again.

 

[FrankDay]

What percentage of power meter owners do you think do that?

Why should I care about that when I do know that 100% of them have the option available

Actually, you are telling me that none of your friends have the new power meters with l/r independent measurement because, as I understand it, it is not possible to do such a calibration on the drive side crank when there are two power meters using current available methods because all communication to the head unit occurs through the left crank. The iCranks people are working on it. Although they say that their cranks should never need calibrating I don't think that it will sit very well with the nerds if they can't calibrate them, both of them.

 

[sciguy]

The "problem" is one (the Velotron) measures power at the wheel and the other (the iCranks) measures power at the crank arm. So, to do a calibration requires making some assumptions regarding chain and bearing losses. The fact that the two files are not the same is not evidence that they are not both accurate (unless, of course, the Velotron's power is higher than the iCranks, which it wasn't). Anyhow, I did something like your suggestion but it isn't necessary to do 15 minutes because in ergometer mode the VT maintains a set power quite closely so all one need do is set the iCranks smoothing to something long and see where it settles out. I did this at various crank lengths to see if I could determine a compensation factor for the different crank lengths. I did this down to 145mm (see chart).

The iCranks are supposedly calibrated for 172.5 crank length so it looks like I am seeing about 10W chain/bearing loss in VT at this power. Note the calibration curve is not a straight line because the strain gauges are not located at the hub of the cranks.

My estimate of 40% low for 125-130 is based upon the further drop I saw compared to my average power for my usual training rides that I had been doing at 145 mm crank length (avg power for rides was typically within 2-3 watts day after day) without doing the whole calibration thing again.

Frank,

That looks like a good start. I don't understand why you didn't run the 125 and 130mm cranks against the Velotron though. Guessing that you were riding the same watts on your course seems totally inadequate compared to a straight up comparison of the two power meters. You aught to run several different wattages as you may find that higher or lower wattages may not match up are equivalent percentages.

I'm thinking 10 watts loss on the drive train is a bit on the high side unless you're in need of maintenance on the bike. By running other wattages you may be able to sift out what the drive train loss is. IIRC if you run in Erg mode with the same chain speed the drive train loss should not be very much different at higher wattage.

Any luck finding folks to build you a computer head that can deal with the compensation factors needed for various crank lengths? Even if you explain to customers that the shorter cranks real artificially low you'll still have some hating to see the lower numbers on a screen.

Hugh

 

[FrankDay]

Frank,

That looks like a good start. I don't understand why you didn't run the 125 and 130mm cranks against the Velotron though. Guessing that you were riding the same watts on your course seems totally inadequate compared to a straight up comparison of the two power meters. You aught to run several different wattages as you may find that higher or lower wattages may not match up are equivalent percentages.

I was simply trying to get some sort of calibration curve for different crank lengths so I could do crank length testing. I wasn't trying to calibrate the iCranks when I did this.

I'm thinking 10 watts loss on the drive train is a bit on the high side unless you're in need of maintenance on the bike. By running other wattages you may be able to shift out what the drive train loss is. IIRC if you run in Erg mode with the same chain speed the drive train loss should not be very much different at higher wattage.

I thought it a little high also but it was what it was and didn't affect what I was trying to do.

Any luck finding folks to build you a computer head that can deal with the compensation factors needed for various crank lengths? Even if you explain to customers that the shorter cranks real artificially low you'll still have some hating to see the lower numbers on a screen.

Hugh

I agree that the average person won't be happy seeing lower numbers and the average person probably won't think to change the calibration constant even if available. Note, the numbers are too high for longer crank lengths also. This could lead some to conclude that longer cranks are more powerful when they are not. Garmin Vector is going to have a similar problem I suspect as they are the only other PM that is crank based that allows for crank length testing.

 

[sciguy]

Actually, you are telling me that none of your friends have the new power meters with l/r independent measurement because, as I understand it, it is not possible to do such a calibration on the drive side crank when there are two power meters using current available methods because all communication to the head unit occurs through the left crank. The iCranks people are working on it. Although they say that their cranks should never need calibrating I don't think that it will sit very well with the nerds if they can't calibrate them, both of them.

My friends are riding SRMs, Quarqs, Powertaps and Power2Maxes. I'm not familiar with the "problem" you've mentioned but do know of individuals who have successfully done static testing on Garmin Vector pedals. Is it the Rotor Crank power meter you're referring too? I've yet to see one of those in the States.

The never need calibrating concept sounds great but I know my reliable old SRM had a slope change over time so I'm all for "testing" rather than trusting.

Hugh

 

[FrankDay]

My friends are riding SRMs, Quarqs, Powertaps and Power2Maxes. I'm not familiar with the "problem" you've mentioned but do know of individuals who have successfully done static testing on Garmin Vector pedals.

They were able to do static testing on the right Garmin pedal? Exactly how did they do that?

The never need calibrating concept sounds great but I know my reliable old SRM had a slope change over time so I'm all for "testing" rather than trusting.

Hugh

Sounds good, I agree, but I wouldn't, necessarily, trust it either until it had proven itself over time.

 

[CoachFergie]

My friends are riding SRMs, Quarqs, Powertaps and Power2Maxes. I'm not familiar with the "problem" you've mentioned but do know of individuals who have successfully done static testing on Garmin Vector pedals. Is it the Rotor Crank power meter you're referring too? I've yet to see one of those in the States.

The never need calibrating concept sounds great but I know my reliable old SRM had a slope change over time so I'm all for "testing" rather than trusting.

Static testing on my Powertap shows it is out but it's dismantling the wheel, sending it from NZ to US and back to get calibrated which is a lot of time and expense. Very easy to do with the SRM I had and the Quarq I have. Would have been nice to do a static test on the Power 2 Max I was testing for the NZ Agent but wasn't set up for it at the time. They claim to not need recalibration. And we now about "claims" around here don't we

 

[Alex Simmons/RST]

Would you be happier if I guaranteed everyone a 10% power improvement over 6 months? Would you call that utter bull****? The few who didn't see 10% would probably be about the same number as return the cranks for money back now, 1-2 per thousand.

And, just out of curiosity, how does the end user "judge" whether their power meter is within the manufacturers specifications?

I'd be happier if you reported the proven outcomes of using the cranks over and above that attainable from training on regular cranks. Since that's SFA, then that's what you should report and claim.

 

[FrankDay]

I'd be happier if you reported the proven outcomes of using the cranks over and above that attainable from training on regular cranks. Since that's SFA, then that's what you should report and claim.

But, that is almost impossible to extract without a control group. Our group we felt was "stable" so we believe most of the improvement was due to the PowerCranks but we cannot state that so all we do is say our typical user sees 40% increase even though all of that may not be due to our device.

This is a problem for any intervention, trying to separate out the effect of the intervention from what would have occurred anyhow. People get a power meter then report improvement. Is it because of the power meter allowing them to train smarter or would it have occurred anyhow? The studies seem to suggest it isn't the power meter but try to tell that to the people who have seen the improvement. At least PowerCranks do have studies suggesting that there is a PC effect but none of them allow us to separate out training effect from PC effect.

So, if you would be happier if I separated out the PC improvement then I suggest that you do the work to figure out the difference and we will use your data. Most of our customers attribute all of the improvement they see to the PC's so you (and we) can't rely on them.

 

[CoachFergie]

I'd be happier if you reported the proven outcomes of using the cranks over and above that attainable from training on regular cranks. Since that's SFA, then that's what you should report and claim.

A little ambitious considering Frank clearly doesn't understand the research reporting process or some pretty basic research principles.

Mind you considering he thinks my study is so good I wish he was grading my Thesis

 

[sciguy]

They were able to do static testing on the right Garmin pedal? Exactly how did they do that?.

One hangs a weight off of it just like the left pedal. Apparently one needs to have a Garmin 510 or 810.

Hugh

 

[FrankDay]

One hangs a weight off of it just like the left pedal. Apparently one needs to have a Garmin 510 or 810.

Hugh

It looks like on the Vector the right crank is the master and the left the slave, backwards from the iCranks. That having been said I just downloaded the Vector Instruction manual. The only mention of calibration involves zero calibration and a "backwards pedaling" calibration (apparently to calibrate pedal location). I see no mention of hanging a weight to calibrate. Can you provide a link as to how one would do this for both cranks on the Vectors?

 

[sciguy]

It looks like on the Vector the right crank is the master and the left the slave, backwards from the iCranks. That having been said I just downloaded the Vector Instruction manual. The only mention of calibration involves zero calibration and a "backwards pedaling" calibration (apparently to calibrate pedal location). I see no mention of hanging a weight to calibrate. Can you provide a link as to how one would do this for both cranks on the Vectors?

I'm not seeing the problem with a master slave set up that you seem to be worried about. Here is what Jens from a Wattage said today to the following question.

Jens,

Are you able to hang a weight from either the left or right Vector pedal to see the torque each see's or does the Master/slave relationship interfere?

Yes, you can measure both pedals independently. Though you need a Edge 510 or 810 to get the torque reading.

I did an extensive new test today. I used two weights (light 6,1kg and heavy 18,5kg). I tested all 3 powermeters, both crankarms / pedals and for SRM and Vector on the pedal body and pedal axle. For PT I used 11 gear combinations.

Here is what I got:

Vector

1. Hanging the weight on the pedal body reduces the torque between 0,5 and 1%.

2. Left and right are the same.

3. Torque with weight on the pedal axle matches the weight from the body weight scale.

4. Very easy and fast to do.

 

[CoachFergie]

http://m.us.wsj.com/articles/SB1000...9536182357290174.html?mod=WSJ_hp_EditorsPicks

Why you measure stuff. Some pretty basic errors in event prep going on here.

 

[Alex Simmons/RST]

But, that is almost impossible to extract without a control group.

Then don't make the claim until you can. Otherwise it is deceitful to do so.

Our group we felt was "stable" so we believe most of the improvement was due to the PowerCranks but we cannot state that so all we do is say our typical user sees 40% increase even though all of that may not be due to our device.

Then you should be making it very clear to people that you cannot say how much improvement will be due to using the device compared with simply training on regular cranks.

This is a problem for any intervention,

Not if you use good protocol and analysis assessment.

So, if you would be happier if I separated out the PC improvement then I suggest that you do the work to figure out the difference and we will use your data. Most of our customers attribute all of the improvement they see to the PC's so you (and we) can't rely on them.

Repeating a lie doesn't make it true.

I would be happier if you made truthful claims.

 

[FrankDay]

Then don't make the claim until you can. Otherwise it is deceitful to do so.

No it isn't. Our claim is what our average person sees who uses them as we prescribe. Most people really don't care how the improvement is broken down and since it is impossible to do so there is no deceit involved. Edit: we also have a video of about 50 cyclists telling their own story. Not all of them report 40% improvement although some report more. What is deceitful about that?

Then you should be making it very clear to people that you cannot say how much improvement will be due to using the device compared with simply training on regular cranks.

I certainly have made that clear here haven't I. As I said, the average customer doesn't care how the improvement is broken down. Get a power meter and all the improvement is training effect (but they think it is the pm). Get PowerCranks and some of the improvement is training effect and some is technique improvement and the overall improvement should be larger than training effect. Does the user care? Me thinks not. All they want to know is what they might expect when they plunk down their money.

Not if you use good protocol and analysis assessment.

Oh phoey. It is a problem. Not insurmountable but hardly worth the effort to break down the improvement until one has demonstrated that there is an improvement. No need to make a study more difficult than necessary.

Repeating a lie doesn't make it true.

We don't repeat any lies. SRM posts power files of tour stages as if to imply that SRM helped those riders to ride at that level so as to sell more product. Is that a lie? The fact that many think that they got better because they got a power meter does not make it a lie if they state that (and many do). It is only a lie if they know it to be false when they state it. The fact that they believe it makes it true to them. All we repeat is the improvement that people report from using the device because that is what people are really interested in. I am told my web site has way too much information and is too busy now. Not sure it would help things if I went into a long explanation as to how our 40% improvement figure is comprised of many parts, blah, blah, blah. If they want more detailed information they can contact us (many do).

I would be happier if you made truthful claims.

I would be happier if you knew what the truth was so you would know that I do make truthful claims.

 

[CoachFergie]

3hrs of information on racing and training with a power meter.

http://www.powertap.com/blogs/train...w-to-train-with-power-by-peaks-coaching-group

 

[coapman]

3hrs of information on racing and training with a power meter.

http://www.powertap.com/blogs/train...w-to-train-with-power-by-peaks-coaching-group

Another form of spam

 

[sciguy]

Another form of spam

Noel,

Fergie doesn't even sell Powertaps so I find it a real stretch to spin his providing a link to that site as an act of "spamming" . This specific thread is here to provide information regarding the use of Power meters. It would seem that the linked site is pretty spot on in that regard.

Now if he owned the company that sold Powertaps that would be a bit of a different story.

Hugh

 

[coapman]

Noel,

Fergie doesn't even sell Powertaps so I find it a real stretch to spin his providing a link to that site as an act of "spamming" . This specific thread is here to provide information regarding the use of Power meters. It would seem that the linked site is pretty spot on in that regard.

Now if he owned the company that sold Powertaps that would be a bit of a different story.

Hugh

They create additional work for coaches.

 

[sciguy]

They create additional work for coaches.

Not for my wife, my self or our dozens of friends who use power meters and are self coached I know a number of athletes who purchased power meters and promptly dropped their coach.

Hugh

 

[FrankDay]

Noel,

Fergie doesn't even sell Powertaps so I find it a real stretch to spin his providing a link to that site as an act of "spamming" . This specific thread is here to provide information regarding the use of Power meters. It would seem that the linked site is pretty spot on in that regard.

Actually, the purpose of this specific thread is found in the first post:

I thought I would start a power meter thread to discuss the science and practice behind this measurement tool.

While the link goes to the practice it certainly has nothing to do with the science (because there is no science behind the use). Let's see, the first video is 56 minutes devoted to the benefits of training with power. Since there is no scientifically demonstrated benefit to training with power that must be 56 minutes of opinion and anecdote. Here are a few: 3:40

we learned a lot about BMX with power meters…helped us to get a silver and bronze medals at 2008 Olympics.

anecdote implying cause and effect.
5:00

You cannot improve that which you cannot measure.

LOL. You cannot improve power without measuring it? Pure marketing BS here.
6:04

RPE=good data
HR=better data
Power=best data

Since we use data for making decisions the implication here is that using power will result in better, more effective, decisions. Unfortunately, there is simply zero scientific support to argue that one of these effort feedback data measurements is better than another for determining outcome.

I will stop there. It is all a bunch of marketing hooey. I guess it might someday be shown to be true (to a small degree) but until there is some scientific evidence to support such ideas all these thoughts have to be considered nothing more than opinion and anecdote. This is a video series given by someone selling software services (and books) and sponsored by a power meter manufacturer. What bias could possibly exist?

 

[CoachFergie]

They create additional work for coaches.

One could also argue that by educating people about using a power meter a coach is creating less work for themselves.

 

[FrankDay]

One could also argue that by educating people about using a power meter a coach is creating less work for themselves.

To me the value of a pm to a coach is the ability to know that the rider is doing the specified workout, rather than relying on reports. I can see why coaches like them (plus to the fact they can point to the improved power number as "proof" that they are worth the cost). That doesn't overcome the "problem" that there is no demonstrated benefit to such desires.