5 power meters, 5 ways of measuring. What's the best?

[jojoma]

I'm still a novice with power, and looking at the following devices, they all have a different way of calculating power:

SRM: Crank
Ergomo: Bottom Bracket
Powertap: Hub
Polar: Chain
iBike: Opposing Forces

Is one way better than the rest? Do they do it differently because of patents or because power measurement is still somewhat new and they are looking for differentiation?

I'm sure over the years there have been contraptions to measure heart rate through the wrist, the forefinger, the neck, or other places, but the industry has standardized on the chest strap. Will power measurement move in one direction too?

 

[Roadie_scum]

I'm still a novice with power, and looking at the following devices, they all have a different way of calculating power:

SRM: Crank
Ergomo: Bottom Bracket
Powertap: Hub
Polar: Chain
iBike: Opposing Forces

Is one way better than the rest? Do they do it differently because of patents or because power measurement is still somewhat new and they are looking for differentiation?

I'm sure over the years there have been contraptions to measure heart rate through the wrist, the forefinger, the neck, or other places, but the industry has standardized on the chest strap. Will power measurement move in one direction too?

SRM and Powertap are the only consistent and accurate powermeters currently available.

 

[Tom Anhalt]

SRM and Powertap are the only consistent and accurate powermeters currently available.

I would add a properly installed Polar to that list...and I've got almost 1 month's worth of "head to head" data with a PT to prove it

 

[lanierb]

I would add a properly installed Polar to that list...and I've got almost 1 month's worth of "head to head" data with a PT to prove it

Problem is... you need a PT to "properly install" a Polar!

 

[Tom Anhalt]

Problem is... you need a PT to "properly install" a Polar!

Not in the least. The only measuring instruments you need are a good tape measure and an accurate scale.

 

[beerco]

I would add a properly installed Polar to that list...and I've got almost 1 month's worth of "head to head" data with a PT to prove it

Tom,

Have you ever done any inter gear comparisons against the power tap? Back when I first bought a Polar back in 2002 I did some hill runs in each gear and noticed a definite pattern where some gears read higher and some lower - Kind of a W pattern.

On average the polar worked fine and gave rational numbers most of the time but sometimes the numbers just seemed too high or two low.

It all became clear to me when I did a TT with the polar - Good performance for me at the time but the power numbers reported were definitely out of my league. I suspect that I probably spent a lot of time in the gears which read higher than average which resulted in a significantly skewed AP number.

Ever notice anything like that?

-Andy B.

 

[Tom Anhalt]

Tom,

Have you ever done any inter gear comparisons against the power tap? Back when I first bought a Polar back in 2002 I did some hill runs in each gear and noticed a definite pattern where some gears read higher and some lower - Kind of a W pattern.

On average the polar worked fine and gave rational numbers most of the time but sometimes the numbers just seemed too high or two low.

It all became clear to me when I did a TT with the polar - Good performance for me at the time but the power numbers reported were definitely out of my league. I suspect that I probably spent a lot of time in the gears which read higher than average which resulted in a significantly skewed AP number.

Ever notice anything like that?

-Andy B.

Sort of. During the time I had both PMs on the same bike, I also did some crits. Robert Chung was kind enough to take a look at the files and attached is a "gear-by-gear" comparison of one of the crits. Here's a question about what the chart represents and his response:

Tom - "The dark line in the boxes represent the mean of the data,
the "height" of the box represents the "spread" of the
data, while the width of the box represents the amount
of data points, right?"

Robert - "Almost. The dark line is the median, not the mean. The median values for the PT were lower, but the peaks for the PT were higher (and both are bounded below by zero) so that tends to make up for the difference when you look at averages."

(Note - in other words, the averages were closer across the gears. Also, the "11" was a place holder for coasting values and the amount of data for the 21 and 23 cogs was so sparce that no conclusion can be drawn from those results).

Now...that also brings up the question about whether the variation between what the Polar is reporting and what the PT is reporting is a function of the Polar, or is it variations in drivetrain efficiency across the gears. See the Kyle and Berto report here and look at the results for the Shimano 27 speed setup.

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

 

[Eden]

One could argue that the device that gives the most consistant results, regardless of accuracy would be the best one. Consider that your actual power numbers are meaningless when it really comes down to a competition. It doesn't matter what your numbers are if you can't out perform those you are racing against. It would be better to use the device to train consistantly. I don't know which of these it is, though I would suspect that the most consistant is probably also the most accurate? Yes, no, maybe? Anyone know which one will give the most consistant results? Does it require a lot of constant fussing and calibrating to get consistancy?

 

[daveryanwyoming]

One could argue that the device that gives the most consistant results, regardless of accuracy would be the best one. ....

Sure, but how exactly do you go about determining which PM out of several choices is the most consistent? Consistent relative to what absolute measure? Good concept, but hard to carry out in practice.

 

[Tom Anhalt]

BTW, here's how the performance of the two matched up using averages and over a longer period of time. Some interesting stuff in there on chain condition and it's effects too:

http://biketechreview.com/phpBB2/viewtopic.php?t=1370