Power Meter Data Files

[Kraig Willett]

For those interested in doing your own analysis on the power meter data that I collected for the
following review (the big three power systems on the same bike for several weeks),

http://tinyurl.com/7jft

I have made the raw data files available at:

http://tinyurl.com/7kxo

--
==================
Kraig Willett www.biketechreview.com
==================

 

[Tim McNamara]

In article <[email protected]>,
"Kraig Willett" <[email protected]> wrote:

> For those interested in doing your own analysis on the power meter data that I collected for the
> following review (the big three power systems on the same bike for several weeks),

Interesting. I didn't have time to scour your report, but I found myself wondering:

1. Did the SRM pick up "spikes" that were in fact not significant in terms of speed (e.g., one
really hard pedal stroke as you stood up on the right pedal on a steep hill). I guess I'm asking
if the device is overly sensitive to this type of input.

2. Does the SRM accurately measure input from the left leg, or are there data losses from torque
through the bottom bracket axle and bearings?

3. Does chain tension reliably measure power?

 

[Andy Coggan]

"Tim McNamara" <[email protected]> wrote in message
news:[email protected]...
> In article <[email protected]>,
> "Kraig Willett" <[email protected]> wrote:
>
> > For those interested in doing your own analysis on the power meter data
that
> > I collected for the following review (the big three power systems on the same bike for several
> > weeks),
>
> Interesting. I didn't have time to scour your report, but I found myself wondering:
>
> 1. Did the SRM pick up "spikes" that were in fact not significant in terms of speed (e.g., one
> really hard pedal stroke as you stood up on the right pedal on a steep hill). I guess I'm
> asking if the device is overly sensitive to this type of input.

Well, one way to look at it that all such "spikes" in power are significant in terms of your speed -
after all, it's not like the extra energy is wasted. However, it is true that the SRM, PowerTap, and
(to a lesser extent) Polar S710 can measure power on a much shorter time scale than may be necessary
for routine use on the road. For other purposes, though, you need as much - or even more - temporal
resolution as these devices offer.

Personally, I find the default value of displaying a 3 second rolling average to be about right -
anything significantly longer than that I find annoying, since it creates a discrepancy/time shift
between what my legs tell me and what the display reports.

> 2. Does the SRM accurately measure input from the left leg, or are there data losses from torque
> through the bottom bracket axle and bearings?

If there are, they're going to be tiny. I would worry more about devices that attempt to measure the
torque applied to the bottom bracket spindle itself (since this is nearly zero when the right leg
drives downward).

> 3. Does chain tension reliably measure power?

Yes - but the question really is, how reliably can you measure chain tension? The answer seems to be
"well enough, but not in every circumstance".

Andy Coggan

 

[Tim McNamara]

In article <[email protected]>, "Andy Coggan"
<[email protected]> wrote:

> "Tim McNamara" <[email protected]> wrote in message
> news:[email protected]...
> > In article <[email protected]>,
> > "Kraig Willett" <[email protected]> wrote:
> >
> > > For those interested in doing your own analysis on the power meter data that I collected for
> > > the following review (the big three power systems on the same bike for several weeks),
> >
> > Interesting. I didn't have time to scour your report, but I found myself wondering:
> >
> > 1. Did the SRM pick up "spikes" that were in fact not significant in terms of speed (e.g., one
> > really hard pedal stroke as you stood up on the right pedal on a steep hill). I guess I'm
> > asking if the device is overly sensitive to this type of input.
>
> Well, one way to look at it that all such "spikes" in power are significant in terms of your speed
> - after all, it's not like the extra energy is wasted. However, it is true that the SRM, PowerTap,
> and (to a lesser extent) Polar S710 can measure power on a much shorter time scale than may be
> necessary for routine use on the road. For other purposes, though, you need as much - or even more
> - temporal resolution as these devices offer.
>
> Personally, I find the default value of displaying a 3 second rolling average to be about right -
> anything significantly longer than that I find annoying, since it creates a discrepancy/time shift
> between what my legs tell me and what the display reports.

Which would be similar to the lag problem with heart rate monitors Kraig mentions early in his
article. OTOH, the power measurement could still be useful for something in training (I'm not really
sure what, to be honest- after all the winner is who gets there first, not who produces the most
power. Cipo has has more power than Lance, but Cipo ain't ever gonna win the Tour).

> > 3. Does chain tension reliably measure power?
>
> Yes - but the question really is, how reliably can you measure chain tension? The answer seems to
> be "well enough, but not in every circumstance".

Good point.

 

[Andy Coggan]

"Tim McNamara" <[email protected]> wrote in message
news:[email protected]...

> the power measurement could still be useful for something in training (I'm not really sure what,
> to be honest- after all the winner is who gets there first, not who produces the most power.

True, but if you can make more power you have a better shot at getting their first. That is, after
all, the whole purpose of training to begin with - measuring the effects of said training (i.e.,
power output) is just provides a way of maximizing the beneft.

Andy Coggan

 

[Kraig Willett]

"Tim McNamara" <[email protected]> wrote in message
news:[email protected]...
> 1. Did the SRM pick up "spikes" that were in fact not significant in terms of speed (e.g., one
> really hard pedal stroke as you stood up on the right pedal on a steep hill). I guess I'm
> asking if the device is overly sensitive to this type of input.

Having extremely limited exposure to power measuring prior to the test, the first thing I noticed
was just how variable power is - this is very different than HR. While climbing Palomar, my HR was
175-180 ish, while my power was 210-300 ish (IIRC).

> 2. Does the SRM accurately measure input from the left leg, or are there data losses from torque
> through the bottom bracket axle and bearings?

I think Andy covered this one.

> 3. Does chain tension reliably measure power?

The Polar device is clever. Though a bit cumbersome to install, the unit works as claimed.

--
==================
Kraig Willett www.biketechreview.com
==================

 

[Andy Coggan]

"Kraig Willett" <[email protected]> wrote in message
news[email protected]...

> Having extremely limited exposure to power measuring prior to the test, the first thing I noticed
> was just how variable power is - this is very different than HR. While climbing Palomar, my HR was
> 175-180 ish, while my power was 210-300 ish (IIRC).

If I had $1 for every new power meter user who expressed the same surprise.


Seriously, I think this almost universal response is partially due to the ubiquitous use of heart
rate monitors - if you've never used one of the latter, and simply think about how hard you press on
the pedals at any given moment, then of course it makes sense that power is highly variable.

> > 3. Does chain tension reliably measure power?
>
> The Polar device is clever. Though a bit cumbersome to install, the unit works as claimed.

Did you try it on a trainer?

Andy Coggan

 

[Kraig Willett]

"Andy Coggan" <[email protected]> wrote in message
news:[email protected]...
> > > 3. Does chain tension reliably measure power?
> >
> > The Polar device is clever. Though a bit cumbersome to install, the unit works as claimed.
>
> Did you try it on a trainer?

Yep, my bad, Andy - I don't want to mislead anybody. I should have stated the following:

Though a bit cumbersome to install, the unit works as claimed with occasional problems with quirky
data. One problematic usage I had with the Polar was while riding a trainer - and I will say that
with a custom shim the issue may have been fixed (but whether or not custom shimming should be
necessary on a $600+ unit is another question).

The problem I had was with elevated power in the 39x14-12 (rarely use those gears) and also in the
53x13-12 (IIRC) on the trainer. It is pretty well explained in the review for those interested.

--
==================
Kraig Willett www.biketechreview.com
==================

 

[jeffchrisope]

Originally posted by Kraig Willett
"Andy Coggan" <[email protected]> wrote in message
news:[email protected]...
> > > 3. Does chain tension reliably measure power?
> >
> > The Polar device is clever. Though a bit cumbersome to install, the unit works as claimed.
>
> Did you try it on a trainer?

Yep, my bad, Andy - I don't want to mislead anybody. I should have stated the following:

Though a bit cumbersome to install, the unit works as claimed with occasional problems with quirky
data. One problematic usage I had with the Polar was while riding a trainer - and I will say that
with a custom shim the issue may have been fixed (but whether or not custom shimming should be
necessary on a $600+ unit is another question).

Hey Kraig, I'm thinking about picking up the Polar Power setup, but your review has made me quite nervous, due to the data quality issues in high gears (as well as some odd cross-gears). What is the final verdict: does the shim eliminate all of these problems, just the trainer issues ...?

Thanks!

 

[Kraig Willett]

"jeffchrisope" <[email protected]> wrote in message
> Hey Kraig, I'm thinking about picking up the Polar Power setup, but your review has made me quite
> nervous, due to the data quality issues in high gears (as well as some odd cross-gears). What is
> the final verdict: does the shim eliminate all of these problems, just the trainer issues ...?

The shim _may_ fix the data issues I observed. I don't know if it will/could for sure, since I
returned all power meters back to the manufacturers after the one month comparison period. IMHO, one
shouldn't have to make a custom shim for a $600 unit to get it to work properly.

If you are second guessing yourself and are concerned about the potential data quality issues of the
polar this late in the game, put your mind at rest and pop the few extra bucks for the more
painless/robust installation of the PT or SRM.
--
==================
Kraig Willett www.biketechreview.com
==================

 

[jeffchrisope]

Originally posted by Kraig Willett
[BIf you are second guessing yourself and are concerned about the potential data quality issues of the
polar this late in the game, put your mind at rest and pop the few extra bucks for the more
painless/robust installation of the PT or SRM.
--
==================
Kraig Willett www.biketechreview.com
================== [/B]

Yeah ... I may have to go that route, unfortunately.

I use the 720 off my bike as well (for running, hiking, and climbing), so I won't be ditching the 720 on my bike of course. I also very much like the PPP 4.0 software, soooo ... are there good utilities for merging in SRM info into an existing Polar .hrm file? That would be ideal ...

 

[Kraig Willett]

"jeffchrisope" <[email protected]> wrote in message
news:[email protected]...
>... are there good utilities for merging in SRM info into an existing Polar .hrm file? That would
>be ideal ...

try searching the wattage list archives on topica or asking the same question over there. someone
will know if one exists.

http://www.topica.com/lists/wattage/read

--
==================
Kraig Willett www.biketechreview.com
==================