comparing power meters

User avatar
tomee
Posts: 297
Joined: Wed Oct 15, 2014 4:30 pm
Location: Vic

comparing power meters

Postby tomee » Tue Jan 23, 2018 2:03 pm

does anyone know a quick easy way to compare power stats from 2 devices for the same ride?

i know DC rainmaker has one but its $5 for 3 compares. Looking for a free alternative to compare 1 ride.
Unless someone has a sub i could use for this?

User avatar
MichaelB
Posts: 14752
Joined: Tue Feb 20, 2007 9:29 am
Location: Adelaide, South Australia

Re: comparing power meters

Postby MichaelB » Tue Jan 23, 2018 4:08 pm

Can't help, but the $5 from DCR sounds reasonable ?

Are the figures different, or just trying to compare accuracy and see which is better ?

User avatar
Alex Simmons/RST
Expert
Posts: 4997
Joined: Fri May 02, 2008 3:51 pm
Contact:

Re: comparing power meters

Postby Alex Simmons/RST » Tue Jan 23, 2018 4:35 pm

Quick, Easy, Cheap. Pick any two.

Depending on how the data was recorded, it can be a little problematic as it's quite tricky to sync data from difference devices.

Much depends on what you are seeking to ascertain about the data - there are various analytic techniques that can be used depending upon what you are trying to work out.

User avatar
kb
Posts: 2570
Joined: Sun Nov 13, 2011 3:22 pm

Re: comparing power meters

Postby kb » Tue Jan 23, 2018 4:41 pm

You could try hunting for a FIT to CSV converter (or install Golden Cheetah) and just graph them in Excel. May need some fudging to line up but that’s probably doable in Excel too.
Image

User avatar
nickobec
Posts: 2271
Joined: Fri Mar 27, 2009 12:51 am
Location: Perth or 42km south as the singlespeed flies
Contact:

Re: comparing power meters

Postby nickobec » Tue Jan 23, 2018 4:43 pm

If you like raw data and I mean raw data

Get Golden Cheetah (it is free), import your ride files, then export the rides as CSV files.

Now you can manipulate the data however you like.

Like opening the CSV files in excel, and use the tools there to compare each second of the ride, graph the two power outputs over time etc.

User avatar
tomee
Posts: 297
Joined: Wed Oct 15, 2014 4:30 pm
Location: Vic

Re: comparing power meters

Postby tomee » Tue Jan 23, 2018 6:05 pm

thanks for the quick replies guys.
MichaelB wrote:Can't help, but the $5 from DCR sounds reasonable ?

Are the figures different, or just trying to compare accuracy and see which is better ?
just trying to see how close the data is from both sources.

will give the golden cheetah and export to CSV a try.
i got a little overwhelmed when using golden cheetah for the first time

User avatar
Alex Simmons/RST
Expert
Posts: 4997
Joined: Fri May 02, 2008 3:51 pm
Contact:

Re: comparing power meters

Postby Alex Simmons/RST » Wed Jan 24, 2018 6:51 am

Attempting to precisely time sync data from two different cyclo-computers is problematic, they just never quite start and stop recording data at the same time, they all have quirks in how they deal with starting and stopping - some add an extra second or two (or three) of made up data while others trigger the "not pedalling" condition more quickly. Some have certain signatures in how they record certain data channels. The move to wireless ANT+ introduced a lot of such artefacts in the data recording process. It can be quite bad on some units, easily over inflating summation power data by several percent for certain types of rides (e.g. criteriums).

The old version of WKO, WKO+ had a feature that specifically allowed one to overlay data channel traces from more than one file but that feature doesn't exist in the same way in the new software.

Again, depending on what it is you are attempting to understand (it's not clear from your posts) there may be easier or more suitable methods.

One is simply to overlay the mean maximal power plot from each ride. That can highlight differences in at least the maximal reported values for all durations of the ride. While not covering all data - it does at least provide a first pass look at whether there are any obvious differences in peak values, and there often are, particularly at shorter duration end of the power duration curve. You can do this with appropriate software such as WKO4 or Golden Cheetah, although it's probably easier to export the chart data to Excel than attempt to get the software to overlay it.

Another example technique is plotting and overlaying the data using a virtual elevation method. This largely removes sync issues as it focuses on distance and calculated elevation - it's a great way to see at what point power files diverge from each other, since the method uses a cumulative process of calculating the energy balance for all the physics involved. It's quite a powerful diagnostic since it nails down where such divergences occur and that can help understand why it might be happening.

User avatar
Alex Simmons/RST
Expert
Posts: 4997
Joined: Fri May 02, 2008 3:51 pm
Contact:

Re: comparing power meters

Postby Alex Simmons/RST » Wed Jan 24, 2018 7:09 am

For those wondering, have a close look at how your meter manages data when you start/stop pedalling. Does it seem to have a repeated values when you stop? On which data channel? Does it treat power and cadence traces differently? For those with dual sided metering, what happens to your power balance data at those times? How is that affecting the summation figures for a ride?

When you start to look into this stuff, you realise there's a bit going on with the data and sometimes the various metrics can be misleading if not considered with due care.

User avatar
tomee
Posts: 297
Joined: Wed Oct 15, 2014 4:30 pm
Location: Vic

Re: comparing power meters

Postby tomee » Wed Jan 24, 2018 9:46 am

great info Alex, thank you for taking the time to write that up.

i should of put this information in the first post.
im wanting to compare a smart trainer power in this case a Neo with a Stages in a higher temperature environment (18 degrees - 28 degrees) over an hour long session.

figured out a rudimentary solution:
Export each as a gpx format file, and then open them in excel.
Pick the time scale from 1, and the power data from both and paste onto 1 sheet.
Make simple xls line chart.

User avatar
Alex Simmons/RST
Expert
Posts: 4997
Joined: Fri May 02, 2008 3:51 pm
Contact:

Re: comparing power meters

Postby Alex Simmons/RST » Wed Jan 24, 2018 12:03 pm

tomee wrote:great info Alex, thank you for taking the time to write that up.

i should of put this information in the first post.
im wanting to compare a smart trainer power in this case a Neo with a Stages in a higher temperature environment (18 degrees - 28 degrees) over an hour long session.

figured out a rudimentary solution:
Export each as a gpx format file, and then open them in excel.
Pick the time scale from 1, and the power data from both and paste onto 1 sheet.
Make simple xls line chart.
Post your results, be interested to see what you find.

It would not surprise me to see some divergence.

The problem then becomes understanding the cause of the difference, which in the case of the set up you mention could be due to either of or both of technical and biological factors.

User avatar
MichaelB
Posts: 14752
Joined: Tue Feb 20, 2007 9:29 am
Location: Adelaide, South Australia

Re: comparing power meters

Postby MichaelB » Wed Jan 24, 2018 2:07 pm

tomee wrote: ....in a higher temperature environment (18 degrees - 28 degrees) over an hour long session.
.....
If 28 degrees is high, do you live in Hobart or something :?:

28 is just getting warm !! :lol: :lol:

User avatar
tomee
Posts: 297
Joined: Wed Oct 15, 2014 4:30 pm
Location: Vic

Re: comparing power meters

Postby tomee » Wed Jan 24, 2018 3:52 pm

MichaelB wrote:
tomee wrote: ....in a higher temperature environment (18 degrees - 28 degrees) over an hour long session.
.....
If 28 degrees is high, do you live in Hobart or something :?:

28 is just getting warm !! :lol: :lol:
:D warm for 5:30am

my trainer sits in the garage and after the couple of days over 38 the garage got warm

User avatar
tomee
Posts: 297
Joined: Wed Oct 15, 2014 4:30 pm
Location: Vic

Re: comparing power meters

Postby tomee » Mon Jan 29, 2018 12:13 pm

although the excel speradsheet method worked it was very time consuming and in the end i went to DCRainmakers analyzer lol

awesome tool!

User avatar
MichaelB
Posts: 14752
Joined: Tue Feb 20, 2007 9:29 am
Location: Adelaide, South Australia

Re: comparing power meters

Postby MichaelB » Mon Jan 29, 2018 1:34 pm

tomee wrote:although the excel speradsheet method worked it was very time consuming and in the end i went to DCRainmakers analyzer lol

awesome tool!
Can you share the results ? Would be interesting

User avatar
tomee
Posts: 297
Joined: Wed Oct 15, 2014 4:30 pm
Location: Vic

Re: comparing power meters

Postby tomee » Mon Jan 29, 2018 3:03 pm

MichaelB wrote:
tomee wrote:although the excel speradsheet method worked it was very time consuming and in the end i went to DCRainmakers analyzer lol

awesome tool!
Can you share the results ? Would be interesting
sure, here was a lower intensity sweet spot session i did yesterday in the garage which was 35+ degrees
https://analyze.dcrainmaker.com/#/publi ... 38ae542208

although i will have to do another test as the two files i did for compare the zwift session had 3 second smoothing, so the next time i compare i will switch it to instant power on zwift.

also the spikes are from ANT+ dropouts

User avatar
Alex Simmons/RST
Expert
Posts: 4997
Joined: Fri May 02, 2008 3:51 pm
Contact:

Re: comparing power meters

Postby Alex Simmons/RST » Tue Feb 06, 2018 5:13 pm

It would seem the primary differences are at low power and at high power.

User avatar
tomee
Posts: 297
Joined: Wed Oct 15, 2014 4:30 pm
Location: Vic

Re: comparing power meters

Postby tomee » Mon Feb 12, 2018 12:52 pm

Alex Simmons/RST wrote:It would seem the primary differences are at low power and at high power.
yeah and im seeing the same pattern on all my other rides regardless of temperature. neo reading lower during low wattage and slightly higher in the higher ranges

not sure whether its to do with drive train losses etc

User avatar
Alex Simmons/RST
Expert
Posts: 4997
Joined: Fri May 02, 2008 3:51 pm
Contact:

Re: comparing power meters

Postby Alex Simmons/RST » Mon Feb 12, 2018 5:00 pm

tomee wrote:
Alex Simmons/RST wrote:It would seem the primary differences are at low power and at high power.
yeah and im seeing the same pattern on all my other rides regardless of temperature. neo reading lower during low wattage and slightly higher in the higher ranges

not sure whether its to do with drive train losses etc
There are a range of possible reasons for the variability in reported power, as well as a difference.

Given one of the devices is a unilateral Stages, there is every chance that your power output is asymmetric across the power range and this is being reflected in Stages data, which assumes perfect symmetry at all times and in all conditions and scenarios. We know this is never the case, everybody has some level of asymmetry and variable asymmetry is also normal.

Of course there is nothing to suggest to Neo is spot on either (other than their marketing claims), some have reported it varies for them depending on cadence, and I doubt it has been subjected to testing of scientific rigour.

Drivetrain losses can account for 1-3% of total power as well, so devices at crank/pedals should read higher than those measuring at the cog/hub/wheel/tyre.

Who is online

Users browsing this forum: jasonc, Mr Purple