iAero vs. CinQo vs. PT

[travispape]

Here's a power meter comparison from my ride this morning. This was a typical result, not the closest match and not the worst. The tilt correction is unit #1 in this post: http://www.ibikeforums.com/viewtopic.php?f=4&t=439

main window.png (67.12 KiB) Viewed 6632 times

full ride.png (56.79 KiB) Viewed 6627 times

mi 00.png (49.38 KiB) Viewed 6630 times

mi 05.png (46.73 KiB) Viewed 6629 times

mi 10.png (47.92 KiB) Viewed 6626 times

mi 15.png (48.65 KiB) Viewed 6627 times

mi 20.png (47.4 KiB) Viewed 6624 times

mi 25.png (48.35 KiB) Viewed 6625 times

mi 30.png (48.18 KiB) Viewed 6625 times

mi 35.png (48.07 KiB) Viewed 6623 times

mi 40.png (48.59 KiB) Viewed 6622 times

mi 45.png (49.63 KiB) Viewed 6628 times

mi 50.png (48.7 KiB) Viewed 6627 times

mi 55.png (48.98 KiB) Viewed 6625 times

mi 60.png (48.46 KiB) Viewed 6622 times

mi 65.png (46.4 KiB) Viewed 6621 times

Code: Select all

------------------------------------------------------------------------
                              IBIKE STATS
------------------------------------------------------------------------

Ride:      Distance: 68.23 mi 
           Duration: 3:39:52
         Start Time: 8/2/2008 7:55:10 AM
         Avg Temper: 85 degF
         Start Pres: 1008.3 mbar
            Avg CdA: 0.350 m^2
                Crr: 0.00700
Speed:          Avg:  18.62 mi/h
                Max: 37.1 mi/h
Wind:           Avg:  19.40 mi/h (dist-based)
                Max: 39.0 mi/h
     wrt Road:  Min: -13.0 mi/h (- is tail; + is head)
     wrt Road:  Avg:  -0.17 mi/h (dist-based)
     wrt Road:  Max: 9.6 mi/h
Slope:          Min:  -8.68%
                Avg:  0.01% (dist-based)
                Max:  11.60%
Elevation:      Min:  155 ft
                Avg:  278 ft
                Max:  435 ft
Cadence:        Avg: 74.9 rpm (ignoring zeros)
                Max: 99 rpm
HR:             Avg: 150.9 bpm (ignoring zeros)
                Max: 170 bpm

------------------------------------------------------------------------
                  IBIKE COMPARED TO ANALYTICCYCLING.COM
------------------------------------------------------------------------
Plug the following parameters(*) into the static forces calculator at 
http://www.analyticcycling.com/ForcesPower_Page.html

        Effective Frontal Area(#)    = 0.701 m^2
        Drag Coef (keep the default) = 0.5 
        Air Density                  = 1.150 kg/m^3
        Weight of Rider and Bike     = 95.7 kg
        Coef of Rolling Resistance   = 0.0070
        Slope                        = 0.00008 rise/run
        Speed                        = 8.325 m/s
        Pedal Cadence                = 74 rev/min
        Crank Length(**)
        Effective Pedaling Range(**)

    Compare the avg power result from Analytic Cycling to the comparable
    result from the iBike 2 calculation engine:

        Propulsion Power = 171.56 W  <---<<< compare to AnalyticCycling(##)
        Rider Power      = 175.06 W

    *  The AnalyticCycling calculator  assumes perfectly calm wind 
       conditions.  Further, it does not account for any accelerations or
       decelerations.
    #  (Drag Coefficient) x (Effective Frontal Area) = CdA
    ** Crank Length and Eff Pedaling Range do not factor into the avg
       power calculation
    ## The AnalyticCycling calculator ignores drive train losses.

------------------------------------------------------------------------
               POWER METERS COMPARED TO EQUATIONS OF MOTION
------------------------------------------------------------------------
           Ride File   Min Pow   Avg Pow   Max Pow			
 AnalyticCycling.com	     0.00	   175.06	   175.11
          Speed Elev	     0.00	   199.14	   960.30
     Speed Wind Elev	     0.00	   198.13	  1010.68
    Speed Wind Slope	     0.00	   195.44	   902.88
           iBike Pro	     0.00	   190.91	   779.22
Quarq CinQo (iB DFPM	     0.00	   185.67	  1028.00
               PT SL	    -1.00	   190.92	   928.92

------------------------------------------------------------------------
                                  R^2
------------------------------------------------------------------------
                 |iBike Pro|Quarq Cin|    PT SL
        iBike Pro|    1.000|    0.730|    0.818
Quarq CinQo (iB D|    0.738|    1.000|    0.832
            PT SL|    0.810|    0.820|    1.000

R^2 calculation is row w.r.t. column.

Before making an adjustment on the wind offset, the iAero average power was 187.4 W

I have turned on the auto-zero feature on the PT. I didn't rezero the CinQo during the ride and as the temperature rose it started under-report power. I should have rezoroed it--I just forgot. I did zero it before starting the ride.

Travis

[travispape]

Here is the ride data:

iBike_2008_08_02_0755_68_Miles_E_Spec.csv

iAero data w/ CinQo as DFPM

(1.44 MiB) Downloaded 306 times

PT 080802.csv

PT data

(412.23 KiB) Downloaded 284 times

[travispape]

Here is data from a recent commute.

main.png (51.7 KiB) Viewed 6611 times

tilt.png (46.4 KiB) Viewed 6608 times

pmc.png (50.26 KiB) Viewed 6609 times

Code: Select all

------------------------------------------------------------------------
               POWER METERS COMPARED TO EQUATIONS OF MOTION
------------------------------------------------------------------------
           Ride File   Min Pow   Avg Pow   Max Pow			
 AnalyticCycling.com	     0.00	   214.19	   214.88
          Speed Elev	     0.00	   257.75	   904.39
     Speed Wind Elev	     0.00	   242.82	   898.98
    Speed Wind Slope	     0.00	   242.56	   790.84
           iBike Pro	     0.00	   240.95	   758.54
Quarq CinQo (iB DFPM	     0.00	   247.29	   839.00
               PT SL	    -0.02	   240.93	   867.25

------------------------------------------------------------------------
                                  R^2
------------------------------------------------------------------------
                 |iBike Pro|Quarq Cin|    PT SL
        iBike Pro|    1.000|    0.830|    0.908
Quarq CinQo (iB D|    0.833|    1.000|    0.881
            PT SL|    0.910|    0.881|    1.000

R^2 calculation is row w.r.t. column.

Both the PT and CinQo were zeroed before the ride.

iBike_2008_08_01_0854_7_Miles_E_Spec.csv

iAero + CinQo data

(166.19 KiB) Downloaded 288 times

PT 080801am.csv

PT data

(37.24 KiB) Downloaded 262 times

[travispape]

Another recent commute:

main.png (51.99 KiB) Viewed 6599 times

tilt.png (46.25 KiB) Viewed 6596 times

pmc.png (49.15 KiB) Viewed 6595 times

Code: Select all

------------------------------------------------------------------------
               POWER METERS COMPARED TO EQUATIONS OF MOTION
------------------------------------------------------------------------
           Ride File   Min Pow   Avg Pow   Max Pow			
 AnalyticCycling.com	     0.00	   220.78	   221.50
          Speed Elev	     0.00	   253.13	   870.82
     Speed Wind Elev	     0.00	   241.90	   872.43
    Speed Wind Slope	     0.00	   242.88	   991.47
           iBike Pro	     0.00	   244.02	   957.28
Quarq CinQo (iB DFPM	     0.00	   247.11	   908.00
               PT SL	    -0.58	   241.92	   920.19

------------------------------------------------------------------------
                                  R^2
------------------------------------------------------------------------
                 |iBike Pro|Quarq Cin|    PT SL
        iBike Pro|    1.000|    0.728|    0.860
Quarq CinQo (iB D|    0.707|    1.000|    0.706
            PT SL|    0.850|    0.709|    1.000

R^2 calculation is row w.r.t. column.

Both the PT and CinQo were zeroed before the ride.

iBike_2008_07_30_1910_7_Miles_E_Spec.csv

iAero + CinQo data

(162.7 KiB) Downloaded 265 times

PT 080730pm.csv

PT data

(36.44 KiB) Downloaded 231 times

[travispape]

Yet another recent commute. I'm just trying to show that these good results are the norm.

main.png (51.46 KiB) Viewed 6588 times

tilt.png (45.72 KiB) Viewed 6588 times

pmc.png (50.75 KiB) Viewed 6588 times

Code: Select all

------------------------------------------------------------------------
               POWER METERS COMPARED TO EQUATIONS OF MOTION
------------------------------------------------------------------------
           Ride File   Min Pow   Avg Pow   Max Pow			
 AnalyticCycling.com	     0.00	   216.99	   217.69
          Speed Elev	     0.00	   251.38	   861.58
     Speed Wind Elev	     0.00	   242.54	   887.38
    Speed Wind Slope	     0.00	   237.69	   832.00
           iBike Pro	     0.00	   237.58	   809.99
Quarq CinQo (iB DFPM	     0.00	   241.99	   752.00
               PT SL	     0.00	   237.59	   719.44

------------------------------------------------------------------------
                                  R^2
------------------------------------------------------------------------
                 |iBike Pro|Quarq Cin|    PT SL
        iBike Pro|    1.000|    0.811|    0.911
Quarq CinQo (iB D|    0.818|    1.000|    0.867
            PT SL|    0.912|    0.863|    1.000

R^2 calculation is row w.r.t. column.

Both the PT and CinQo were zeroed before the ride.

iBike_2008_07_29_0849_7_Miles_E_Spec.csv

iAero + CinQo data

(167.05 KiB) Downloaded 274 times

PT 080729am.csv

PT data

(37.46 KiB) Downloaded 286 times

[rruff]

They do match very well most of the time. Do you have any idea what was going on during the times when the iAero reading was off compared to the other two?

[travispape]

They do match very well most of the time. Do you have any idea what was going on during the times when the iAero reading was off compared to the other two?

Keep in mind that what is going on here is that the iBike is making instantaneous power measurements by dynamically summing frictional losses, work done against gravity, and acceleration. The iBike is fundamentally different technology than the strain-gauge technology of both of the DFPM power meters shown, so the fact that the iBike power matches the other guys so well is no small feat. It's like balancing a check book where the DFPMs are just tracking the cash in the bank, but the iBike is keeping track of all the checks and deposits reconciling balance sheet every second.

When it's time to pick nits and focus in on the differences, some of the possibilities include

* rider out of nominal position, which impacts both CdA and tilt due to changing weight distribution
* wind turbulence messing up the wind speed measurement
* cross-wind changing the rider's effective CdA
* a gust of tail wind can add power since the iBike can't measure wind passing the rider
* significantly rougher or smoother road surface than road used for coast-downs

In general, the iBike has to do more filtering on its sensor readings than the DFPM power meters do for their one sensor, so again, it is no small feat for the iBike to match up so well with the DFPM power vs. time plots. When you compare the results over a longer time scale, a well-calibrated iBike will be as good as or even better than the DFPM. For example, the CinQo started under-reporting as the temperature rose and its zero offset became invalid. (BTW, I didn't do that on purpose--I just forgot to rezero it later in the ride.) Since the iBike is different technology, any differences would be most visible in instantaneous power plots such as those above which is why we show them--it validates the iBike technology. You don't have to have a DFPM power meter installed in your drive train to get accurate power readings--the iBike is accurate on both long and short time scales.

[Ratman]

Hi, Travis,

May I assume that all of the screens you've shown are post-iBike2 tilt profile, wind and power processing? Also, when (and how many?) wind offset calibrations are you typically doing during these rides?

Thanks!

[travispape]

Hi, Travis,

May I assume that all of the screens you've shown are post-iBike2 tilt profile, wind and power processing? Also, when (and how many?) wind offset calibrations are you typically doing during these rides?

Thanks!

Yes, these are post-processed. You can download the CSV file to view the results before processing. You'll see they are close.

I only did one wind offset cal before beginning the ride in each case. In fact, I don't normally redo the wind offset cal for my evening commute so it is using the wind offset cal result from the morning.

Travis