PowerPod over reading on all bikes

[darthmonkey]

My personal struggle with my PowerPod and its accuracy have a well documented, long and jaded history.
Very long story short, I've been laboring under the impression that after numerous calibration rides, I'm getting somewhere near accurate results, (in my estimation about 10% out) an example of which is a 309w FTP (ish) ride.
However, after purchasing a Kickr Core indoor smart trainer, which boasts a +-3% accuracy, and doing an FTP ride, I find that my FTP is more like 240w.
Let's be generous, and say it's (even though it has its own calibration) it's -3% out, so it's actually 247w, that would put the (remember multiple calibrations) PowerPod out by 22%.

Given this, it appears I need to use Issac software to adjust the profiles of all three bikes (even though I've been told in this forum that I shouldn't need to). I'm unsure exactly what I'll need to change, but was hoping for some assistance.
I've attached the profiles for all three bikes, Road, Cyclocross, and MTB, along with my latest road calibration ride on the cyclocross bike.

Although I know it's estimation based on an assumption, I'd like to be able to get a 230w -250w FTP, rather than the 290w -310w ones I've been getting.

I would also like to calibrate the PowerPod against the Kickr, but as I think it'd take a lot of faffing, and another weekend 'on the tools' I'll leave that for another post.

Thanks in advance.

P.S. There may be duplicate attachments, as it doesn't seem possible to add two at a time, then when I do, it misreports which attachment I've successfuly uploaded.

[Velocomp]

There was only one ride file attached (the other ride file attached was a duplicate of the first) so I can comment only on that ride file. If you have more ride files, from other bikes, that you'd like me to look at, either post them or email them to me at jhamann@velocomp.com

Also, you attached only one profile.

The ride file you posted indicates a solid calibration. So, I have no reason to believe your setup is wrong.

This is only a very short ride (4 minutes) but it includes enough to do an independent check, using this website: http://www.aerocoach.com.au/power-from-speed/.

Basically, I select a section of ride where your parameters are nearly constant (speed, wind speed, slope) and load them into the website. I check to see if the website results are close to PP measurements.

For the section of ride I chose, PP reads 178.8W and the website reads 181.0W. Extremely close...

You were "on the bike" during this entire ride: if you were riding on the flats (as the data suggests), if you weren't battling much ground wind (as the data suggests) and you were riding at a constant speed of around 30km/h (as the data suggests), and your height and weight parameters are correct, then I have no reason to believe PP data is incorrect.

So, if you did an FTP test on this bike, with this profile, I would have see no obvious reason to disbelieve it.

I can't comment on the Kickr Core results; there is no data from it, and no way to independently verify its measurements. This raises an obvious question: why should the Kickr be believed?

[darthmonkey]

Thanks very much John for checking my ride data, and running those calculations. I tried my own calculations, and came up with wildly different results based on minute differences in wind speed, and gradient.
You'd expect these to play a factor, but after adding my known and close estimate settings for a section I know well, a difference of 0.3% gradient changes the required watts to maintain an example of 36.2km/h over 1.57km, from 228w to 260w, a 13% difference. Add to this that I know I had a tailwind, but can only guessimate from the location away from the measured wind and direction of 17km/h from the SW, so halving it to provide an 8.5km/h tailwind, gives me a wildly different 191w.
In short using that calculator is fantastic, assuming you know your wind speed, and exact gradient (average is not always great even for tiny slopes of less than 0.3%.
This means that if I estimate the unknown variables with a tiny difference in either gradient or wind speed, I can say I was putting out anything from 180w (-10kw/h wind speed) to 264w (0kph wind speed) without even touching the gradient.
Assuming I got all the other esitmates correct, (I even calculated air density), that wold put my Strava estimate of power closer to the 'aerocoach' site than my PowerPod by about 50w for an example ride.
With all that said, I'm (personally) happy that my calibrated Kickr (checked by a Wahoo tech) is more accurate than the PowerPod, so would like to adjust my profile (we'll start with just the one I sent you), so that it more accurately estimates the Kickr's FTP readings of 240w.

What settings should I change in order to reduce the general PowerPod readings by 50w in all conditions?

[Velocomp]

Thanks very much John for checking my ride data, and running those calculations. I tried my own calculations, and came up with wildly different results based on minute differences in wind speed, and gradient.
You'd expect these to play a factor, but after adding my known and close estimate settings for a section I know well, a difference of 0.3% gradient changes the required watts to maintain an example of 36.2km/h over 1.57km, from 228w to 260w, a 13% difference. Add to this that I know I had a tailwind, but can only guessimate from the location away from the measured wind and direction of 17km/h from the SW, so halving it to provide an 8.5km/h tailwind, gives me a wildly different 191w.
In short using that calculator is fantastic, assuming you know your wind speed, and exact gradient (average is not always great even for tiny slopes of less than 0.3%.
This means that if I estimate the unknown variables with a tiny difference in either gradient or wind speed, I can say I was putting out anything from 180w (-10kw/h wind speed) to 264w (0kph wind speed) without even touching the gradient.
Assuming I got all the other esitmates correct, (I even calculated air density), that wold put my Strava estimate of power closer to the 'aerocoach' site than my PowerPod by about 50w for an example ride.
With all that said, I'm (personally) happy that my calibrated Kickr (checked by a Wahoo tech) is more accurate than the PowerPod, so would like to adjust my profile (we'll start with just the one I sent you), so that it more accurately estimates the Kickr's FTP readings of 240w.

What settings should I change in order to reduce the general PowerPod readings by 50w in all conditions?

I took another look at your ride file.

It is only 4 minutes long.

During the first 2 minutes your PP was recalibrating the accelerometer (your PP mounting angle changed from its prior ride). Because calibration was off, during the first 2 minutes the watts there were way too high. This skewed your average watts for the entire 4 minute ride to 296W.

In the final 2 minutes, where calibration was correct (and where I analyzed your file) average watts were 165W. This seems to be in line with what you are expecting.

To prevent watts errors for the first 2 minutes of your rides you can do the following:

1) make sure PP is attached securely to the bike and cannot rotate

2) do not remove PP from the bike between rides

3) at the end of the ride, click the button 5 times (Solid Yellow/Green/Red/off) to turn off PP. This will prevent "wrong" data from being added to PP when you're transporting your bike

Of course, your bike/rider weight, wheel circumference, tire type, and rider height/ride position also needs to be correct. I don't see anything in your profile information that suggests problems with those parameters.