AeroPod, PowerPod and Newton

After several attempts and many factory resets, i think i finally got a decent calibration ride done last night. When i'm on the climbs, the power numbers coming from PowerPod was higher than those from the Quarq Riken AL (Quarq Riken AL is paired to Edge 520 while PowerPod is paired to Edge 800). On flatter roads, the power numbers from both devices would match a big portion of the ride but there were times when the power numbers from PowerPod would drop by a huge margin. I'm talking about 80 to 90 watts even though i was putting out the same effort. John if you don't mind reviewing the IBR files and help me figure this out, that would be great!

Some details on my setup:
- Mounted on the handlebar (Please refer to the attached photo for the location)
- Bike has knobby 700x33c tires (The closest choice in the edit profile menu in Isaac is 700x32c)
- Using Garmin hub-mounted speed sensor
- No cadence sensor

I have attached the original IBR file and also one that i merged with the TCX file from the same ride with the Quarq Riken AL power meter.

No two power meters are going to match exactly 100% of the time. Each type uses its own kind of filtering, measurement, etc...

Remember too that PowerPod is very "sensitive" to any changes of ride position, particularly on downhills, where aerodynamics become very important (aero watts vary as the cube of bike speed).

The only place I see more than a momentary difference is on a slight downhill around mile 4.9 to 5.3, where PP reads low by about 30W. This could be a result of you changing ride position (sitting up a bit?), or even just shifting your weight a bit on the bike (altering its tilt calibration slightly).

I don't see any sustained portion of the ride where PP is reading 90W low...

I think your unit is working just fine.

I don't expect the power meters to match 100% of the time but the differences between the two (my bike computers were displaying 3-second power averages, not instantaneous power) on several occasions were alarmingly large, even though there was no change in my riding position (on the hood).

I did a DCRainmaker style power meter comparison with 30-second average:
I also did a 30-second average of power differences between the two. When the difference is negative, it means PowerPod is displaying higher power number than the Quarq. When the difference is more than zero, Quarq has a higher power number than PowerPod. Looking at the second chart, you can tell in general, PowerPod is reporting lower power number than the Quarq power meter, except when i was on the climbs (initially).
What would explain the PowerPod giving a higher power number on the climbs? I would estimate on average, i was getting 30 - 40 watts more on PowerPod.

Would an out-and back ride using the power number from the Quarq PM be able to better calibrate the PowerPod?

I saw your post after returning from dinner last night, so I did not have time to respond...

1) PowerPod factory settings are designed to give consistent and accurate results for the "average" rider. Most riders don't have two power meters on their bike, but if you happen to to have two PMs, then when doing this level of detailed power meter comparison, care has to be taken make sure PP settings are correct:
A. When you adjusted your profile, the wind scaling reading was inadvertently changed away from the factory value of 1.4, to 1.325. I put it back to 1.4.
B. Your knobby tires have higher rolling resistance, so I set your Crr a bit higher, to 0.0065.
C. Your CdA is a bit low. I set it to 0.40.

All of these tweaks are minor refinements which aren't necessary unless you're trying to get the last bit of comparative accuracy out of your PowerPod.

I've made these corrections in your ride file and profile, and attached them below.

2) Your ride is 24 minutes long. The first 8 minutes of any ride is where PowerPod does its fine-tuning. Your first climb was from minute 2 through minute 7. PowerPod was about 30W high until minute 5, when its calibration corrections began to factor in. You'll notice after minute 5, the numbers match nearly exactly, and in the second climb, from minute 12 through 20, PP was dead-on.

3) Knobby tires are "softer" than road tires (a great thing for comfort). However, softer tires also mean that any front-to-back movement of your body on the bike can cause small changes in tilt "offset" (the softer tires are a bit like shock absorbers). There are a few places on your ride (say, minute 8.5 to 9) where there is a momentary (but minor) difference. Very possibly this a result of a slight shifting of bike position.

4) Especially when using short filtering, the power numbers you see on your bike computer screen are going to be highly dependent on the internal averaging that is done inside the PM sensor. PowerPod's internal filtering algorithms are quite different from other PMs. PP intentionally reduces the amount of filtering (because of Powerstroke) so our numbers tend to jump around a bit more (look at your sprint at the end of the file, where PowerPod actually shows the within-the-stroke power variation, but Quarq does not). But that's also why we use Dynamic Power Smoothing. I think you'd find performance much better then.

The original files are pretty close, and they are even closer when the tweaks are added. NP, IF, TSS scores are identical, VI differs by 0.02, average watts are within 3W of each other, and max watts within 6%.

Regarding your original topic, "Inexplicable drop in power numbers". The answer to this one is simple: you changed PowerPod's filtering from Dynamic Power Smoothing to "zero second" filtering. On the zero second setting you're getting raw data out of the PowerPod. Raw data from ANY power meter is jumpy. The other guys apply filtering before sending it out, and there is no way to override their internal filter setting. I suggest you return your PowerPod to DPS.

Corrected ride file and profile are attached.

Thank you John for the detailed response. It's much appreciated, as usual.

In one of my earlier calibration attempts, the wind scaling reading was actually set by PP to be closer to 1.0.

I read this forum post by speedracer and saw that after his calibration, the wind scaling was set to 1.0 so that's why i thought that calibration attempt was not successful.

http://www.ibikeforum.com/viewtopic.php?f=93&t=3518

If i remember correctly, the 1.325 number came from the most recent calibration (Thursday night). It's also possible, like you wrote, that i have inadvertently changed the wind scaling.

If you look at the 30-second average power comparison i made, starting around 10:30 to 12:00 and then from 19:00 to 23:00, the power numbers between the two devices diverged quite a bit, with PP showing lower power numbers. From the IBR file, are you able to determine what caused this lower power number from PP?

http://www.ibikeforum.com/download/file.php?id=2761

I disabled dynamic smoothing because of DCRainmaker's review. What was the factory dynamic smoothing?

Another question. How do i use Isaac to adjust the ride file post-ride, like what you did?

I will apply your updated profile to my PP and will do a longer ride to test it. Thanks again!

Velocomp wrote:I saw your post after returning from dinner last night, so I did not have time to respond...

1) PowerPod factory settings are designed to give consistent and accurate results for the "average" rider. Most riders don't have two power meters on their bike, but if you happen to to have two PMs, then when doing this level of detailed power meter comparison, care has to be taken make sure PP settings are correct:
A. When you adjusted your profile, the wind scaling reading was inadvertently changed away from the factory value of 1.4, to 1.325. I put it back to 1.4.
B. Your knobby tires have higher rolling resistance, so I set your Crr a bit higher, to 0.0065.
C. Your CdA is a bit low. I set it to 0.40.

All of these tweaks are minor refinements which aren't necessary unless you're trying to get the last bit of comparative accuracy out of your PowerPod.

I've made these corrections in your ride file and profile, and attached them below.

I don't see a way to tweak the Crr and CdA for the profile. Do the tweaks you did only apply to the ride file and therefore can only be done post-ride? Is it possible to have these tweaks made to the profile?

ktula wrote:I read this forum post by speedracer and saw that after his calibration, the wind scaling was set to 1.0 so that's why i thought that calibration attempt was not successful.

http://www.ibikeforum.com/viewtopic.php?f=93&t=3518

If i remember correctly, the 1.325 number came from the most recent calibration (Thursday night). It's also possible, like you wrote, that i have inadvertently changed the wind scaling.

As far as I understood the wind scaling value is a constant one. Depending on the mount position the factory values are 1.4 (mount which is shipped with PP) and 1.7 (e.g. K-EDGE Combo Mount). Still I don't understand really why there is so much difference between the values. I hope to get an Answer from John here. From my point of view the 1.7 are a bit too high but since I cannot compare to another power meter I leave it now at that value.
The wind scaling of 1.0 was after I created the profile for the 1st ride. It is set by ISAAC - from my point of view it is done incorrectly as a chose "front mount".

The difference in wind factors is because the air pressure incident on the bike changes, depending on where the PP is reading it.

Post a .ibr ride file where you ride solo and we will figure things out for you