Power meter comparision (iAero vs DFPM) - incredible

[wellmt]

I did a calibration of my iAero a couple of days ago using a 4 mile cal ride and a dozen coast downs. It was done in less than ideal conditions.

Yesterday I fitted a DFPM (Powertap) and rode a 20 min interval with 5 min cool down as an experiment to see how accurate that calibration was... (it was going to be multiple intervals but the weather got bad).

When I look at the stats in iBike I can press the button to view either the iAero's measurements or the Powertaps....I was amazed to see that if I take the 20 min interval on it's own, I get a tiny 1.9 watt average difference. If I include the 5 min cool down as well the difference is 0 (average watts). The only difference is the max watts measurement.

Remember this iAero was calibrated without using a Powertap! Seems too good to be true? Am I missing something. If not then I am stunned!

I'll post some pictures up tomorrow.

[dtrousdale]

I just installed a DFPM (CinQo Saturn), and I had the opposite experience. So far I've assumed I probably did not do something right and will perform all the setup and calibration again when I get a chance. I did only the 4-mile ride and no coast-downs, per the instructions as I understood them.
Below is a picture of my first ride. Unless I'm interpreting this wrong, in the bottom panel, it looks to me that there far more than I expected, and pretty significant (=100W and -100W), CinQo spikes relative to the iAero Gen3.

iAero-CinQo_comparison.jpg (98.26 KiB) Viewed 13526 times

[racerfern]

You've got some setting wrong, but I don't know what, since I don't have a DFPM. The top graph pretty much lines up. I'm looking for Coach Boyd or someone else who is knowledgeable to jump in.

[wellmt]

I just installed a DFPM (CinQo Saturn), and I had the opposite experience. So far I've assumed I probably did not do something right and will perform all the setup and calibration again when I get a chance. I did only the 4-mile ride and no coast-downs, per the instructions as I understood them.

We're not comparing the same things here, although the results are interesting.

1. I'm talking about the average power over a ride being incredibly close/identical - which frankly I still find stunning (what are your overall averages?)
2. My iBike was calibrated using a 4 mile ride, 12 coastdowns and without estimated CRR. Without using a Powertap wheel. Your cal method was different (cal with a powermeter)
3. If I view the same overlay graphs as you, I agree the power outputs do not line up either - but the overall measurement is the same.

Regarding the second point above:

I attempted a cal ride afterwards using the same method as you (a 4 mile cal ride) with the Powertap wheel. The resulting profile was nonsense. I'm putting this down to rain in the wind port. I'd have to try again

[wellmt]

OK here are my comparisions - again let me emphasize that the iBike was calibrated without a DFPM.

compare1.jpg (113.61 KiB) Viewed 13481 times

Here is the comparision. I can't see how it is possible to make them align more closely?

compare2.jpg (103.76 KiB) Viewed 13479 times

[wellmt]

Sorry, don't mean to take over the forum! But I've just done another quick flat ride and the numbers match up so well - I'm really pleased.

Looks like I won't be needing to lug a Powertap wheel around. Well done Velocomp!

powcomp2a.JPG (78.28 KiB) Viewed 13422 times

Comparision:

powcomp2.JPG (126.34 KiB) Viewed 13424 times

I'll try a hilly ride next time, I think that will really show any differences. I think it's the high power points that cause any differences, the Powertap is always reading higher.

[dtrousdale]

All I can say is that I wish my results looked like yours. No such luck, however.
Assuming that my first cal ride and profile were flawed, I did a second cal ride today on the same road as I did my first cal ride using the same protocol set forth in the doc: 4-mile ride with no CDs when done with a DFPM. Everything was the basically the same--same tt bike, same clothing, Cinqo Saturn, iAero G3, etc.--except that I was one pound lighter, the road segment I used was a little different (but mostly the same), and there may have been slightly less pressure in the tires. After the cal ride, I rode up-and-down the same road afterwards so I could have a fresh sample ride to which I could apply both my old and my new profile, went home and processed the cal ride to create a new profile, and finally applied both profiles to the ride I did today. The differences look much worse with the new profile. The old profile doesn't look so bad right now
Here are the results for my first profile:
Aero 0.418
CdA 0.264
Fric 10.426
Crr 0.0037
Here are the results for my second profile:
Aero 0.150
CdA 0.101
Fric 27.155
Crr o.0041

Here are some pictures, plus I'll attach my ride file:

Exhibit A: PM comparison in iBike3 using the old profile

pm-compare_profile-01+dfpm.jpg (166.18 KiB) Viewed 13411 times

[dtrousdale]

Oooops! Don't know what happened. I wasn't done yet with my post.

Here is Exhibit B: With the first profile, the stats don't look too bad:

profile-01+dfpm.jpg (48.92 KiB) Viewed 13372 times

Now here's where things get interesting. The differences are indeed "incredible" but not in a positive sense. Here is Exhibit C, comparing PMs in iBike3--Same ride but using the profile I created today:

pm-compare_profile-02+dfpm.jpg (166.65 KiB) Viewed 13395 times

Here is Exhibit D, comparing stats using the profile I created today against the CinQo:

profile-02+dfpm.jpg (71.07 KiB) Viewed 13371 times

Looks like I have to post my ride file separate since the max is 3 files per post. I'd appreciate any insights as to what is wrong with all this.

[dtrousdale]

Sorry for the fragmented post. Here is the ride file to which I referred in my previous posts, the ride with my strange new profile:

[wellmt]

There is definitely something wrong with your second profile. Even for a TT bike that CDA is too low, that's what is causing the low numbers. The question is why?

Is there any chance the wind port was blocked in your cal ride, do you have the Remote Wind Sensor?

For the record, my cal ride with a DFPM gave me a CDA of .22 which is too low for a regular road bike.

[dtrousdale]

There is definitely something wrong with your second profile. Even for a TT bike that CDA is too low, that's what is causing the low numbers. The question is why?

Is there any chance the wind port was blocked in your cal ride, do you have the Remote Wind Sensor?

I, too, noticed that the CdA is way too low. The iAero is located exactly where it was long before I ever got the DFPM, but I never got such an unbelievable number before. I'm pretty meticulous about calibration: there was no blockage, tilt/wind cal performed immediately before cal ride, beautiful sunny day, little traffic, road is smooth, straight as an arrow with little rollers and no stop signs or traffic light. I do own a RWS but have not installed it yet because I did not want to change everything at once to simplify any necessary troubleshooting.

BTW you mentioned that you got a profile that was "nonsense" when you tried to calibrate using your PT. You mentioned rain blocking the wind port as a possible cause of the problem. Have you tried this again? Now I'm wondering if calibration with DFPM-only or with no CDs does not always work.

I hope iBike Sports Tech Support can weigh in here at some point.

[lorduintah]

You might send direct mail to John or Allan rather than waiting for them to read through the forums.

Tom

[wellmt]

BTW you mentioned that you got a profile that was "nonsense" when you tried to calibrate using your PT. You mentioned rain blocking the wind port as a possible cause of the problem. Have you tried this again? Now I'm wondering if calibration with DFPM-only or with no CDs does not always work.

I hope iBike Sports Tech Support can weigh in here at some point.

Unfortunately the weather has conspired to stop me. Having said that - I'd be doing it for kicks as the non-DFPM obtained profile is so good.

I do need to get a profile for my Winter/bad weather bike, so I will be trying again in the near future. Your conditions for obtaining a profile sound really good compared to mine, weird.

[coachboyd]

Sorry for the fragmented post. Here is the ride file to which I referred in my previous posts, the ride with my strange new profile:

Sorry for the delay, I was away in a VERY rural area of Western North Carolina this past weekend where the words cell phone and internet have no meaning to the locals.

But I just took a look at this file and there are a few things that was causing the results you saw. The first thing was that you had a riding tilt in your profile of -2.4%. This was causing you to have a VERY high friction number in your profile and counteracting it with a very low cda. I changed the riding tilt to a more real world number and that made the results a lot better. The new profile is embedded in the file attached here. Open this file and then go to profiles->edit profiles->extract, and the new profile will be added to your list of profiles.

The other thing that worries me is your tilt compared to slope (the green line). This is going all over the place meaning that you aren't getting good slope data. I can correct it afterwards in the software, but on the road the slope is always trying to 'catch up' to the barometric pressure data. More than likely this is a case of your mount not being secured enough (or in an area where it can flex a lot). If the iBike is constantly flexing up and down a lot, then this will cause the wattage readings to be all over the place. Can you post a picture of how you have your iBike mounted on your TT bike?

[dtrousdale]

Thanks for getting back. Here are some more questions and the pics you requested:

The first thing was that you had a riding tilt in your profile of -2.4%. This was causing you to have a VERY high friction number in your profile and counteracting it with a very low cda. I changed the riding tilt to a more real world number and that made the results a lot better.

Not sure how to handle this in the future. The nearest place I can find where I can get a really good 2-miles out-and-back is 30 minutes away by car. Once I get there, I have not been able to find a reasonably flat surface on which to do the tilt calc that now has to be done no more than 30 minutes before doing the cal ride. It took several tries, but I did get a "good tilt" message. Does this mean I cannot always rely on the tilt calc function?

The other thing that worries me is your tilt compared to slope (the green line). This is going all over the place meaning that you aren't getting good slope data. I can correct it afterwards in the software, but on the road the slope is always trying to 'catch up' to the barometric pressure data. More than likely this is a case of your mount not being secured enough (or in an area where it can flex a lot). If the iBike is constantly flexing up and down a lot, then this will cause the wattage readings to be all over the place. Can you post a picture of how you have your iBike mounted on your TT bike?

As you can see my photos below, I have a Topeak Extender attached to my step and below the aero bars as suggested in the iAero doc. And it is on there tight.

I'm seriously considering changing the location of the iAero to accommodate the RWS that I recently purchased. The problem with the RWS is that it does not stay inserted in the iAero if it is bent at too much of an angle in order to get the other end alongside the handlebars. I tried putting it on the stem without the the Topeak Extender and above the Aero bars, but it was impossible to remove the iAero from the mount with out first removing the left arm pad every time. Now I'm thinking of locating the iAero using the Topeak Extender on the stem, further back on the Extender (not so far out to the end as I have it now), and above the aero bars. I'm hoping this will allow me bend the RWS tubing less such that it does not pop out of the iAero but does allow me to route the tubing over the end of the handlebar. What do you recommend?

P1012357.jpg (114.55 KiB) Viewed 13198 times

P1012360.jpg (89.34 KiB) Viewed 13196 times

[tonyt]

I would suggest (as you said you would do) moving it as close to the stem as possible on that extender.


For the on-topic bit - I agree. I thought the intervals would throw it off a little - not even close!

t.png (127.49 KiB) Viewed 13137 times

[wellmt]

OK, I can't get the untweaked and independently calibrated outputs of the Powertap/iBike to not match*

Just wanted to finish my input on this thread by saying once calibrated correctly, the iBike just works. I'm an iBike believer.

[* The first five mins are sometimes out, but the iBike's tilt correction feature even sorts this problem out]

[dtrousdale]

I will add as my coda to this thread that, thanks to Coach Boyd, I'm now getting similar matching results from my iAero/Cinqo combo. After relocating and attaching my iAero in a way similar to Boyd's setup, this all finally seems to work the way I had hoped.
This also seemed to have solved another problem that I mentioned on another thread and that that was also causing a lot of frustration: the iAero frequently losing the Cinqo signal. The iAero is now physically closer to the Cinqo and has fewer metallic obstructions between the iAero and the Cinqo.
Finally, I abandoned a step in DFPM calibration procedure (described in the iAero doc) that rarely worked and turns out to be completely unnecessary. This was another source of frustration mentioned in another thread and on which I spent way too much time trying to figure out what I was doing wrong. Thanks to Jim @ Quarq and Coach Boyd, I now rely only on what Quarq calls Auto Zero (spinning backwards 4-5 times) and do not bother with Manual Zero as implemented with the iAero. The iAero doc seems to imply that one must do each of these in succession, but that is not true in this case.