If your intention is to use it for CdA testing... it won't do that, since it lacks testing software. It just gives you an inaccurate real-time value.

You're better off getting a Weatherflow Weathermeter and a mount (and a phone if you don't have an android) and using the CdACrr App: https://forum.slowtwitch.com/...oid_app__P5881854-6/
http://cdacrr.blogspot.com/

I've used it a bunch of times in the last few months and it's working well for me.

Interesting. I was assuming that since it has a corresponding Garmin Connect IQ app, it would store the real-time CdA data so I could do a few laps with setup A, and then a few laps with setup B in a controlled environment and see the avg. CdA of each lap.

I got mine haven't used it much, thinking of renting it
(the software is a bit old school, my computer is old/slow as shit so I've been impatient)

It probably stores CdA, but that isn't going to help you. It uses the barometer for elevation, which is inaccurate and also drifts. That's the biggest issue. It's doesn't do laps or use the real elevation in any way, adding a big unnecessary source of error.

Plus you need to mount it either way out in front or off to the side so it isn't influenced by your configuration changes. That isn't a big deal; you'd just need to make/buy the mount, and is true of any airspeed sensor you want to use for testing.

The airspeed data would be useful if you made your own spreadsheet or software to use it. But I'm pretty sure you can get that with the Powerpod also, which is much cheaper.

But the cheapest and best route (ya, both at once) is the Weatherflow meter and CdACrr app. $70 for the weathermeter, $10 for the mount, and $50 for the phone if you don't have one, and the app is practically free ($2?). There are loops or out-back test options that make use of known elevation points. The app performs the analysis in the field and no faffing with spreadsheets required. I'm happy to help if you want to use that, as I have plenty of recent experience.

I used both this and the Notio device (borrowed from another tester for a bit who was helping me). The AeroPod was almost useless. The Notio was better for testing incremental changes, but you need to focus on lap calculations. Real time data is more difficult. The Notio validated decently well in the tunnel.

AeroPod uses a barometric pressure sensor to measure elevation changes, but does not use any baro sensor measurements in its CdA calculations. So, baro sensor drift is irrelevant for AeroPod.

Many users perform the kind of A/B measurements described.

AeroPod laps are marked by touching the "Lap" button on a Garmin, or double-touching the "temple" on Raptor AR glasses. When a lap is marked, the prior lap average CdA is displayed for a few seconds on the screen; after that, live CdA resumes.

Second-by-second CdA measurements are stored in the ride file. Isaac software is useful in comparing CdA test results, and in further refining CdA measurements.

AeroPod MSRP is $399, $100 more than PowerPod V3, and $1000 less than the last-seen price of Notio.

---

John Hamann

Thanks for your reply John

dbeitel

Right, it uses acceleration to determine slope. The baro sensor is used to correct this periodically (tilt errors). But the fact remains that elevation isn't determined with the precision necessary for good CdA measurements. If you are doing actual laps and look at the elevation profile you'll see that the elevation doesn't overlap precisely, and there is drift as well if there is barometric pressure drift. To compute CdA accurately you need coincident beginning and ending points (net elevation change = 0) and software that makes use of this fact.

I'm basing this off data people posted shortly after the Aeropod came out. I don't know if you've updated the analysis software since then.

I find the software pretty unwieldly.

It's designed so that you use the "Lap" button on your Garmin to mark the beginning and ending of each test segment. Makes sense.

In April, I tried to use it to get some simple "directional" data on three helmets. Wasn't looking or expecting to tease out 1W difference, just wanted to see which of the three worked best for me.

Protocol: Out and back on a 2 mile stretch of road with each helmet. Repeat. So 6 laps. Cutting out the turnarounds, this meant 12 tested segments:

1. Helmet A out
2. Helmet A back
3. Helmet B out
4. Helmet B back
5. Helmet C out
6. Helmet C back

7 to 12: Repeat above.

First problem: When I get home, either the Aeropod or the Isaac software split the ride into 3 separate files. And I can't easily tell which is which. And it seems that the files split some of my test segments. I don't really spend the time trying to piece the files together - I'm just frustrated that I wasted a morning. Oh and it shouldn't be a file size issue - the entire test session was just over 24 miles. But three files.

Second problem: The software's CdA analysis gives you a graphic chart like this:



The laps/test segments are marked by the vertical lines in the chart. But like I said, it's hard to know which line corresponds with which test segment. And sometimes I'm not quite sure if a vertical line on screen is a lap marker or part of the background grid that the software uses to present the data.

OK assuming you figure that out, where are the numbers? Well, in order to see your CdA for any of your test segments, you need to click just within the lap marker on the chart, drag your mouse horizontally just inside the ending lap marker, and that will bring up the portion of the ride that you selected (more or less the test segment, depending on how close your mouse clicks are to the actual lap markers), and then you will be shown the CdA for the manually selected portion of the chart.

What I wish it did:

1. Don't separate my single test session into multiple files. Once you do that, you split/ruin my test session. I need all the laps in one place and in the order they were performed.
2. Give me a simple table - Show the laps vertically, and then give me some simply columns, more importantly CdA for that lap, but maybe some other data would be helpful - AP, NP, wind, I dunno... at this point I would love just the table with the laps numbered and my CdA for that lap. It seems so simple but for whatever reason the software doesn't do this.

(BTW if you look at the chart above, you can see that there is definitely signal and not just noise. The longer stretches between lap markers are my test run; the shorter stretches are the turnarounds. You can definitely see CdA jump up during the turnarounds, this is where I am sitting up and getting on the brakes before straightening out and settling in for the next test run. So I do hold out hope that the AeroPod could be useful with better, more user-friendly software.)

---

Amateur recreational hobbyist cyclist
https://www.strava.com/athletes/337152
https://vimeo.com/user11846099

You should be able to export the data in .csv format; then you can analyze the data better. I don't know why the file was split, but that doesn't seem normal.

I'd like to see what your detailed elevation profile looks like for each lap.

Where do you have the device mounted?

In the past we have used elevation data to correct for tilt-related user errors, such as a loose mount, or bumping the power sensor out of position. If such a tilt error occurred, it would be automatically corrected.

In AeroPod testing we found that this error checking/correction method was introducing a certain amount of "noise" into the CdA measurements, so in April we implemented firmware 7.06, which removes any influence of the baro sensor on tilt correction. This FW does improve the stability and consistency of CdA measurement. The tradeoff, of course, is that with FW 7.06+, tilt-related user errors will no longer be automatically corrected; however, we believe that AeroPod customers are more careful about following instructions.

---

John Hamann

I had their old model, ibike i think it was called. Its an interesting concept that works kind of. You do a calibration ride and then the sensors figure out power based on wind, tilt, etc. And it was more effective than you would think compared to a power meter on like 10 second smoothing. The problems I had were the unit has no idea if you sit up and use the base bar or if you're on the aero bars. It also would slip out every once in a while and just read super high or super low watts. Maybe the new units are more reliable but it wasn't consistent enough to be viable.

In general regarding CdA field testing... it's hard to get good results... and reducing any variable's uncertainty is a good thing.

One of those very important variables is elevation/slope. The way you are currently computing CdA actually introduces error compared to a VE protocol, because you are using your computed profile rather than having start/finish points coincident. If the software was able to correct the data file so that the elevation at start/finish were identical, this would help. I've been advocating building an elevation map of the test course, and using this in the CdA computation. Lots of variables need to be measured during a test but the elevation profile is constant and knowable, and should be fixed. The Aeropod would be good for this mapping since it's a better instrument for determining elevation changes than anything else readily available (like GPS or a barometer).

Another one is airspeed. Attempting to measure this where the biker/rider influences the airstream is not going to be very accurate. I understand the mounting location for the Aeropod was selected for convenience and esthetics, and because it is a PM it needs a rigid mount for the acceleration measurements. But for good airspeed measurements the device needs to be mounted either far in front or to the side. It's particularly important that any changes in position or equipment not affect the airspeed calibration. A vs B testing with a calibration bias can lead to poor conclusions.

I appreciate your comments.

I'm pretty sure many readers in this forum know the following but, in case it is unclear...AeroPod does not use elevation or slope data in its CdA computation. AeroPod processes data directly from the calibrated accelerometer, wind sensor, DFPM, and speed sensor. This is all that is required to obtain CdA measurement. Among other things this means that AeroPod does not require closed loops. It also means that, as currently implemented, knowing the precise elevation profile of a course would not improve the accuracy of AeroPod CdA measurement.

We like closed loops too, but reason we like them is because the course does not change between tests.

We understand the concern about airspeed. So far, in our testing at least, we have not seen airspeed influenced much by bike and rider movement. Of course, the mounting location of AeroPod DOES make a difference; one purpose of the calibration ride is to determine the effect of AeroPod location on wind measurement.

Finally, I thought it might be interesting to show how the latest AeroPod firmware improves CdA measurement.

AeroPod transmits its live CdA measurements to a bike computer. AeroPod also records the real-time CdA signal in the HR channel of the ride file (to convert HR to CdA, multiply by 4 and divide by 1000).

In the original AeroPod (FW release 7.01) there was more noise than we'd like in the CdA signal.



We made a big change in April with FW 7.06 that largely got rid of the noise. CdA signal is much easier to find.



These graphs are from my ride this morning. There are two AeroPod devices on my bike (and two separate DFPMs). Note that where the CdA dips measurably toward the end of the ride is when I went into a TT position for a few minutes.

We agree with rruff that CdA is a difficult measurement to make, and we're always looking for ways to improve. If you have suggestions, or if you'd like to discuss more off-line, please email me at jhamann@velocomp.com.

---

John Hamann

Do you also record air speed in the ride file?

---

What's your CdA?

Yes. The ride file records, with one second resolution:

AeroPod Power
DFPM Power
Cadence
Bike Speed
Wind Speed
Elevation
Slope
CdA
PowerStroke

---

John Hamann

You get slope from the accelerometer. Saying that you "don't use elevation or slope data" does not help your case, since errors in the acceleration measurement will just as surely result in a slope error and a CdA error.

Knowing the elevation profile would certainly improve the accuracy, if you used the profile in the computation. Currently the accelerometer measurements (slope) have no error checking. You just run with whatever it says, which is the problem. If you do a series of laps, every lap will have a different profile. In reality it will be the same each lap... at least VE makes use of that fact! But we can do even better by mapping the entire course. With an accurate profile, even wind can be inferred... it doesn't need to be measured. But if you have airspeed measurement as well you can see how CdA and Crr vary at different locations, plus you have a greater ability to check for anomalies and input variables being off.

The Aeropod probably does a decent job of determining "just riding along" CdA... but useful field testing requires greater precision and accuracy. The hardware is probably good enough, but you need a strict protocol (ideally several) and software that makes use of it. The software to do this well (accurate, near idiot proof, easy to use) would not be trivial.

BTW, I'd love to be proven wrong... ;) If you are doing laps on a windy day and not changing anything, what sort of SD in the CdA measurement would you expect with the Aeropod?

What would be very interesting would be a head-to-head test against a "known good" method of field testing CdA, such as the "Chung Method." It shouldn't be that hard to do a few laps, calculating CdA using both the Aero Pod and the Chung Method, and see how cloesly they match. Anyone?

Seeing how it supports airspeed recording in the Garmin .FIT file I can add this to MyCdA.app fairly easily and you should be able to compare. If someone sends me a .FIT file from Aeropod I can hook this up.

---

What's your CdA?

For VE you need basically windless conditions, and an airspeed measurement is just adds extra noise when that is the case.

I have lots of data using the CdACrr app and the weathermeter, in very challenging wind conditions. SD of .003m^2 is typical, which I think is pretty good, considering. Would be nice to do better though... and I'd really like some software that uses a mapped elevation profile. I think this would be the key ingredient for making field testing viable, even for windy and day to day comparisons.

I was also getting very good day to day consistency using the same configuration at first (avg CdA of .205, .205, .205, .208, .208, ..208,... .222 ?!). I then changed configurations (a better one) and my numbers are consistent again in the ~.212 range. I still haven't nailed down what happened, but the one big difference was that I started testing in the morning rather than afternoon. I suspect road heating and Crr effects are the cause.

Here you go (file attached below), with the laps reported by Strava:



Lap 1 was the calibration run. Tests began with Lap 2. Here are the helmets for each lap/test run:

2 & 4 (out & back): Giro Aerohead (M)
6 & 8 (out & back): POC Cerebel (M)
10&12 (out & back): Specialized (S)

14 & 16 (out & back): Giro Aerohead (M)
18 & 20+22 (out & back): POC Cerebel (M)
24 & 26 (out & back): Specialized (S)

On my last leg on the Cerebel, I hit a light and had to slow, and so used the lap button to cut out a portion of that run. So the data from 20+22 is probably a bit janky and is likely unreliable.

---

Amateur recreational hobbyist cyclist
https://www.strava.com/athletes/337152
https://vimeo.com/user11846099

I'm moderately sure that's not the case: VE will use airspeed if you have it.

 
Hmm.. based on John's explanation above I would have expected to see all these values (CdA, airspeed, etc..) but they seem to be missing from your file. I know they're saving them under different .FIT fields (e.g. CdA -> heart rate) but I don't see them.

Here are all the data fields I see:



Did you get the .FIT file straight from your Garmin or did you export it out of Strava or something similar?

---

What's your CdA?

I downloaded from Garmin Connect. I can see if that .fit file is still on my Garmin - it probably is as I have the AeroPod Connect IQ app on an old Garmin 820 that I no longer use and only kept around for testing purposes.

---

Amateur recreational hobbyist cyclist
https://www.strava.com/athletes/337152
https://vimeo.com/user11846099