Powertap Users--please help with funky download numbers

I just purchased a PT Pro, and I’m getting some funky info when I download the file. The actual bike computer reads fine (or so it seems to me) and gives me what appears to be accurate info regarding average watts and speed (I’ve taken my Garmin with me on the last 2 rides, and the computer and Garmin are fairly close for average speed). But when I download the files into the Power agent, the numbers on the computer and the downloaded numbers are quite a bit different. Now, I have the bike computer set to eliminate zeros from power and speed, and I realize that could be a reason for some of the difference. But everything is slightly different, and the differences are bizarre and inconsistent.

Here is an example from my ride today:
Bike computer/Power agent numbers
229 ave watts/213
22.5mph ave/21.6
137HR/140
Time 2.14.25/2.14.18
Ave cadence 82/76
1719kj/1714
Distance 47.19–same on both computer and Power agent (and both wrong; off by about 2.2 miles when compared to my Garmin, which has always been extremely accurate and consistent on this course). So, the time is off by 7 seconds, the distance is the same, but the overall speeds differ by almost 1mph (and both speeds are wrong if we look at the distance and time–it doesn’t add up; how can the software miscalculate average speed using its own distance and time numbers?).

For what it’s worth, I’ve ridden with an SRM in the past and the bike computer seems much closer to correct in terms of watts and average speed (despite the distance inaccuracy). The Poweragent numbers seem very low. I’d like to get this straightened out so I can get accurate to download into Cyclingpeaks.

I’m hoping like hell I’m doing something wrong with the Poweragent and that the hub isn’t the problem. Thanks for the help!

Did you change the storage rate on your computer?

Do you have the correct wheel size selected?

Also try uploading directly to Cyclingpeaks to see what you get.

Sometimes the PT drops and you can get separate files in the same workout. Did you merge it when uploading into poweragent?

I remember having a similiar problem myself, although I use SRM now so I’m probably not to much help.

I also have the same problem…the unit showed avg power to be 205 but after downloading into Poweragent it was 169. Also after downloading the ride there were multiple “files” that I had to “merge”. I don’t use the interval mode. Some of the files were very short duration with no power numbers.

IS yours the SL 2.4?

I had that for a about 4 months, and I would try to download my workout and would need to merge like 40 files. Lots being 10 seconds or so with no power. This was becuase the 2.4 tends to drop, ALOT.

If its the not the 2.4, then dont know what to tell you…

It’s an older Pro model (yellow hub cover). It appears the unit creates a new “file” when you stop (speed=0) and you resume or the signal drops out.

Thanks for all of the suggestions, guys. I did indeed merge all the files, checked my storage rate, have the wheel size correct, and so on.

I also uploaded directly into Cyclingpeaks and got yet another different set of numbers. How is that even possible for God’s sake?

And just to add to the headache, this ride data is from a cleared unit. Before I rode today, I cleared the thing out completely so there would be no confusion with old ride data. So, I have only one ride on there and three different sets of data. Bike computer data, Poweragent data, and Cyclingpeaks data.

Get rid of the “no zeroes”. You want the zero values in there if you want to have any real representation of what your rides are like. All info is valuable.

Checking the data from my last ride Everything downloaded into Cyclingpeaks matches the CPU except Avg Speed. I hadn’t noticed before because I never look at that display on my CPU. The CPU Avg. Speed is incorrect to the ride time shown on the CPU. So I’m “assuming” it’s using some sort or rolling average algorithm to save CPU memory?? I would go with the raw data loaded to cyclingpeaks.

I’ll give it a shot–thanks.

you should really average in the “zero” value because you are resting or coasting during that time, but you are still riding your bike.

I don’t think your watts are low they seem kind of high considering you are not averaging in the “zero” values. I mean i don’t know how much you weigh and how much elevation you had but at those watts(slightly less at 210Watts) i averaged 22Mph with about 2800ft of elevation gain on a 50 mile ride(that is comparable to your ride). But i was averaging in the Zeroes and WATT goes up must come down. So i may have had a lot of zeroes on the down hills.

The difference in the actual bike computer numbers and the software numbers maybe because some of them do smooth averaging.

But come to think of it i never actually check my average watts on the bike computer after i finish riding, i just wait til i get home and look at the entire ride on the computer.

so who knows Maybe I have the same problem you do, but i never check it right after the ride.

My powertap is an SL

I was going to say the same thing about the Watts looking pretty high on his data. A nice ride nonetheless but it looks like not pulling in the zeros is causing abnormally high avg watts.

I’ve had good luck calling Saris/PowerTap with questions. Their reps are pretty knowlegeable and technical. I would think they could help you with this question.

Thanks, I’ll give them a call and see what they have to say. I dropped them an email, but a call is probably a better idea.

The ride is on hilly and rolling terrain and I’m a big guy. My threshold watts are fairly high (former roadie), and I’ve ridden with power for years, so I know the wattage numbers are actually pretty close on the bike computer (just not the download). The road is in the country and has no stop signs, lights, etc., and I don’t coast or soft pedal on this ride (ever), so there wouldn’t be any zeros in the file anyway. We basically get out there and TT for 2.5 hours.

But, I’ll give it a shot anyway and see if the computer and download numbers line up. Couldn’t hurt to try!

Once again, that’s a nice ride.

Thanks. I’ve got some very strong training partners that keep me honest.

Trevor,

Ave speed on the CPU display seems to be impacted by stopping and, therefore, becomes inaccurate. Mine is always off in training but accurate when racing (since I never stop). My stored data is always accurate.

Thanks, Chris

Thanks, I’ll give them a call and see what they have to say. I dropped them an email, but a call is probably a better idea.

The ride is on hilly and rolling terrain and I’m a big guy. My threshold watts are fairly high (former roadie), and I’ve ridden with power for years, so I know the wattage numbers are actually pretty close on the bike computer (just not the download). The road is in the country and has no stop signs, lights, etc., and I don’t coast or soft pedal on this ride (ever), so there wouldn’t be any zeros in the file anyway. We basically get out there and TT for 2.5 hours.

But, I’ll give it a shot anyway and see if the computer and download numbers line up. Couldn’t hurt to try!

Heh, heh. This just means you’ll be all the more shocked at how much time you really do spend at zero. Even a small % pulls your averages down more than you’d think. You don’t know how many times I’ve been riding solid for an hour, seeing nothing but 200w values seemingly constantly, and then see the average is 178w, etc!

yeah, i am always surprised how low my average watts are with the Zero averages in. i mean even when i stop the cadene to reach down and get a my bottle for a drink i try to not stop pedaling cause i know that drops my watts.