Venting on computrainer power variations
I'm trying to get my head around some Computrainer power variability issues. I did a comparison of the CT power measurement early in a workout and late in a workout and the difference was HUGE--about 18%. Here's my research design: I warmed up on the CT for 25 minutes at an average output between 160 and 170 watts, then calibrated. Then I start a new workout starting my CT and Quarq powermeter at the same time. The CT shows about 10% less average power than my Quarq/Garmin is showing early in the workout. After 80 minutes of cycling, I put my CT in ergo mode and ride at 179 watts then I reset my Garmin for a new lap and it shows power average of 166, almost 8% lower than the CT. So, my CT has gone from undermeasuring my power by 10% early in the workout to overmeasuring by 8% near the end (yes, the Quarq could be the culprit, but it does not seem to be succeptible to the fade that I have documented in the CT based on calibration numbers). This blows me away. The CT Lab is supposed to be highly accurate. This really makes me wonder what kind of information I have been getting from my CT and whether it's reliable.
Comments
My husband uses PT and computrainer and finds the number for each. Very different by like 20watts on the test. He sets his workouts for indoors using the computrainer numbers and when riding outside uses the PT numbers. He is not worrying about it. You have your baseline. Now work hard and the numbers for each will go up.
If the answer is a calibration thing- a WSM will be able to help you. Calling Tom Glynn. He knows everything.
a coupe of thoughts on this to consider. First, not sure how you determined it was your CT that had the problem and not the Quark or maybe both were off.
As to CT calibration, the offical policy is to warm up for at least 10 minutes before calibrating. The machines are more accurate at >150 watts and > 14 mph. I typically test at about 5 minutes to get in the ball park and then again between 10 and 15 minutes. I have a PT also, which I use to feed speed and cadence data into TrainerRoad and if I sense there is a signifcant difference between the two, then I recalibrate again. If the calibration number computed by the CT remains very close (within 5-10 points), I assume the issue is with the PT, otherwise I assume the CT isn't warmed up enough and a go for another couple of minutes and recalibrate. When the numbers are consistent between calibration tests, I assume the CT is warmed up and accurate.
I believe the only real way to determine the accuracy of the CT is to send it back and have it tested.
As Carrie noted, I tend to have "indoor numbers" and "outdoor numbers" and don't stress about the differences. Indoors, it's CT numbers.
As a side note, power meters will typically not have the same numbers from different models. They all measure strain from different areas, so the farther the strain gauge is away from the force (your foot), the lower the power number should be. So a power meter measuring power in the pedal should have higher watts than a strain gauge in the hub.
I think Steve was venting about the massive amount of drift that the CT must have had in it's power reading over the course of an 80 min workout (assuming that the Quarq was constant). I think it's reasonable to think that the Quarq was pretty constant as they are generally only known to drift over time if there are severe changes in temperature or barometric pressure, or maybe with heavy humidity or moisture, etc. Neither of those things were likely to have varied much inside on a trainer. If the CT absolutely has a much different calibration reading 2 mins in vs 10 mins in (hence the reason for warming it up before you calibrate) the question is will this drift level off and stop at some point or will it continue to "warm up" over the course of the whole 80 mins? what about after 3 or 4 hours? If you're doing a 5.5 hr Race rehearsal on a CT, does 180W 30 mins into the workout mean the same thing as 180W at hour 5? We are conditioned to believe that a Watt is a Watt is a Watt. Well it might not be the same effort by your legs if the calibration drifts by a lot over time...
Nothing to add, but just wanted to quote this
Another dumb thought but have you looked to see if your software is dropping data? You can look at the raw data and if you see a ton of zeroes when you know you were pedaling, it might be a Garmin / TrainingPeaks / WKO issue.