I went on a 2 hour Computrainer ride over the weekend and noticed the average watts from the Powertap was 17 less than the Computrainer result. My Powertap Pro+ was purchased new just a few weeks ago and I’m using a Garmin Edge 500 head unit set to auto-calibrate on coast (which I did several times). Also, I’m not sure if this is related or not, but the average speed reported via the PT was about 1MPH higher than the CT.
Is there an explanation for the discrepancy?
the computrainer measures power after watts are lost at the tire. due to this there is usually a rolldown calibration test you can do, to make for more accurate readings for whatever tire/pressure/clamping force you are using.
did you do that process?
in the end, the computrainer is guessing a bit no matter what you do, and may just read higher. it might be a good idea to record the difference at various wattages so you have an idea if it is a constant difference or a percentage.
I did a winter of computrainer classes last year with my powertap. My powertap was always low no matter how much roll down calibration I did with the computrainer.
Yes, I did the rolldown calibration → My Garmin HU is set to auto-calibrate, and as mentioned above, I coasted several times throughout the workout. It’s also set to display average power @ zero cadence, but that shouldn’t matter much since I didn’t stop pedaling for more than 10 seconds to calibrate the PT. I understand there will always be discrepancies between the two, but a difference of 10+ average watts is significant and if I’m going to be using the PT for training/racing, then I would like have an accurate reading, and at the least, understand whether the discrepancy is static or dynamic.
are both set to ignore 0s in their averaging?
anyway I would sooner be suspicious of the computrainer than the powertap if the powertap is new =)
Is the discrepancy constant over time and over different wattages. If so, then does it really make a difference to your training. If all you need to do is apply some arithmetically correction.
Basically if both power devices are precise over the range of wattages, temperatures and over time that you use it, it shouldn’t really matter to you if they are truly accurate.
First of all, 10w is only significant if your power is <200W. Second, there are errors associated with measurement with both devices. It could be that the actual power you’re putting out is neither that reported by the PT or that reported by the CT (after taking into account the difference in wattage expected due to the difference between the measurement points of each device). As long as you understand how each device reports your power, then the number reported doesn’t matter in the absolute sense.
Say, for instance, your PT reports 10W high at all power increments 0-400W. You go out on a bike ride and average 250W for an hour, and your peak and min output stays between 0 and 400W the whole time. It doesn’t matter that you’re actually only putting out 240W. You know it’s 10W high, so it is what it is.
That said, you have to decide which power reading to accept, and which to “take with a grain of salt.” I would personally believe the PT over the CT since it’s newer, and it’s what you’ll be using the most. Do a step test, riding your PT wheel on the CT to determine the relationship between the power displayed by the two at a range of wattages you’ll expect to output in a ride. You could do something like 2 min at each of ~10 different wattages, say 150- 250 in steps of 10W. Use one device or the other to measure your incremental change in wattage, and note whether or not the other increases by the same amount. Run through this procedure 3 times, the first time through go 150-250, then recalibrate, then bring your wattage well above 250 then go 250-150, and a third time you could do a random order. That procedure ought to rule out hysteresis…
Both of our computrainers are well within both of our powertaps margins of accuracy. This holds true in the 100W to 400W range (and average workouts value), and I’ve also found this to be true irrespective ot the powertap calibration number (between 2 and 3 anyway). I have never witnessed more than 3% difference between the two (10 watts at 300 watts). 17 watts is likely too much unless you ride at 400+ watts… Do a static test on your power tap, and, if OK, send the computrainer for recalibration. Or simply ride your computrainer with your power tap.
Power tap will always read velocity slightly higher. Since all resistance comes from tire friction, the press-on force results in a smaller apparent diameter. Basically, like if you were trying to measure your wheel circumference with severely under inflated tires.
Francois in Montreal
I see the same thing. I have 2 powertaps: SL+ and PRO+ and a computrainer. After many rides of riding either PTAPs in parallel with the computrainer, the following trend is constant for me…
My PTAP PRO+ (training wheel) reads a good 4.5% lower than the computrainer.
My PTAP SL+ (racing wheel) reads about 1.7 to 2.0% lower than the computrainer.
The percentage are calculated on a CP20 part of a longer session, with the computrainer calibrated several times during a long warmup. While these differences do not seem like much, the reading difference between PRO+ and Computrainer at 300w is almost 15 watts… not negligible!!!
The speed difference you see is normal… the roller of the trainer is small, and by using a fair amount of pressure to avoid tire slippeage, deforms the tire more which results in a smaller effective wheel radius compared to the radius would would get with the wheel on the ground. For a given rear-wheel-RPM, the computrainer will see a smaller speed than your normal bike cpu.
Seems like one (or both) of your hubs is off.
Have you done a stomp test to see if they are consistant with each other?
I have the same problem. For me it’s been constant, so it’s not a big deal. When I test, I record both values. My CT result is important in setting a proper FTP for ErgVideo use, but beyond that, I don’t think it’s that important. My bike based PM are what I compare when looking at my WKO+ files. I would not get worked up over 10 watts, that is consistent from workout to workout. I had a time when it bothered me and I tried to make them match. I have the best results keeping the roll down number between 2 and 2.50. Above 2.50 the disparity was greatest.
Not a CT user, but I recall reading something about “drift” in the CT resistance unit over time it heats up with use.
My CT reads a bit over 10% less over the whole wattage regime. CT speed is also lower than PT wheel speed. My CT is over 10 years old, so perhaps some internal electrics are a bit degraded. Nonetheless, I never assume a CT should match a PT or any other powermeter. Too much difference in methodology. Just like I wouldn’t expect a PT to match an SRM, or even SRM to match a Quarq.
Don’t worry about it and just note the differences. Adjust your training for the discrepancy depending on which tool you are using that day.
Basically if both power devices are precise over the range of wattages, temperatures and over time that you use it, it shouldn’t really matter to you if they are truly accurate.
Unless you are really serious about power based training and you would like your data from this year to be comparable to data from years ago and into the future. Comparable across multiple devices is also not too much to ask. A watt is a watt and power files should not be marked with an asterik to describe the device being used to measure them. Anything less is settling.
To the op, you could do a stomp test on the pt and check its error and then compare that to a calibrated ct. There’s your answer. If either is off by more than 2-3% I would send it for repair.
Do you know how old that computrainer is? I have 3 of them. 1 2011 LAB, 1 2011 Pro and a 1994 Pro. The 2 recent ones report power about 2-3% lower than my Quarq (which is spot on since the Quarq measures power at the crank). The 1994 Pro reported power about 15% higher.
There is a calibration pot behind the sticker on the load generator near the serial # sticker. If you really wanted to get them to match you could play around with it, or you could send it back to Racermate to be recalibrated.
Quick update. I contacted Saris about the discrepancy and they told me that the accuracy of the PT is within 1.5% vs. 5-10% on the CT (depending on the model). The rep also told me that my PT is not in need of service since the calbration factor of 519 is within their spec. Having trained on a CT before buying the PT, I was shocked to see my FTP drop by nearly 20 watts and I was really questioning a lot of different things (saddle, bike fit, lack of training, etc.).
Which leads me to my next question. My sprint race bike splits this year have been between 33-35 minutes on a road bike w/no aero-bars and I probably spend about 60% of my time riding with my hands on the hoods. Let’s assume the grade ranges between +/-4% with a steady 10mph headwind. Does a 2.85W/Kg seem consistent with the results?
But if both devices are precise, presumably the OP isn’t getting a new PM device every year. So to the effect that data on a particular device is precise and can be used to compare to prior efforts when measured on the same device.
Also if both devices are precise even if not truly accurate, then some formula could be used to convert power from one to the other. So comparisons can be made.
But if both devices are precise, presumably the OP isn’t getting a new PM device every year. So to the effect that data on a particular device is precise and can be used to compare to prior efforts when measured on the same device.
Also if both devices are precise even if not truly accurate, then some formula could be used to convert power from one to the other. So comparisons can be made.
It takes less energy to ensure your power measuring devices are both accurate and precise than it would to follow through on your vague plan.
If you have a fan blowing on you and the Powertap it may change the temp which would affect the reading. Someone mentioned on another forum to reset the zero offset after a few minutes of riding.
Having trained on a CT before buying the PT, I was shocked to see my FTP drop by nearly 20 watts and I was really questioning a lot of different things (saddle, bike fit, lack of training, etc.).
Why are you assuming your wattage dropped. You never answered the question about how old that CT is and if anyone has done a sanity check to see if it’s calibrated.
Without checking your actuals with an PM that is known to be properly calibrated you really have no way to know which one is correct, or perhaps neither of them are accurate.
In any case, if your PT is consistent from session to session then you can still use it to properly do your workouts and assess improvements.