Swapping Power Meters...see a big difference

[Lewis Moon]

I just bought a Powertap Pro+. For a long time I've been riding with a PowerCal (pseudo PM) but now, on the same course, I'm putting out ~20W more with the Pro+. I did some calculatin' and I can get the Pro+ numbers to work for the course but the PowerCal numbers seem to be really low; I put the "measured power" into the formula and my predicted time is a LOT slower than measured, even if I add in a tail wind. Heart rates match up within 2 BPS.

YMMV

[bewheels]

If I understand what you are saying...you may be comparing numbers that should not be compared. Regardless of what a formula tells you, if you can produce the same amount of power using your old method of measurement, you should produce roughly the same result in your event...regardless of what the new PM system's numbers are.

If your new system of measurement is providing different numbers, that does not mean you will end up with a different result. It just means you will now be working with different numbers.

To a certain degree it is not the actual numbers, it is about using a system that allows you to measure the output in a consistent fashion and work from there.

If you use a home scale to see how much you weigh, its accuracy to a scientific scale is somewhat irrelevant (provided it is not way off). What matters is that it is consistent every time you step on it.

[Joachim]

I just bought a Powertap Pro+. For a long time I've been riding with a PowerCal (pseudo PM) but now, on the same course, I'm putting out ~20W more with the Pro+. I did some calculatin' and I can get the Pro+ numbers to work for the course but the PowerCal numbers seem to be really low; I put the "measured power" into the formula and my predicted time is a LOT slower than measured, even if I add in a tail wind. Heart rates match up within 2 BPS.

YMMV

I thought you worked in science? :)

[Lewis Moon]

I thought you worked in science? :)

Heh...I do. Just stating that I saw a pretty big difference between the two meters on the same course. When I calculated the watts needed to do the course at the same rate of speed, all things being more or less equal, the Pro+ came out spot on. I do intervals on this course, so I have maybe 25 reps to compare.
The powerCal infers power, it doesn't measure it. There's the rub.

[kgreene10]

I have only ventured into the outer-orbital regions of the black hole that is power calculations, but it seems to me that PMs produce different numbers the same objective power for three reasons:

1. Placement -- all equal, power at the pedals should higher than at the crank, should be higher than at the the hub.

2. Strain gauges -- all watts are calculated from (cadence x torque) and whereas the former is easy to calculate continuously, the latter depends not just on the tech in the strain gauges but how many gauges there are.

3. Algorithm -- different head units calculate watts differently, so even the same PM with different computers can yield different numbers. A related issue is how the PM deals with seeming power surges that come from force on the pedal that doesn't translate into power at the rear wheel, such as hitting a bump in the road (biggest deal for pedals).

All of this applies to actual real-life PMs. I don't know what to say about PowerCal and derived power more generally. I'm pretty amazed that they even get in the ballpark and I honestly have no clue how they do their calcs, but frankly I would be much more surprised if derived power gave you the same numbers as the more direct measurement method. Then again, even a broken clock...

[zachateseveryth]

stuff that just infers power like the ibike, kinetic power computer, powerCal, et. can be all over the place.

you'll see differences between physical measurement devices like an SRM vs a Powertap but they'll generally correlate reasonably well to each other, at least for steady-state efforts. Absolute magnitudes can be all over the place. I think most of us have seen guys use power meters that are horribly out of calibration on Strava.

[kramnnim]

So, now that you own a real PM...do you feel the Powercal was worth the cost?

[christian]

I can't say I'm surprised that a strain gauge-based powermeter is better at measuring power than an instrument which infers it from heart rate?

I would be surprised if the difference between two accurately calibrated strain gauge-based meters varied that much, but in the end it doesn't matter, so long as the precision of one gauge is consistent across efforts.

[ergott]

The powerCal infers power, it doesn't measure it. There's the rub.

So you already have your answer. The Powercal is known to estimate on the low side and also be slow to react to shorter efforts. It can be good for longer intervals and managing effort. You can use the Powercal to gauge improvement as well.

Any serious work with a coach and regimen requires a power meter that doesn't estimate or infer numbers.

[Joachim]

Heh...I do. Just stating that I saw a pretty big difference between the two meters on the same course. When I calculated the watts needed to do the course at the same rate of speed, all things being more or less equal, the Pro+ came out spot on. I do intervals on this course, so I have maybe 25 reps to compare.
The powerCal infers power, it doesn't measure it. There's the rub.

Yes, so you found out that something that's not a real power meter doesn't do a good job of measuring power. Not being facetious, but you were stating the obvious. Alternatively, all real power meters are incorrect and powercal is accurate.

[nebraskacycling]

Any serious work with a coach and regimen requires a power meter that doesn't estimate or infer numbers.

I'm not sure why this is. Is there scientific evidence that this provides better results than other training (naked or HR)? I agree that having a PM might make it easier for the coach, but is it necessary for a training regimen?

Thanks,
TrainingNoob

[Lewis Moon]

So, now that you own a real PM...do you feel the Powercal was worth the cost?

I think DC Rainmaker had it right. If you use it within its limitations it's OK. What you're functionally getting is a realtime calculation not unlike the calculated values on Strava. The reason I moved to the strain guage PM was I wanted more accurate readings during changes, like at the start of an interval or TT. Your heart rate lags behind your power output until you've over shot and blow up. :no: The powerCal helps, but is not as accurate or fast.

[Lewis Moon]

Yes, so you found out that something that's not a real power meter doesn't do a good job of measuring power. Not being facetious, but you were stating the obvious. Alternatively, all real power meters are incorrect and powercal is accurate.

I am Obvious Man!

I was also just startled that the difference would be so large. It also started me to thinking that my physiology might not fit the model and that may account for the difference. I have a "Kawasaki" rather than a "Harley" heart. My RP has never dipped below ~55, even when I was running in high school (10miles in 60.06) or racing bikes in college. At 56 I had a measured Max HR of 196 which is well in excess of wat the formulas say it should be.
I'm just weird that way.

[laupsi]

I'm not sure why this is. Is there scientific evidence that this provides better results than other training (naked or HR)? I agree that having a PM might make it easier for the coach, but is it necessary for a training regimen?

Thanks,
TrainingNoob

short answer, yes! check out recent article on Team Sky, think it was Velo News or Procycling; either way, good story on how repeated success was achieved using the minutia of training by power.

[RedRider]

Like the scale that says you weigh less, the power meter that has a higher watt readout is clearly the superior product! :banana:

[Lewis Moon]

Like the scale that says you weigh less, the power meter that has a higher watt readout is clearly the superior product! :banana:

Heh...It also significantly raises the bar. Now, instead of a 275 watt FTP I want a 295 watt FTP. Two edge sword. I measure my progress by racing, not horse power.

[Lionel]

Yeah, powercal under-estimates quite a bit for me, compared to the powertap.

[christian]

short answer, yes! check out recent article on Team Sky, think it was Velo News or Procycling; either way, good story on how repeated success was achieved using the minutia of training by power.

1) Funny.

2) A powermeter is undoubtedly useful to ensure that the time spent training is most efficiently allocated to target the improvements you want, and to objectively measure those improvements. For getting out of the 5s, just riding lots and doing some LTHR 2x20s is clearly enough, but objectively measuring improvement is difficult without a PM.

[nm87710]

Good Luck and Enjoy!

[Ti Designs]

A powermeter is undoubtedly useful to ensure that the time spent training is most efficiently allocated to target the improvements you want, and to objectively measure those improvements.

Power output is a number. What you do with that number is the real question. The power meter replaced the HR monitor as a training tool because the HR monitor was subject to outside parameters and is a trailing indicator. As I've always told the riders I coach, if your HR is above your AT in the middle of a climb, it's already too late... So my first question is how a PM can base it's output on HR??? Making up for missing data is a marketing thing...


How one trains based on power is the question I've been working on for the past month. My simple answer is that most people with power meters don't - this is given a sample set of 60 riders and two coaches' power based training plans. I must have read the power training book a half dozen times, highlighting and taking notes - I understand it all very well, I just can't pinpoint how gains are achieved, or tested for. Even the basis for the whole program - FTP is a complex variable which isn't well defined. Fatigue profiles are also all over the map...

[Lewis Moon]

Power output is a number. What you do with that number is the real question. The power meter replaced the HR monitor as a training tool because the HR monitor was subject to outside parameters and is a trailing indicator. As I've always told the riders I coach, if your HR is above your AT in the middle of a climb, it's already too late... So my first question is how a PM can base it's output on HR??? Making up for missing data is a marketing thing...


How one trains based on power is the question I've been working on for the past month. My simple answer is that most people with power meters don't - this is given a sample set of 60 riders and two coaches' power based training plans. I must have read the power training book a half dozen times, highlighting and taking notes - I understand it all very well, I just can't pinpoint how gains are achieved, or tested for. Even the basis for the whole program - FTP is a complex variable which isn't well defined. Fatigue profiles are also all over the map...

This is one of my main issues. Call it reckless exuberance, whatever, but I have a real tendancy to overshoot my body's ability to keep up.
Last night I saw it on my first interval. I have a 0.7 mile course that starts flat and ends at around 16%. I started too fast and then compounded it by really grinding out the last 200 or so feet. All the rest of my intervals suffered, even though I reigned it in and stuck my nose on the power meter for the rest of the workout.
Had I stayed within my planned parameters, the overall workout would have been much better, instead of an exercise in attrition.

[LesMiner]

Here is the DC Rainmaker (http://www.dcrainmaker.com/2012/11/cycleops-powercal-in-depth-review.html) review of the PowerCal. Lots of great information and comparisons. One can draw any number of conclusions. An interesting point in the review is the PowerCal measured a higher power than the PowerTap over the same course. Quote from the review.

Now, it should be noted that virtually all power meters display some sort of second to second variation that often concerns folks new to power meters. For example, a typical power meter will look like the below string (consider each number the following second). This is from a PowerTap hub across 13 seconds.:


114,117,117,118,104,130,116,120,108,106,106,119,10 8

Now, let’s take a look at some numbers from a PowerCal (these are aligned to the same 13 seconds as the PowerTap above)


35,68,130,163,225,258,260,230,200,105,75,0,0

What you see here is far greater variation between each second, as well as far more ‘super-high’ and ‘super-low’ numbers. For example, the 0’s and the 35/68.

Now, when looking at averages (and I just picked 13 random seconds), we see the following:


PowerCal: 135w
PowerTap: 114w
Another variant is from Kurt Kenetic. Their pseudo power meter is based on their representation of speed versus power. Kurt assumes an average rider wind resistance profile and has made their fliud resistance unit macth that curve. So for example somewhere between 16 mph and 17 mph a rider produces 200 watts. The shortcoming is the contact between the tire and the resistance unit shaft. Unless perfectly coupled, the Kurt power meter will read higher than a PowerTap.

Another interesting point is that PowerTap does not appear on the Cyclops website. It has it own website PowerTap (http://www.powertap.com/)

The health club I go to starting using spin bikes with a power meter from Freemotion (http://www.freemotionfitness.com/webapp/wcs/stores/servlet/Product_-1_10001_10002_68501_176153). The power sensor is on the left crank only along with cadence. So if your pedal stroke is errratic between legs then your average power will be higher. The smoother you spin the lower the power reading. Put more emphasis on the left leg with the same cadence and the power reading will go up. Not all that useful. The only thing useful is the cadence and maybe the HR reading. Another case where a lot of assumptions are applied to calculations that are estimates.