I've read through that before and its good and useful work, but germain to the point at hand all I conclude is you trust the calibration procedures from the manufacturer and hope that they are doing it correctly. thats certainly a valid and expedient way of doing it, but in no way does it address the issue of "does you particular PM give accuarate mesurements." For the most popular meters (Quark, PT, SRM etc) I'm certain its sufficient for any training purposes, but its certainly doesn't rise to NIST level of tracebility.

Ok I'll play along. How do you measure the accuracy of a PM without, at some time, assuming one is the Gold Standard, or relying on the company's word that theirs is calibrated properly.

For most users all we need is a PM that gives a reasonably accurate, repeatable number. It really doesn't matter if that number is watts, HP or just a number.

Thanks for the update. You could always put another unit on the other crank arm and have separate L/R channels for anyone that wants that at 2x the price (mostly kidding, but I bet someone would buy it, still cheaper than some other PMs).

JGell wrote:

We have seen this before....

remember the Metrigear (speedplay-based system), which made a HUGE splash and never came to market. Now its the "Garmin Vector" and by all accounts, that is a dead product...

I've never done this, but I believe there is a process to check the calibration of a powertap by hanging a known weight on a pedal and reading the torque, etc. Obviously, care needs to be taken to ensure it's done properly, but it is my understanding that you can check the accuracy of a powertap in this manner. I just assume the auto-zero function works on mine. I haven't found a need to check the accuracy.

The guys who are performing field tests to try and estimate their CdA, etc., will need the accuracy, for more accurate numbers.

It can be done.

This is just bugging the heck out of me. I work in calibration, torque and speed (among other measurements) are things I do every day.

Accuracy is what everyone is after not so much precision since they all display in 1 watt increments.

Accuracy is how close you are to the standard.
Precision is at what degree you can measure with as in 10 watts, 1 watts or .1 watts.

Every PM I've seen measure with the same precision, 1 watt, however they may not have the same accuracy. Please stop using accuracy and precision interchangeably.

jaretj

styrrell and others:

There will be much more comparison data available soon. We have been testing for over 2 years and yes our bikes look as "interesting" with all the devices to record the data. We did end up creating an app so we can make recording multiple devices much easier and with data that overlays each other perfectly. Not to mention we got rid of some of the head units on our bikes. Our website will be going live on Tuesday and will have much more information. We will continue to provide comparison data, we have ridden thousands and thousands of miles with just about every make of power meter out there at the same time as ours. Yes, we have ridden with a Quarq on the other side of our power meter.

We also have a very high end Dyno, so we can test all the different variations of power meters.

The bottom line, we are as accurate and in many cases more accurate than the competition on our measured power. Yes, we double this and assume that your right and left leg are doing the same amount of work. From our thousands of miles compared to all of the competition, we don't see an impact at all on how you would use the data to train. Yes, some riders have a slight offset on one leg but it is usually less than 4 to 5%, but from ride to ride it is very consistent. Meaning that you can base your training and gains from it.

As to the comparison of the Ergomo, we don't use an optical sensor like they did. We are using a custom array of strain gages to measure the torque applied to the crank arm.

Thanks for noticing that we have been quiet! We were very intentional about this. We like you, have heard the many promises over the past few years. We certainly did not want to make a promise before we were ready. Our design is complete, we are simply testing the final version of the hardware and processes. So January it is.

Hope this is helpful!

Thanks,

Pat
Stages Cycling

Hmmm...let's see...I've got a 6 year old PT wheel that still is accurate, precise, and stable (I can check it, and do so)...oh, and a couple of Quarqs (prototypes, no less) that range in age from ~1-3 years old that are also accurate, precise, and stable.

So if the PT and Quarq meet your standards, how close to them will the Stages have to be to also meet your standards? Within 1%? 0.5%? Or will you want it to come within a certain percentage of the true standard from the testing rig referenced in this thread?

If there's one thing I've learned when playing around with a bevy of PMs over the years, it's that it's important to be able to at least check the calibration (torque slope for strain gauge based devices) "as mounted", even for units that are "factory calibrated". Things tend to change when you mount them on a bike...

What do you do if the calibration is off? Can you adjust it somehow? I'm a PT user so I just "zero" it once a week or so.