Login required to started new threads

Login required to post replies

Prev Next
Re: Official power2max support thread [Frost] [ In reply to ]
Quote | Reply
Hi frost,

I would not advise changing the slope on a power meter, because the tools you have at your disposal to determine the trueness are too blunt to do so. Regarding two P2Ms: I use 4 different ones and was never able to detect a difference, nor have we had reports. Within spec the potential differences are so small that it would be extremely difficult to notice.

Best
Nicolas

---
power2max
http://www.power2max.com/northamerica
official power meter of Movistar Team
Quote Reply
Re: Official power2max support thread [aaronechang] [ In reply to ]
Quote | Reply
@aaronchang:

I follow these discussions from afar but don't get involved, because there is a lot of "dangerous half knowledge" involved and it is difficult for us to get involved without going into areas of our IP. power2max's design goals were to deliver a reliable and consistent product that works as specified. Electronics can fail, which is why the unit has internal checks, which you see as flashes of the LED when you insert the battery. A red LED that comes on after green flashes is like the check engine light.

Regarding calibration and the points you raise: according the NSK (http://www.nskamericas.com/...and_Drive_Torque.pdf) starting friction is 2 to 2.5 larger than dynamic friction, which would amount to about 2% in your with an ideal BB. This document doesn't talk about the pre loading effect of canting the bearing.

In the example you cite with a test of starting torque you assume that starting torque remains constant and doesn't increase with preload and thus canting, which is an untested assumption.

During our R&D work our engineers tested different configurations and found that some configurations had enough friction to make calibration biased and unreliable. Please also remember that the friction can be affected by changing the installation (e.g., by increasing pre-load), which make it difficult to get results that can be reproduced with confidence. In short - if you get different results, you can't tell if it because something in the installation changed or because of another reason.

In our experience the test is flawed and not prone to giving results that can be reproduced with confidence. It is also easy to get something wrong in execution.

Best
Nicolas

---
power2max
http://www.power2max.com/northamerica
official power meter of Movistar Team
Quote Reply
Re: Official power2max support thread [power2max] [ In reply to ]
Quote | Reply
power2max wrote:
@aaronchang:

I follow these discussions from afar but don't get involved, because there is a lot of "dangerous half knowledge" involved and it is difficult for us to get involved without going into areas of our IP. power2max's design goals were to deliver a reliable and consistent product that works as specified. Electronics can fail, which is why the unit has internal checks, which you see as flashes of the LED when you insert the battery. A red LED that comes on after green flashes is like the check engine light.

Regarding calibration and the points you raise: according the NSK (http://www.nskamericas.com/...and_Drive_Torque.pdf) starting friction is 2 to 2.5 larger than dynamic friction, which would amount to about 2% in your with an ideal BB. This document doesn't talk about the pre loading effect of canting the bearing.

In the example you cite with a test of starting torque you assume that starting torque remains constant and doesn't increase with preload and thus canting, which is an untested assumption.

During our R&D work our engineers tested different configurations and found that some configurations had enough friction to make calibration biased and unreliable. Please also remember that the friction can be affected by changing the installation (e.g., by increasing pre-load), which make it difficult to get results that can be reproduced with confidence. In short - if you get different results, you can't tell if it because something in the installation changed or because of another reason.

In our experience the test is flawed and not prone to giving results that can be reproduced with confidence. It is also easy to get something wrong in execution.

Best
Nicolas

I've calibrated a lot of different crank-based PM's, somewhere between 50 to 75 different units and performed the calibration well into the 100's of times. The precision and accuracy is far greater than what you suggest. I've never seen BB condition or pre-load affect the results, and that includes taking a crank out of a bike where the crank is barely spinning and putting it in a new, perfectly-adjusted BB. It's simply not an issue that actually impacts the results in any appreciable way in my experience. I've also done enough dynamic calibration checks with a Powertap and crank-based PM to completely trust the method of performing a static calibration.

I'm not going to spend a bunch of time arguing about it on the internet, though. I'll just continue to use and recommend to my clients a power meter that can be user-calibrated. While it's frustrating to me that you guys won't unlock this capability, there are other options out there that do.
Quote Reply
Re: Official power2max support thread [power2max] [ In reply to ]
Quote | Reply
power2max wrote:
@aaronchang:


I follow these discussions from afar but don't get involved, because there is a lot of "dangerous half knowledge" involved and it is difficult for us to get involved without going into areas of our IP. power2max's design goals were to deliver a reliable and consistent product that works as specified. Electronics can fail, which is why the unit has internal checks, which you see as flashes of the LED when you insert the battery. A red LED that comes on after green flashes is like the check engine light.

Regarding calibration and the points you raise: according the NSK (http://www.nskamericas.com/...and_Drive_Torque.pdf) starting friction is 2 to 2.5 larger than dynamic friction, which would amount to about 2% in your with an ideal BB. This document doesn't talk about the pre loading effect of canting the bearing.

In the example you cite with a test of starting torque you assume that starting torque remains constant and doesn't increase with preload and thus canting, which is an untested assumption.

During our R&D work our engineers tested different configurations and found that some configurations had enough friction to make calibration biased and unreliable. Please also remember that the friction can be affected by changing the installation (e.g., by increasing pre-load), which make it difficult to get results that can be reproduced with confidence. In short - if you get different results, you can't tell if it because something in the installation changed or because of another reason.

In our experience the test is flawed and not prone to giving results that can be reproduced with confidence. It is also easy to get something wrong in execution.

Best
Nicolas

Unverified claims. Prove it. What could possibly be so secret about a means to validate the accuracy of power meter?

Let's assume for a moment the BB friction or installation (e.g. pre load) affects the power meter's slope calibration as you suggest.

That sounds to me like a perfect reason to permit a user to check and verify slope calibration with the power meter installed on their bike. How can the P2M factory possibly know how much such factors might affect slope once installed?

Static torque tests have been demonstrated to be a reliable and accurate means to test spider based power meter accuracy, and published in the scientific literature. Link to full research paper here:
http://www.researchgate.net/publication/227000666_A_static_method_for_obtaining_a_calibration_factor_for_SRM_bicycle_power_cranks


I'd suggest you'd need to provide some solid data to suggest such research is flawed.


Not permitting an adjustment to a power meter's slope, is IMO a significant omission on the part of P2M.
Quote Reply
Re: Official power2max support thread [power2max] [ In reply to ]
Quote | Reply
On a positive note...

P2M Type-S installed in an hour, (including time to drive up to the hangar to use the vice!). Seems really nice. I've done four two hour turbo sessions and just wanted to say it seems a really well made component.

I think some of my friends on their third and fourth Stages will be very interested.

Praxis rings are a big step up from FSA/Rotor Q. Can't believe what I've been missing...

Happy Christmas to all at P2M.
Quote Reply
Re: Official power2max support thread [Watt Matters] [ In reply to ]
Quote | Reply
@roady, watt matters: the topic of calibration is always discussed with much passion. power2max has chosen not to make it adjustable not to be mean or any other reason, but because R&D led to the conclusion that the outcome would be inferior to a precise and reliable factory calibration.

Regarding the linked article by watt matters - the procedure described therein does not involve a bottom bracket, so it is not the same method.

Regarding the observation that the spider and hub based power meter agreed when calibrated with the same method: of course they will, that should be the case, because both would be affected by the same loss in torque using the method.

Best
Nicolas

---
power2max
http://www.power2max.com/northamerica
official power meter of Movistar Team
Quote Reply
Re: Official power2max support thread [power2max] [ In reply to ]
Quote | Reply
power2max wrote:
@roady, watt matters: the topic of calibration is always discussed with much passion. power2max has chosen not to make it adjustable not to be mean or any other reason, but because R&D led to the conclusion that the outcome would be inferior to a precise and reliable factory calibration.

Regarding the linked article by watt matters - the procedure described therein does not involve a bottom bracket, so it is not the same method.

Regarding the observation that the spider and hub based power meter agreed when calibrated with the same method: of course they will, that should be the case, because both would be affected by the same loss in torque using the method.

Best
Nicolas
IMO this decision makes P2M an inferior product, and I see no reason to change that opinion until you can supply evidence to suggest otherwise.
Quote Reply
Re: Official power2max support thread [Watt Matters] [ In reply to ]
Quote | Reply
Hello Watt Matters,

you are of course entitled to your opinion, like everybody else. Just to clarify - it is of course possible to take readings by hanging a weight off the crank arm of a power2max. We simply doesn't recommend the test because it is not a reliable, unbiased, reproductible test, based on the our experience. We suggest testing against a Cyclus2 or to do a climb test. In our factory we have even more precise testing capabilities.

Have a great holiday and best,

NIcolas

---
power2max
http://www.power2max.com/northamerica
official power meter of Movistar Team
Quote Reply
Re: Official power2max support thread [power2max] [ In reply to ]
Quote | Reply
power2max wrote:
We simply doesn't recommend the test because it is not a reliable, unbiased, reproductible test, based on the our experience. We suggest testing against a Cyclus2 or to do a climb test. In our factory we have even more precise testing capabilities.

Have a great holiday and best,

NIcolas


Your experience runs counter to that of a lot of other folks. Again, I'm not mad about it, there are other options. Used SRMS and new Quarqs are roughly the same price. I just think it's worth nothing that P2M is definitely giving up some sale by keeping a hard line on the user-calibration thing.

Side note: the Cyclus2 has a stated accuracy of 2%. That's not as accurate as a correctly-done static measurement.
Last edited by: roady: Dec 23, 14 13:28
Quote Reply
Re: Official power2max support thread [Watt Matters] [ In reply to ]
Quote | Reply
No it doesn't. There is more to the process than hanging weights. The timing and algorithm can change the final output. I've been in calibration for 16 years and know what he's talking about.
I've strain gauged hundreds of shafts to do things close to what a power meter does.

sampling torque data and angular velocity at different rates poses a few challenges. Getting an energy balance is not as easy as you think it is
Quote Reply
Re: Official power2max support thread [jaretj] [ In reply to ]
Quote | Reply
jaretj wrote:
No it doesn't. There is more to the process than hanging weights. The timing and algorithm can change the final output. I've been in calibration for 16 years and know what he's talking about.
I've strain gauged hundreds of shafts to do things close to what a power meter does.

sampling torque data and angular velocity at different rates poses a few challenges. Getting an energy balance is not as easy as you think it is

That's great, but in practice static calibration is repeatable within well under half a percent--and there's published research with demonstrates repeatability of .01 Hz/NM. And when dynamically checked (such as against a powertap), I've never seen slope errors from a static calibration. So, on the one hand you have theoretical posturing that it's not accurate or precise; on the other hand you have the actual experience of a lot of people who've calibrated their own power meters, which runs completely contrary to these theoretical problems. And it's not just me--I've never really heard of anyone having issues with a static calibration on an SRM, ever.

If there is something special about a P2M that makes static calibration inaccurate, then I have a hard time trusting it to measure power accurately (although those who claim that the units "don't need calibration" frequently uses static calibration resolve to demonstrate that this is the case? Go figure....)

Sorry, you're swimming upstream when you question static calibration of power meters, because it's a scientifically-validated method that's been proven to work well.
Quote Reply
Re: Official power2max support thread [roady] [ In reply to ]
Quote | Reply
"Sorry, you're swimming upstream when you question static calibration of power meters, because it's a scientifically-validated method that's been proven to work well"

For those of us that do it for a living, we know the limits and circumstance that a static calibration is valid. This is one of the cases of it depends on the angular sample rate, torque sampling rate and the integration with respect to time
Quote Reply
Re: Official power2max support thread [jaretj] [ In reply to ]
Quote | Reply
 I have my A2LA accreditation and my six sigma to do that type of testing
Quote Reply
Re: Official power2max support thread [jaretj] [ In reply to ]
Quote | Reply
jaretj wrote:
I have my A2LA accreditation and my six sigma to do that type of testing

Your argument from authority falls completely flat to those with experience calibrating power meters and in disagreement with the published scientific research on the subject.

If a strain gauge-based power meter can't reliably reproduce force measurements when a weight is applied to them, I certainly don't trust it to accurately display power. Fortunately, they can--which you seem to completely ignore in your posts.
Quote Reply
Re: Official power2max support thread [roady] [ In reply to ]
Quote | Reply
And you have ignored the time component for that particular device which is most likely different from the SRM and is important for calculating power. The torque measurement is only part of the measurement.
You guys are trying to say that the devices are identical and the published research from the SRM should apply but you don't know that for sure.
I have practical experience with this. If you choose not to listen, that's OK with me.
Quote Reply
Re: Official power2max support thread [jaretj] [ In reply to ]
Quote | Reply
Guys, a lot of discussion on calibration of the powermeter and the ability to do this yourself.
This might be interesting for just a small group of people, but for most people a powermeter should just do its job. It should be flawless, easy to install, easy to maintain, reliable and accurate (in a certain window). But for me, just a cyclist, it should be accurate between rides over a longer period and used as a training aid. I do not have another powermeter and therefore, what does it mean if my powermeter is 2% off compaired to another one? For me it is okay if the powermeter shows 300Watt instead of 306Watt as long as it shows these values all the time. With this I can use it as a training aid to improve my overall performance.

Giving the possibility to change/adjust these powermeter yourself might also introduce a lot of problems by people not exactly knowing what they are doing and also giving P2M a bad name, because it are always these people who start complaining.
Just my 2 cents.
Quote Reply
Re: Official power2max support thread [R1Phrankey] [ In reply to ]
Quote | Reply
R1Phrankey wrote:
Guys, a lot of discussion on calibration of the powermeter and the ability to do this yourself.
This might be interesting for just a small group of people, but for most people a powermeter should just do its job. It should be flawless, easy to install, easy to maintain, reliable and accurate (in a certain window). But for me, just a cyclist, it should be accurate between rides over a longer period and used as a training aid. I do not have another powermeter and therefore, what does it mean if my powermeter is 2% off compaired to another one? For me it is okay if the powermeter shows 300Watt instead of 306Watt as long as it shows these values all the time. With this I can use it as a training aid to improve my overall performance.

That's great, until you have 2 power meters at the same time, say one on your road bike and one on your TT bike--one reading 2% low and another reading 2% high. That ends up being a 12W difference at 300W, and frankly that's enough to affect pacing during both training and racing. This is exactly why I switched from a Powertap to SRM. Another issue is tracking historical data over a long period of time. It's nice to be able to know that the data are consistent across different power meters, and I'd rather not have to use the same power meter for 15 years.

R1Phrankey wrote:
Giving the possibility to change/adjust these powermeter yourself might also introduce a lot of problems by people not exactly knowing what they are doing and also giving P2M a bad name, because it are always these people who start complaining.
Just my 2 cents.

Funny, someone at P2M made that same argument. SRM has allowed user-calibration from the beginning and Quarq have their own app to do it. I've never heard of an issue with users with people inputting an incorrect slope and complaining about it or inducing a lot of problems? That's not surprising since, as you said, the average user isn't going to bother with it; however, those that will calibrate their PM will generally take the time to do it correctly. And, for a lot of those people, the inability to do so is a deal-breaker. It is for me.
Quote Reply
Re: Official power2max support thread [power2max] [ In reply to ]
Quote | Reply
power2max wrote:
Hi frost,

I would not advise changing the slope on a power meter, because the tools you have at your disposal to determine the trueness are too blunt to do so. Regarding two P2Ms: I use 4 different ones and was never able to detect a difference, nor have we had reports. Within spec the potential differences are so small that it would be extremely difficult to notice.

Best
Nicolas

I have 2 and they read differently. My first one was ok when I got it based on a static test, but drifted within a couple of months to about 4% out. I contact P2M about it and was told the static test I was using was not valid. I was pretty sure that it was out as FTP tests and tests against my powertap all seemed to confirm this. I returned the unit for the L/R balance and temp updates and asked for a recal then but you didn't seem to do this. Anyway before Kona I decided to buy another one (great Kona deal :-) ) and this one has stayed in cal. I done a cross check using each P2Max with the Powertap on the rear wheel and it shows the 4% discrepancy. Anyway, I've now finally returned the initial unit and am hoping you will re-calibrate it for free given it went out of cal so soon.

In short, the static test seems to be pretty accurate, although I'm sure with a very worn BB this may not be the case.

I'm sure their are many other P2Max out there that are out of cal, but how are people going to know and report back unless they carefully carry out a static test which you say isn't accurate? If a PM drift out by 5% over a year most people will think they've just got worse/better.

Sorry for rant, but I'm just as frustrated with myself for leaving it this long to send the unit back.
Quote Reply
Re: Official power2max support thread [Nobbie] [ In reply to ]
Quote | Reply
Hi Nobbie,

are you located in North America or in Europe? Not always easy to figure out who is writing due to screen names.

Best
Nicolas

---
power2max
http://www.power2max.com/northamerica
official power meter of Movistar Team
Quote Reply
Re: Official power2max support thread [power2max] [ In reply to ]
Quote | Reply
power2max wrote:
Hi Nobbie,

are you located in North America or in Europe? Not always easy to figure out who is writing due to screen names.

Best
Nicolas

Sorry,

based in UK, I've just sent the unit back. To be fair to you, that's what P2Max europe have said to do before and I just haven't got round to it until now.
Quote Reply
Re: Official power2max support thread [Nobbie] [ In reply to ]
Quote | Reply
Hi Nobbie,

ok, sounds like we have a way to resolve your question.

Some points of clarification: our power meters are designed to keep their calibration for a long time. When power2max came to market our power meters offered a new set of features that were previously thought not to be possible:
- Ability to change chain rings without affecting calibration
- Precise cadence without a magnet
- Temperature compensated
- Stable calibration

We regularly test units at the factory and are able to review calibration. The testing procedure has been qualified and provides repeatable results. Home testing the power meter whilst installed in a bottom bracket introduces a bias of unknown magnitude, which makes the comparison unreliable. Change your bottom bracket or change the lateral play on your cranks and you might change the measurement by a percent or two (I am just guessing here at the magnitude). If you take a reading you may think the calibration of your power meter has changed, but it hasn't.

This is at the heart of why we don't recommend it. Testing against a Cyclus2 or doing a steep long climbing test are easier to do well, in our experience.

Best
Nicolas

---
power2max
http://www.power2max.com/northamerica
official power meter of Movistar Team
Quote Reply
Re: Official power2max support thread [power2max] [ In reply to ]
Quote | Reply
Nicolas,

I can fully understand why you don't endorse home testing of calibration. As you say, many uncontrolled variables and varying skill/knowledge levels of those carrying out the tests. It could easily lead to loads of units being returned which are perfectly in calibration, but which the customer is sure are under-reading their massive power output ;-)

In my case, I'm an engineer by training and have a pretty good grasp of the issues and have many comparison points to confirm the repeated static test results - powertap and cross check with other P2Max, perceived effort at threshold, power to hold 40kph (I aero test a lot). While any one of these other measures could be wrong, it is unlikely that all are wrong and that the first P2max is the one that is right. What appears to have happened is that the unit arrived with accurate calibration, but soem sort of drift has occurred in the first few months that has now stabilised. I look forward to hearing what your engineers have to say when they test my unit.
Quote Reply
Re: Official power2max support thread [Nobbie] [ In reply to ]
Quote | Reply
Hi Nobbie,

fully understand your concerns as you said it will be checked and you will have an answer.

Yup, the concern about wrong conclusions is a valid one.

In my humble experience tests like power at threshold, etc and speed for 40kph are almost impossible to do to a precision where you could detect any meaningful differences. For example, I went to a wind tunnel and shaved off almost 30 watts at 40kph (I am a big guy with a pretty high CDA) just by varying my head and back position - no change in bike setup or equipment. Within trials comparisons I believe are easier to validate, cross trial comparisons extremely difficult to get to a meaningful margin of error.

Best
Nicolas

---
power2max
http://www.power2max.com/northamerica
official power meter of Movistar Team
Quote Reply
Re: Official power2max support thread [power2max] [ In reply to ]
Quote | Reply
power2max wrote:
Some points of clarification: our power meters are designed to keep their calibration for a long time. When power2max came to market our power meters offered a new set of features that were previously thought not to be possible:

- Ability to change chain rings without affecting calibration



In my experience, the slope on an SRM doesn't change with a chain ring swap. Even if it did, SRM supports a field calibration, a home user can do it pretty easily.
Quote Reply
Re: Official power2max support thread [R1Phrankey] [ In reply to ]
Quote | Reply
R1Phrankey wrote:
Guys, a lot of discussion on calibration of the powermeter and the ability to do this yourself.
This might be interesting for just a small group of people, but for most people a powermeter should just do its job. It should be flawless, easy to install, easy to maintain, reliable and accurate (in a certain window). But for me, just a cyclist, it should be accurate between rides over a longer period and used as a training aid. I do not have another powermeter and therefore, what does it mean if my powermeter is 2% off compaired to another one? For me it is okay if the powermeter shows 300Watt instead of 306Watt as long as it shows these values all the time. With this I can use it as a training aid to improve my overall performance.

Giving the possibility to change/adjust these powermeter yourself might also introduce a lot of problems by people not exactly knowing what they are doing and also giving P2M a bad name, because it are always these people who start complaining.
Just my 2 cents.

High accuracy may not be important for your particular use case, but it is necessary for lots of scenarios. An inaccurate PM is useless for aero testing, for instance, as well as tracking power across multiple power meters / bikes.

I've never heard of problems with SRMs or Quarqs related to people changing the slope. You can always revert to the original slope that came with your power meter. power2max prints it on a certificate, Quarq prints it on a slip of paper, and SRM puts it on a sticker on the back of the power meter itself. It's not that big of a deal.

Static torque testing has been the de factor standard for testing power meter accuracy since power meters first started being used. SRM, Quarq, et al. are all open about using it verify the accuracy of their units. Quarq has a video tutorial on how to use Qalvin to do a static torque test - https://www.youtube.com/watch?v=4IZL3QXtYv0. Garmin Vector also has a section in their manual describing how to perform a static torque test to verify calibration. I don't remember where I read it, but I believe the forthcoming 4iiii's calibration procedure also involves a static torque test. So I'm a bit surprised we're even having this discussion.

If the slope on a power2max was adjustable, then it would all be a moot point. Because at the end of the day I shouldn't really care what power2max's position on static torque testing is - I just want my device to be properly calibrated and to be able to verify it's working properly. If I want to use static torque testing to verify the accuracy of my power2max unit, that's fine. If I don't care either way and trust power2max, that's fine. If I want to use some other verification method, great.

But if I have to send my unit in to be re-calibrated, then that's when I do care what their views on calibration are.
Quote Reply

Prev Next