Power metering without a power meter...

yaya

This is a technical discussion so some may find it boring or irrelevant so apologies and please feel free to ignore:-)

On Di2 groupsets and the Shimano software, can one programme it so it knows the cassette/ sprockets and chainrings sizes and the crank arm lengths?

Basically if you have a computer and sensors that can read your speed (wheel speed and/ or GPS), cadence and know which gear you are in, then you have all the parameters needed for calculating Watts output, perhaps more accurately than a dedicated power sensor/ meter...

So the question is how come Shimano (maybe with Garmin) hasn't done it yet and why is everybody spending R&D and marketing resources on selling us another device?

For most of us power metering isn't really necessary, at least not enough to justify spending the extra money on a power meter, but it is a "nice to have" part of our training?

Or am I completely off base here and am missing some critical component????

BR
y

Alex_Simmons/RST

yaya wrote:

Or am I completely off base here and am missing some critical component????

This.

To measure power you need more than knowledge of velocity and/or rotational velocity, you also need to measure the applied forces or torque. Which is why power meters have force sensing strain gauges.

cougie

You're missing out so many things there. Your calculations aren't allowing for incline or wind or weight or aerodynamics.

trivial_poursuivant

cougie wrote:

You're missing out so many things there. Your calculations aren't allowing for incline or wind or weight or aerodynamics.

This.

You could do the exact same route 2 days in a row using the same gears and doing the same speed but one day you have a 20mph headwind. The difference in power applied to maintain the same speed will be very different and without a meter you have no way of calculating it. The opposite would apply if you went much slower but if you had a meter telling you that although you were slower you still pushed as hard to counter the conditions.

The idea of using power is knowing what you can maintain for a given period of time. So if you tried to do a big climb you knew how to pace yourself so you don’t blow from going too hard too quickly.

I think a lot of people get into the mindset of sitting on a smart turbo or wattbike etc which is a benign environment. You don’t do all your riding solely in them conditions and to improve you need to ride outside in the real wind and on undulating unpredictable roads to get any proper benefit.

Alex_Simmons/RST

cougie wrote:

You're missing out so many things there. Your calculations aren't allowing for incline or wind or weight or aerodynamics.

Yes, one can focus on estimating all of the energy demand factors and estimate power that way, or simply measure the energy supply side by measuring the force applied and velocity of application (most power meters).

The problem with the former approach is, as you point out, is the large number of variables and assumptions required.

To give the OP some idea of the requirements, to estimate power output (e.g. for each second say) you need to know:
ρ - air density (kg.m^-3)
Va - air velocity relative to direction of travel (m.s^-1)
Vg - ground velocity (m.s^-1)
CdA - coefficient of drag x frontal area (m^2)
Fw - wheel rotation factor, expressed as incremental frontal area (m^2)
m - total mass of bike + rider (kg)
l - wheel moment of inertia (kg.m^2)
r - outside radius of tyre (m)
Vgf - final ground velocity (m.s^-1)
Vgi - initial ground velocity (m.s^-1)
Tf - final time (secs)
Ti - initial time (secs)
Crr - coefficient of rolling resistance (unitless)
g - acceleration due to gravity (m.s^-2)
Gr - gradient rise/run (unitless)
Ec - Efficiency of drivetrain (unitless)

That's quite a list to accurately measure or know for every moment of a ride. Some factors are reasonably constant, while others will vary considerably from moment to moment.

Link to explain the mathematics involved:
http://www.aerocoach.com.au/math-model/

Much easier simply to measure the applied force and velocity of that applied force. Hence power meters.

yaya

Thank you all for your responses, much appreciated!

I should have pointed out that my current situation is that I can only use my bike on a turbo, and have been for several months and possibly for the foreseeable future. Hence (I guess) why I don’t consider wind and gradient as factors.
Also the trainer I use is a “dumb” fluid one. All you do is change gears on the bike. Resistance changes a lttle when the rotor spins faster/ slower in the fluid but that’s about it.
BR
y

Alex_Simmons/RST

yaya wrote:

Thank you all for your responses, much appreciated!

I should have pointed out that my current situation is that I can only use my bike on a turbo, and have been for several months and possibly for the foreseeable future. Hence (I guess) why I don’t consider wind and gradient as factors.
Also the trainer I use is a “dumb” fluid one. All you do is change gears on the bike. Resistance changes a lttle when the rotor spins faster/ slower in the fluid but that’s about it.
BR
y

Wheel speed can be used as a reasonable proxy for power on many indoor trainers with some care about warming up the resistance unit, use of same tyre, tyre pressure, press on force. Many trainers have a speed-power curve and various businesses provide ways to convert such data to a power estimate.

Here for example is one list of power-speed relationships for a wide variety of trainers:
http://www.powercurvesensor.com/cycling ... er-curves/

Keep in mind there are various factors that will impact on the accuracy and repeatability of such power estimates.

cougie

The trainerroad application can work well with virtual power if you have a turbo that they've tested. All you need is that it's consistent week on week.

timothyw

Probably worth mentioning that there is a power meter out there that works differently to most - the Velocomp Powerpod.

It needs to be attached to a wheel speed sensor, you pre set it with your height and weight, then it has it's own sensors to measure acceleration, incline, wind speed, air pressure - using all that it is able to get similar results to a 'proper' power meter.

Main drawback is that when used with a turbo trainer it is only as accurate as virtual power readings based purely off wheel speed.

Alex_Simmons/RST

TimothyW wrote:

Probably worth mentioning that there is a power meter out there that works differently to most - the Velocomp Powerpod.

It needs to be attached to a wheel speed sensor, you pre set it with your height and weight, then it has it's own sensors to measure acceleration, incline, wind speed, air pressure - using all that it is able to get similar results to a 'proper' power meter.

Not unless it can also provide all the missing variables? e.g. CdA, Crr, wind

timothyw

Alex_Simmons/RST wrote:

TimothyW wrote:

Probably worth mentioning that there is a power meter out there that works differently to most - the Velocomp Powerpod.

It needs to be attached to a wheel speed sensor, you pre set it with your height and weight, then it has it's own sensors to measure acceleration, incline, wind speed, air pressure - using all that it is able to get similar results to a 'proper' power meter.

Not unless it can also provide all the missing variables? e.g. CdA, Crr, wind

The wind as mentioned is determined from a sensor built into the powerpod, CRR and CdA are approximated based on you telling it what sort of tyres you have and what position you are riding in (so if you're on a flat bar bike, usually on the hoods, usually on the drops or on a tt bike).

This approximation is then refined based on it calibrating itself as you ride - for best results you do an out and back ride. CRR is also automatically adjusted in response to the vibrations it detects (so it can tell if you're riding on cobbles, chip seal, good tarmac etc).

There is of course an issue there that the watts will read a bit low if you sit up on the tops, or a bit high when you go into the drops, but most people spend the vast majority of their time riding in the same position and you can argue that this is in itself useful information - IE if you get low in the drops then even though you're making the same effort, you can see that you're getting an extra X watts of aerodynamic benefit from the low position.

Conversely if you're doing intervals and so taking care to hit your numbers you just need to stay in your normal position.

Better still though is if you combine the powerpod with a conventional power meter, then you can get live CDA data (to be clear, I'm not certain if you can do this with a powerpod yet - you might still need to get their 'Newton' bike computer instead of a powerpod).

redvision

TimothyW wrote:

Alex_Simmons/RST wrote:

TimothyW wrote:

Probably worth mentioning that there is a power meter out there that works differently to most - the Velocomp Powerpod.

It needs to be attached to a wheel speed sensor, you pre set it with your height and weight, then it has it's own sensors to measure acceleration, incline, wind speed, air pressure - using all that it is able to get similar results to a 'proper' power meter.

Not unless it can also provide all the missing variables? e.g. CdA, Crr, wind

The wind as mentioned is determined from a sensor built into the powerpod, CRR and CdA are approximated based on you telling it what sort of tyres you have and what position you are riding in (so if you're on a flat bar bike, usually on the hoods, usually on the drops or on a tt bike).

This approximation is then refined based on it calibrating itself as you ride - for best results you do an out and back ride. CRR is also automatically adjusted in response to the vibrations it detects (so it can tell if you're riding on cobbles, chip seal, good tarmac etc).

There is of course an issue there that the watts will read a bit low if you sit up on the tops, or a bit high when you go into the drops, but most people spend the vast majority of their time riding in the same position and you can argue that this is in itself useful information - IE if you get low in the drops then even though you're making the same effort, you can see that you're getting an extra X watts of aerodynamic benefit from the low position.

Conversely if you're doing intervals and so taking care to hit your numbers you just need to stay in your normal position.

Better still though is if you combine the powerpod with a conventional power meter, then you can get live CDA data (to be clear, I'm not certain if you can do this with a powerpod yet - you might still need to get their 'Newton' bike computer instead of a powerpod).

BUT the powerpod cannot be used indoor and the OP specified he is currently only able to ride on the turbo.

Plus, the powerpod is generally regarded as a pile of ****. For the same (similar) money you can buy a 4iiii which is a proper power meter (albeit left side only).

timothyw

redvision wrote:

BUT the powerpod cannot be used indoor and the OP specified he is currently only able to ride on the turbo.

Plus, the powerpod is generally regarded as a pile of ****. For the same (similar) money you can buy a 4iiii which is a proper power meter (albeit left side only).

Generally regarded by who? DCRainmaker was positive https://www.dcrainmaker.com/2016/03/pow ... eview.html and he continues to be fairly positive in his annual round ups.

It's true though, it's not really one you should be using on your turbo, although again assuming that the setup is consistent and depending on your trainer there's no reason that the numbers shouldn't be consistent enough to use for training purposes - there's plenty of people with proper power meters that can't hit the same numbers on the road that they hit on the turbo.

If it was my cash and I was buying now I'd either get a powertap hub or a watteam powerbeat. I don't think I'd ever buy single sided - I honestly think it's amusing that you wouldn't trust a powerpod but would trust to faith that you have consistent power balance (and wouldn't start unconsciously favouring the left leg to hit numbers).

Anyhow, given the OPs question the powerpod is of interest as a power meter that works differently.

yaya

TimothyW wrote:

redvision wrote:

BUT the powerpod cannot be used indoor and the OP specified he is currently only able to ride on the turbo.

Plus, the powerpod is generally regarded as a pile of ****. For the same (similar) money you can buy a 4iiii which is a proper power meter (albeit left side only).

Generally regarded by who? DCRainmaker was positive https://www.dcrainmaker.com/2016/03/pow ... eview.html and he continues to be fairly positive in his annual round ups.

It's true though, it's not really one you should be using on your turbo, although again assuming that the setup is consistent and depending on your trainer there's no reason that the numbers shouldn't be consistent enough to use for training purposes - there's plenty of people with proper power meters that can't hit the same numbers on the road that they hit on the turbo.

If it was my cash and I was buying now I'd either get a powertap hub or a watteam powerbeat. I don't think I'd ever buy single sided - I honestly think it's amusing that you wouldn't trust a powerpod but would trust to faith that you have consistent power balance (and wouldn't start unconsciously favouring the left leg to hit numbers).

Anyhow, given the OPs question the powerpod is of interest as a power meter that works differently.

So the PowerPod does sound interesting and on their website it says that i can be set to "trainer" mode, which I guess means that it'll only use speed (and cadence?) input and calculate power based on the rider's & bike's weight, plus it can display the power on your head unit.

Which brings me back to my original questions...why can't Garmin etc. do a similar thing? And if there was a way to enter crank arm length and gear ratios, wouldn't that make it even better?

The Power curve software looks interesting and it does support my trainer, but it's Windows Only and won't send data to the head unit...

I'm in no rush so am happy to wait and get some more opinions and keep looking for a working solution that is not too complicated or expensive...

thanks again
y

cougie

If you're just turbo training then look into trainerroad.

timothyw

The main reason that such a device as you are describing doesn't exist is because it doesn't work as a means to determine power.

The cadence you are pedalling at, and for that matter the gear you are using make no difference to the power output you are generating - the power is determined by the combination of how hard you are pressing the pedals and how often you are doing it - and the speed that you are travelling at is a poor approximation for the power you are outputting - at the simplest, if you are riding down a hill you can be going very fast indeed, but not outputting any power (you might even be turning the pedals at the same time, but perhaps not fast enough to have any effect!)

For trainer usage, as mentioned, numerous training programs (zwift, trainerroad etc) will guess your power output from the speed that your wheel is turning at. This isn't particularly accurate as they don't know what resistance setting you are on, if you've tightened up the roller too tight, if your tyre is flat etc. Direct drive turbos can give better approximations because there are less of these variables to consider.

Elite offer a Misuro sensor to complement some of their cheaper turbos, it takes the speed that the turbo is going round and converts it into a power reading:
http://www.wiggle.co.uk/elite-misuro-b- ... ce-sensor/

That's what I use with my turbo muin and for my limited needs it's fine. The powerpod can do the same thing but is a lot more expensive, and wasted if you aren't using it outside.

redvision

TimothyW wrote:

Generally regarded by who? DCRainmaker was positive https://www.dcrainmaker.com/2016/03/pow ... eview.html and he continues to be fairly positive in his annual round ups.

It's true though, it's not really one you should be using on your turbo, although again assuming that the setup is consistent and depending on your trainer there's no reason that the numbers shouldn't be consistent enough to use for training purposes - there's plenty of people with proper power meters that can't hit the same numbers on the road that they hit on the turbo.

If it was my cash and I was buying now I'd either get a powertap hub or a watteam powerbeat. I don't think I'd ever buy single sided - I honestly think it's amusing that you wouldn't trust a powerpod but would trust to faith that you have consistent power balance (and wouldn't start unconsciously favouring the left leg to hit numbers).

Anyhow, given the OPs question the powerpod is of interest as a power meter that works differently.

DCR didn't rave about it though and was critical of its price point. All you have to do to get an unbiased opinion is read user reviews of it. Most users (myself included) have found it to be inconsistent at best. A power meter doesn't have to be bang on accurate with the power number, but it has to be consistent. The powerpod just isn't.

OP there is only one way to reliably and consistently measure power output and that is using a device with a strain gauge.

Alex_Simmons/RST

TimothyW wrote:

Alex_Simmons/RST wrote:

TimothyW wrote:

Probably worth mentioning that there is a power meter out there that works differently to most - the Velocomp Powerpod.

It needs to be attached to a wheel speed sensor, you pre set it with your height and weight, then it has it's own sensors to measure acceleration, incline, wind speed, air pressure - using all that it is able to get similar results to a 'proper' power meter.

Not unless it can also provide all the missing variables? e.g. CdA, Crr, wind

The wind as mentioned is determined from a sensor built into the powerpod, CRR and CdA are approximated based on you telling it what sort of tyres you have and what position you are riding in (so if you're on a flat bar bike, usually on the hoods, usually on the drops or on a tt bike).

This approximation is then refined based on it calibrating itself as you ride - for best results you do an out and back ride. CRR is also automatically adjusted in response to the vibrations it detects (so it can tell if you're riding on cobbles, chip seal, good tarmac etc).

There is of course an issue there that the watts will read a bit low if you sit up on the tops, or a bit high when you go into the drops, but most people spend the vast majority of their time riding in the same position and you can argue that this is in itself useful information - IE if you get low in the drops then even though you're making the same effort, you can see that you're getting an extra X watts of aerodynamic benefit from the low position.

Conversely if you're doing intervals and so taking care to hit your numbers you just need to stay in your normal position.

Better still though is if you combine the powerpod with a conventional power meter, then you can get live CDA data (to be clear, I'm not certain if you can do this with a powerpod yet - you might still need to get their 'Newton' bike computer instead of a powerpod).

Sorry to disappoint, but it's not a power meter, it is however one step up from a random number generator.

Given the level of imprecision in such an instrument, it's of no more use than using perceived exertion or HR monitor and a watch.

Mikey23

Was going to mention the powerpod... i can assure you it is not a pile of **** . If calibrated properly works really well and is as accurate as most and at a fraction of the price. Much bike snobbery around which says that a power meter has to be strain gauge based.

But my reason for posting is the point the OP towards a couple of excellent articles about the powerpod, one by rainmaker and the other one you can google. Should help you to understand the principles and the maths regardless of whether you purchase or not...

Mikey23

@alex-simmons... no i’m not going to respond because i would be wasting precious seconds of my remaining life...

redvision

Mikey23 wrote:

If calibrated properly works really well and is as accurate as most and at a fraction of the price.

But it really doesn't - the readings are inconsistent at best. And it's actually only a few quid cheaper than the 4iiii.

It's not snobbery, it's fact, to measure power accurately you need a strain gauge.

yaya

TimothyW wrote:

The main reason that such a device as you are describing doesn't exist is because it doesn't work as a means to determine power.

The cadence you are pedalling at, and for that matter the gear you are using make no difference to the power output you are generating - the power is determined by the combination of how hard you are pressing the pedals and how often you are doing it - and the speed that you are travelling at is a poor approximation for the power you are outputting - at the simplest, if you are riding down a hill you can be going very fast indeed, but not outputting any power (you might even be turning the pedals at the same time, but perhaps not fast enough to have any effect!)

For trainer usage, as mentioned, numerous training programs (zwift, trainerroad etc) will guess your power output from the speed that your wheel is turning at. This isn't particularly accurate as they don't know what resistance setting you are on, if you've tightened up the roller too tight, if your tyre is flat etc. Direct drive turbos can give better approximations because there are less of these variables to consider.

Elite offer a Misuro sensor to complement some of their cheaper turbos, it takes the speed that the turbo is going round and converts it into a power reading:
http://www.wiggle.co.uk/elite-misuro-b- ... ce-sensor/

That's what I use with my turbo muin and for my limited needs it's fine. The powerpod can do the same thing but is a lot more expensive, and wasted if you aren't using it outside.

Thank you for the link! My needs are probably as limited as your and and this seems to provide what I'm looking for, which is to display power on the Garmin head unit. It doesn't have to be accurate but consistency is important.
Only thing is that my Elite fluid trainer is too dumb and is not listed as being supported...

I tried Zwift a few times, which also shows power but I am not yet convinced that Zwift suits me...

Alex_Simmons/RST

Mikey23 wrote:

@alex-simmons... no i’m not going to respond because i would be wasting precious seconds of my remaining life...

My views on this sort of thing are not confined to this unit. there are a number of power meter products that suffer similar issues but for different reasons.

When some decent testing of the unit is published, let me know. There are very few power meters that have actually had decent testing conducted with scientific rigour on their performance across the power duration spectrum.

What level of power meter performance one needs is individual and much depends what applications of power data one intends to use or apply it to. If you are only interested in low-fi applications (e.g. monitoring of intensity while riding), then by and large such things can be done just as effectively without a power meter.

Mikey23

Ok, I’m not so grumpy today... i didnt really post to advertise powerpod or any other product. Simply to point the OP towards a couple of useful articles which outline the underlying principles under which power meters work in order to help to answer his original question...

But having said that, i would refer you to DCRainmakers tests in which he compares the accuracy of multiple power meters in different circumstances. In which a properly calibrated powerpod does pretty well when compared with others. And certainly well enough for the majority of users to whom fractions of watts are not vitally important. And consistent enough for individual use. Not perfect i agree but good enough. And i suspect there is no such thing as a perfect power meter

Not worth falling out over and if someone would like to buy me a set of garmin vector 3s i would not object!

Alex_Simmons/RST

Mikey23 wrote:

Ok, I’m not so grumpy today... i didnt really post to advertise powerpod or any other product. Simply to point the OP towards a couple of useful articles which outline the underlying principles under which power meters work in order to help to answer his original question...

But having said that, i would refer you to DCRainmakers tests in which he compares the accuracy of multiple power meters in different circumstances. In which a properly calibrated powerpod does pretty well when compared with others. And certainly well enough for the majority of users to whom fractions of watts are not vitally important. And consistent enough for individual use. Not perfect i agree but good enough. And i suspect there is no such thing as a perfect power meter

Not worth falling out over and if someone would like to buy me a set of garmin vector 3s i would not object!

DCRs tests are an OK start but they are not the sort of test rigour I'm talking about (that's a lot harder to do) although there is one relatively simple test he could add.

Power meters work in a variety of ways, as I said before, even various strain gauge models have issues the majority are happily ignorant of.

Glad you're feeling better, I've had a pretty crappy week myself.

cruising on empty

An Ultegra Stages Power meter can be had for £310 any day of the week on eBay. They are not luxuries anymore, they are necessaries.

imposter2.0

cruising on empty wrote:

An Ultegra Stages Power meter can be had for £310 any day of the week on eBay. They are not luxuries anymore, they are necessaries.

They may not be 'luxuries' any more - but they ain't 'necessaries' either...