Crank based, single or dual sided vs hub power

[s14phoenix]

I have a question (or few) with regards to power meters and the output power.

 

I understand single sided power meters have certain limitations with the dynamic way in which we generate power between left and right legs. And I understand that a lot of power meters use a strain gauge to measure power output along with some super magic formulas etc. 

 

I compared a crank based single sided power meter with a power meter hub and found a few "issues".

 

When I pedal at a cadence of 80 and lets say the 2 power meters are showing +/- 225W at the same speed on the same road at the same time - which makes sense

 

 I then increase cadence to 95 keeping the same speed in a lighter gear the crank based power meter drops power to say 188W and the Hub based power meter still shows 225W - which I agree with the hub based power.

 

So my question is why does it seem that the less "strain" on the crank is the only power determining factor on the crank based power meter?

Or am I missing something?

Or do I have a faulty power meter on the crank or hub?

Or is this just how the crank based strain gauge meters work? (badly IMO).

 

I am trying to get away from the hub type meters due to getting a new wheelset which I really don't want to fit a power meter hub

 

 

 

[DieselnDust]

If you have changed gears and therefore the chain linear velocity and your hub based pm still gives the same power output then something is amiss with it.

The strain gauge is just measuring the strain in the material I.e. the force deflecting the material. That input force will change as your cadence has changed.

A hub based meter will so less variability in the numbers since the hub is further down the drivetrain and the twist in the shaft is somewhat smoothed by your mass (whereas with a crank based meter your mass is continuously trying to twist the cranks material by varying degrees depending on where your legs are in their range of motion), the number should still drift up and down and should track the crank pm numbers.

[Dexter-morgan]

I started with Hub based, older powertap hub is k@k heavy, then got Garmin Vectors............MMM so much information.

[Capricorn]

If you have changed gears and therefore the chain linear velocity and your hub based pm still gives the same power output then something is amiss with it.

The strain gauge is just measuring the strain in the material I.e. the force deflecting the material. That input force will change as your cadence has changed.

A hub based meter will so less variability in the numbers since the hub is further down the drivetrain and the twist in the shaft is somewhat smoothed by your mass (whereas with a crank based meter your mass is continuously trying to twist the cranks material by varying degrees depending on where your legs are in their range of motion), the number should still drift up and down and should track the crank pm numbers.

 

[Karman de Lange]

Do the real world calcs and see if the power nunbers are within the tolerances.

 

your hub might read 5% high and crank 5% low.  both still in range, but makes large number difference.

 

My quarq and kickr is same at 150w , 8% off at 250W.  at 1000w its ~10% off, all within norm.  This also changes based on cadence. but that is averages.

 

but to answer your question

 

its strain gauge as per above. if you have slight 1-2% difference between left/right leg, this will also exaggerate the error with cadence changes on single sided as it multiplies the power on one side with 2 where hub see continues power.

 

I have left/right difference between high/low cadence .. up to 2% difference, so single sided will be up to 4% wrong depending on my cadence.

[s14phoenix]

SO... If I have it right - if you measure power at the tyre, output at various cadence at the same speed and road conditions will remain say 300w since that is the force the tyre puts out onto the road to remain at say 35kmh even if that translates to 250w at 98rpm on the crank or 320w at 60rpm.

 

so if this is correct then typically one cannot compare wattage numbers between these types of power meters at all as there is too many variables that could influence it. Since this is the case you can also not use these power readings directly against each in scenarios like Zwift or others as 250W does not equal 300w and does not equal 320W.

 

I believe this is also why there is so many different over and under readings in Zwift and Rouvy etc. You can also not compare your FTP and power on one trainer with another and also not on different power meters.

[Barend de Arend]

Power meters need to measure speed for the math to work, but they care about "speed of the strain gauge."

 

In crank based power meters that means cadence.  No cadence, no power.

 

 

All power meters have an error bar, but most claim < +- 2%.  The difference shouldn't be more.  Yours is more.  One of them is wrong.

 

One of the problems of two power meters (pedals or crank-arms -- not hub or crank spindle spider) is that the error can compound in funny ways.  Power meter engineers aren't always statisticians.

 

 

Strain gauges are temperature sensitive, and most indoor trainers have a viscous liquid that's also temperature sensitive, so a lot of indoor trainers are inaccurate for the first 15 minutes or so of a ride until everything warms up.

[RocknRolla]

Remember the basis of power training is having a number to work to, so long as you train with the same equipment, that number will be a "constant" and not 5% easier or more difficult as you are switching between equipment. (Smart trainer VS bike outdoors)

 

I try and not worry about the actual value of the number, but more making sure you hit the required number when training.

 

Training outdoors with a powermeter also generates a variance between training indoors due to environmental factors.

[Karman de Lange]

SO... If I have it right - if you measure power at the tyre, output at various cadence at the same speed and road conditions will remain say 300w since that is the force the tyre puts out onto the road to remain at say 35kmh even if that translates to 250w at 98rpm on the crank or 320w at 60rpm.

 

so if this is correct then typically one cannot compare wattage numbers between these types of power meters at all as there is too many variables that could influence it. Since this is the case you can also not use these power readings directly against each in scenarios like Zwift or others as 250W does not equal 300w and does not equal 320W.

 

I believe this is also why there is so many different over and under readings in Zwift and Rouvy etc. You can also not compare your FTP and power on one trainer with another and also not on different power meters.

 

 

ja

 

only true meassure is at ground level (tyre might take some watts out) , or smart trainer without wheel calibrated correctley (but they do drift with temprature and age).

 

I have 2 power sets .. one for my indoor training and one for outdoor. really irritating but not much one can do.

[s14phoenix]

ja

 

only true meassure is at ground level (tyre might take some watts out) , or smart trainer without wheel calibrated correctley (but they do drift with temprature and age).

 

I have 2 power sets .. one for my indoor training and one for outdoor. really irritating but not much one can do.

 

This sounds EXACTLY what one would need to do. Have a FTP to determine zones for each power meter and situation.

[DieselnDust]

SO... If I have it right - if you measure power at the tyre, output at various cadence at the same speed and road conditions will remain say 300w since that is the force the tyre puts out onto the road to remain at say 35kmh even if that translates to 250w at 98rpm on the crank or 320w at 60rpm.

 

 

Hub power meters measure torque in a torque tube  inside the hub shell so chain tension is providing the twisting force. The nature of the tension lends itself to providing an "average Force" to the torque tube. hence the numbers appear more consistent but this is after drivetrain losses. A worn chain will deliver a lower force to the torque tube due to slippage of the chain.

 

A crank based power meter is measuring the twist in the crank arm as it loads up and unloads on every revoluion. The numberss appear more erratic as a result. But  this is also before drivertrain losses so the number should read slightly higher than a hub unit will. 

The crank or pedal based systems will always give you a more accurate idea of the power your legs are producing whereas the hub is telling you what power is in the drivetrain. It sorts of similar to automotive engine dynamometer testing vs chassis dynamometer testing where power at the fllywheel and power at the wheels are not the same.

 

so if this is correct then typically one cannot compare wattage numbers between these types of power meters at all as there is too many variables that could influence it. Since this is the case you can also not use these power readings directly against each in scenarios like Zwift or others as 250W does not equal 300w and does not equal 320W.

 

its not a good idea to compare the numbers between different meters measuring at different locations but the different technologies still deliver very similar results. My quarq reads slightly higher than my NEO but its very consistent in being that 4-6W higher. So with historical data I can compare the two. There are others factors that introduce variability 

 

I believe this is also why there is so many different over and under readings in Zwift and Rouvy etc. You can also not compare your FTP and power on one trainer with another and also not on different power meters.

 

well actually you can as long as there PM is calibrated correctly and compensates for temperature and the batteries are in similar condition. Battery voltage is also very important to get consistent readings. 

Apps like Zwift and Rouvy work out their own power. I find my ower in Zwift is different to when using TTS4 or Tacx trainer app

Just use one tool and K.I.S.S.

[madbradd]

If the meters are calibrated, then the numbers should be similar (note, not SAME) regardless of where they are on the bike. Yes there are losses and inaccuracies, but if you look at DCRs comparisons he does most rides with multiple PMs - hub, crank, pedals. His numbers are always in the same ballpark and generally track each other consistently.

 

Power is Force x Speed (Rotational in this case). For a crank, that is the strain gauge in the crank x cadence. For the hub it is the strain gauge in the hub x speed of hub.

 

The numbers should be directly comparable (again, not necessarily the same). Note also that the inaccuracies stated by manufacturers will be within a certain power range.

 

Remember as well that they assume that you will be operating in a common cycling range (if you're riding at 200W, a crank based power meter will be more accurate when riding at a cadence of 80-100 than it will be if you're riding at say 40, or 140. Same for hub. If you're riding at 200W at 5km/h or 40km/h or 100km/h, the accuracy will change.

[JA-Q001]

SO... If I have it right - if you measure power at the tyre, output at various cadence at the same speed and road conditions will remain say 300w since that is the force the tyre puts out onto the road to remain at say 35kmh even if that translates to 250w at 98rpm on the crank or 320w at 60rpm.

 

so if this is correct then typically one cannot compare wattage numbers between these types of power meters at all as there is too many variables that could influence it. Since this is the case you can also not use these power readings directly against each in scenarios like Zwift or others as 250W does not equal 300w and does not equal 320W.

 

I believe this is also why there is so many different over and under readings in Zwift and Rouvy etc. You can also not compare your FTP and power on one trainer with another and also not on different power meters.

Not really like you said. If you keep a mass moving at a constant speed, you have to apply a constant force to overcome friction ( this may be an incline, rolling resistance, bearing friction, chain friction and air resistance). In cycling we use wattage to measure this force. If you are going at a constant speed, and all opposing forces stays the same, then your power output should remain the same.

 

The powermeter uses a rotation speed and the Force in Newtonmeters to determine your power. If you stay at a constant speed, and you change your cadance your force output into the pedals change. Higher cadance, lower force and the opposite if your cadance drops.

 

Getting to the crank vs hub part. The crank should be equal to the hub minus drivetrain losses. Drivtrain losses do go up with a higher cadance, but the 2 should track parallel.

 

I actually prefer the hub on my mtb. Its a lot more stable on rough terrain. Crank based PMs use accelerometers these days and if you are freewheeling down a hill on a rough track you get a lot of false readings with the accelerometer picking the shaking up as rotations, the hub works with wheel rotations, so this is not a problem.

[DieselnDust]

Not really like you said. If you keep a mass moving at a constant speed, you have to apply a constant force to overcome friction ( this may be an incline, rolling resistance, bearing friction, chain friction and air resistance). In cycling we use wattage to measure this force. If you are going at a constant speed, and all opposing forces stays the same, then your power output should remain the same.

 

The powermeter uses a rotation speed and the Force in Newtonmeters to determine your power. If you stay at a constant speed, and you change your cadance your force output into the pedals change. Higher cadance, lower force and the opposite if your cadance drops.

 

Getting to the crank vs hub part. The crank should be equal to the hub minus drivetrain losses. Drivtrain losses do go up with a higher cadance, but the 2 should track parallel.

 

I actually prefer the hub on my mtb. Its a lot more stable on rough terrain. Crank based PMs use accelerometers these days and if you are freewheeling down a hill on a rough track you get a lot of false readings with the accelerometer picking the shaking up as rotations, the hub works with wheel rotations, so this is not a problem.

 

 

I have not found that the crank based systems give false readings. My finding are the exact opposite

[s14phoenix]

...

 

I actually prefer the hub on my mtb. Its a lot more stable on rough terrain. Crank based PMs use accelerometers these days and if you are freewheeling down a hill on a rough track you get a lot of false readings with the accelerometer picking the shaking up as rotations, the hub works with wheel rotations, so this is not a problem.

 

 

I have not found that the crank based systems give false readings. My finding are the exact opposite

 

I think you are both also correct in that Hub power meters are a lot smoother and there is less error when off-road shaking etc but that a crank based power meter is a lot more direct with its reading thus being as inconsistent as you are or as smooth and consistent as you are.

[Fat Boab]

I have a question (or few) with regards to power meters and the output power.

 

I understand single sided power meters have certain limitations with the dynamic way in which we generate power between left and right legs. And I understand that a lot of power meters use a strain gauge to measure power output along with some super magic formulas etc. 

 

I compared a crank based single sided power meter with a power meter hub and found a few "issues".

 

When I pedal at a cadence of 80 and lets say the 2 power meters are showing +/- 225W at the same speed on the same road at the same time - which makes sense

 

 I then increase cadence to 95 keeping the same speed in a lighter gear the crank based power meter drops power to say 188W and the Hub based power meter still shows 225W - which I agree with the hub based power.

 

 

What happens if you decrease cadence by selecting a bigger gear for the same road speed etc? 

 

Secondly, is the behaviour you've noted consistent at other constant power levels?