Quarq Cinqo

[JasonW]

Nice review on http://www.pezcyclingnews.com/?pg=fullstory&id=6991&status=True&catname=Latest%20News

Assuming you need a new crankset, this seems to be a very good alternative pricewise to Powertap.

Does anybody have any plans to import them here?

[Minion]

Quarq do not currently supply outside the US and Canada. They say they are conducting electrical compliance tests and should have them completed later this year. I recall reading a post here in July last year that said the same thing.

 

You could get one indirectly from Wiredbike.com, but they seem a bit pricey.

 

[Christie]

Assuming you need a new crankset' date=' this seems to be a very good alternative pricewise to Powertap.[/quote']

 

Or have Cannondale Hollowgram cranks

 

I don't think it will be an issue to get one, we should be able to buy on-line, via shops or ebay.

 

I think the quark has is a potential SRM killer - it has more strain guages, for thousands of rands less.

 

Limited handlebar displays is a disadvantage, for now. 

[PPWTF]

 

 

I stand to be corrected, but engineering friends with no cycling background practically p*ssed themselves with laughter at the marketing blurb trying to pass off more strain gauges as giving greater accurate than less.

 

3 dimensions = 3 axes... how may other plains of movement do we think take place in a crank!? Location of strain gauge plays a far bigger part, but even then any more than 4 is considered engineering overkill - splitting hairs.

 

As for accuracy of measurement, if the major manufacturers continue to use an averaged cadence, then we will continue to get averaged power, with all the room for error that it includes. The greater the cadence pick up, the truer the reading of actual power.

 

What's most interesting about Quark is surely the open source header unit, and it's retail price if you already have a garmin or A N other compatible unit

 

As for pricing, it's a matter of time before power meters become as common place as HRM, prices will drop massively (maybe 5 years) A Canadian firm looked at a shoe insole with a heal mounted "pod", the idea being it was moveable between bikes, that the hardware was on the person. It was inovative, but ran into accuracy and developemnt cost issues, only after showing at Interbike 2 years in a row. What I'm saying is the market is approaching a tipping point, and power meters will become wide spread, sold at Sportsmans warehouse alongside Polar HRM etc - it's just a matter of time

 

(long dull post ends)

 

PPWTF2009-04-30 12:06:11

[Christie]

Here is why I think more is better: (sorry, it is a bit long winded)

1. Power transfer from cranks to sprockets via the chain happens in a plane (2d), force in the 3d cartesian axis would make a negligable contribution to moving the bike forewards.

2. Power generated by our legs varies over the pedal stroke.

3. Power is calculated by measuring the deflection if the spider, which can be directly converted to a torque number, from which power can be calculated.

 

4. The more accurate the measurement of the deformation of the spider, the better the accuracy of the power calculated.

 

5. I would guess that there are 5 locations of 2 strain guages each, one in the radial and one in the tangential direction.

 

6. 4 Guages probably just as accurate as 5 when measuring a solid disk, but >90% of cranks have got 5 arms, which means you need 10 guages to fully capture deformation.

 

7. Powermeters sample at fairly high frequencies, SRM at 200Hz, Powertap at 67 Hz if I remember correctly. I think an averaged value is displayed and stored, else the a handlebar unit will run out of memory in a short time.

 

8. Re cadence: 1 magnet would give the average rotational speed over 2 pedal strokes. A candence of 90rpm is 1.5 rotations per second,or 3 pedal strokes. If power is stored/displayed at an update interval of 1 second, a single magnet will give accurate average crank speed over that 1 second. Not ideal, but good enough, I guess. 2 or more magnets would be better, but they would have to be spaced at prescise angles from one another, and this would be a b!tch to set up at home, for the Quark configuration. A static pick-up on the frame, with the magnets in the crank would make sense, but that complicates the product.

Christie2009-04-30 13:13:59

[Minion]

 

Your friends may need to think a bit harder about the strain gauge problem.

 

My immediate thought for the benefits of additional strain gauges is for temperature compensation - attach additional gauges to an unloaded piece of metal and you can cancel out most of the effects of thermal expansion and contraction. I'm not, howeverm sure whether any power meters on the market actually do do this.

 

This article: http://www.biketechreview.com/archive/pm_review.htm, describes the use of additional gauges to compensate for the transverse stresses that will be present at each gauge.

 

It's an interesting point about the averaged crank speed measurement from cadence. I don't think it detracts from the fact that improving the accuracy of one of the measured variables (torque with strain guages) should still improve the overall accuracy of the system. It could be that they could get even greater than their claimed accuracy values by improving speed measurement.

 

Edit: I see Christie beat me to most of it.

 

Edman2009-04-30 14:01:23

[Minion]

Here're some documents on how the SRM works:

SRM manual - p92 onwards

SRM patent (in German)

SRM patent - PCT

 

Looking at the pic on p92 of the manual, it would appear that SRM does not use automatic temperature compensation. It also illustrates why SRM meters come with 2,4 or 8 gauges. The Quarq with a five-arm spider comes with 10.

 

[PPWTF]

Wheatstone bridge

http://en.wikipedia.org/wiki/Wheatstone_bridge

 

used in the correct way cancels out both transfer stresses and temperature variations.

 

Some strain gauges are less effected by temperature variation, ultimately it's not about the amount of strain gauges, it's about the quality of the gauges, and how much you pay for them is somewhat directly reflected

 

As for sampling rates and storage issues, this may be true in that they are measuring power - BUT as cadence (angular velocity) is measured only once in a cycle it doesn't matter how many samples of power you take between cadence samples you are still only working with one angular velocity around the entire cycle - which means it is already averaged.. you will not get a true power until your angular velocity samples tend to a very very large number (infinite) you would have to take the limit - in one revolution. If you picture the graph of a power cycle over one revolution - you would be getting close to this but it wouldn't be exact, it could never be exact as there are too many variables with wind, linear speed, cross sectional drag - the list goes on.. cycling on a road is not a laboratory, this may be obvious, but you will not get an accurate power reading unless all these things are taken into account, all you need to do is look at the meter you have, and work with the graphs, they won't be true power readings, but they don't need to be, the are a tool used for training, all they need to be is constant.

 

Mrs PPWTF - BSc. (Hon.) Eng - who happens to be a very occasional recreational cyclist

(not claiming to be the oracle, just giving an engineering explanation to something driven by market hype, lots of smoke and mirrors, and a degree of misunderstanding)

 

PPPS.... all power to open source Quarq (the nucleus of an atom)

 

 

 

[bruce]

The powertap uses a torque tube, i.e. strain gauges measure the deflection of the tube.  The gauges are arranged in a wheatstone bridge formation and hence the powertap is reasonably stable under temperature variation.

 

I agree that the power measurement will only be 'really' accurate when both torque and velocity are sampled at a high enough frequency to get a true picture of the variation.  4 times rpm would be a reasonable starting point for velocity (given that there are 2 peaks per revolution, nyquist and all that good stuff).

 

One of the things that i'd like to see is the phase difference between velocity and torque, and how it get's impacted by inertial damping.  This could give insight as to why an indoor trainer and outdoors feels different, yet the power meter reads the same output.  Given that all power meters sample velocity info once per revolution, this information is not available.  In essence I'd like to test if the torque peak and velocity peak change in relation to each other, does this effect muscle fibre recruitment and hence alter the fatigue felt when riding indoors as opposed to outdoors.

[Windbreaker]

 

The powertap uses a torque tube' date=' i.e. strain gauges measure the deflection of the tube.  The gauges are arranged in a wheatstone bridge formation and hence the powertap is reasonably stable under temperature variation.

 

I agree that the power measurement will only be 'really' accurate when both torque and velocity are sampled at a high enough frequency to get a true picture of the variation.  4 times rpm would be a reasonable starting point for velocity (given that there are 2 peaks per revolution, nyquist and all that good stuff).

 

One of the things that i'd like to see is the phase difference between velocity and torque, and how it get's impacted by inertial damping.  This could give insight as to why an indoor trainer and outdoors feels different, yet the power meter reads the same output.  Given that all power meters sample velocity info once per revolution, this information is not available.  In essence I'd like to test if the torque peak and velocity peak change in relation to each other, does this effect muscle fibre recruitment and hence alter the fatigue felt when riding indoors as opposed to outdoors.

[/quote']I know that this went stale but bruce, have you uncovered any more relating to the indoor vs outdoor effects?

 

[bruce]

Nope, nothing new yet