Power measurement on turbo: too inaccurate to be helpful?

topcattim

After spending a few years using a hand me down trainer, I'm looking now to buy a new one. I have tried to follow a heart rate based plan but am aware that cardiac drift makes this a bit unreliable. So I'm looking for a turbo with power measurement. Reading some reviews has taught me that power is often inaccurately measured when calibrated against a Powertap or similar, hardly surprising given that I plan to spend about £300 on a turbo, while power meters cost much more.

But if my intention is just to meter my efforts on the turbo, rather than to work out my actual road powers, then I don't think this matters? If I just work out my zones based on a max power test on the turbo then at least it will all be internally consistent? Or do I need to look out for a turbo with a much better reputation for accurate power measurement?

Herbsman

Judging by the comments in other threads about turbo trainers with power measurement, few if any can be relied on for accuracy or consistency. Save up a bit longer and buy a proper power meter.

charlie_potatoes

Herbsman wrote:

Judging by the comments in other threads about turbo trainers with power measurement, few if any can be relied on for accuracy or consistency. Save up a bit longer and buy a proper power meter.

Also judging by those comments it seems to me that a trainer will be more consistent if your bike is permanently hooked up to it and the tyre pressure remains constant?
Or is that just another one of my excuses to buy a new bike? 8)

amaferanga

If you're patient you could probably pick up a used PowerTap for around £300.

fearby

amaferanga wrote:

If you're patient you could probably pick up a used PowerTap for around £300.

and the rest

amaferanga

fearby wrote:

amaferanga wrote:

If you're patient you could probably pick up a used PowerTap for around £300.

and the rest

Eh? Given that the lower end hubs only cost about £500 new then £300 for a used one is quite possible. The old wired ones might be even cheaper.

bigpikle

The basic 'pro' hub is still >£600 before being built to a wheel in most places, and the newer and better G3 is way more than that! They are still commanding silly prices on any forum and ebay - hard to see any quality rim or something that isnt years old and done '000's of miles for <£500.

amaferanga

Bigpikle wrote:

The basic 'pro' hub is still >£600 before being built to a wheel in most places, and the newer and better G3 is way more than that! They are still commanding silly prices on any forum and ebay - hard to see any quality rim or something that isnt years old and done '000's of miles for <£500.

Only an idiot would pay more than a few hundred quid for a used Elite+ hub or wheel or an older wired system. Plenty of them on eBay I guess.

bigpikle

plenty of idiots out there it seems

I could really do with a 2nd PT wheel for indoor sessions but people keep paying silly money. Will wait for the Rotor crank system I think instead...

danowat

I got an SL+ with Joule 2, new, for £500......

ziglar

where from?

twotyred

With a decent turbo using speed to measure power is maybe not accurate but its pretty precise and repeatable. Using the power calibration curve for my Kurt Kinetic Road machine estimated power is about 9% higher than actual power as measured by the power meter on the bike. However session to session the ratio of speed vs actual power is consistent so I'd say that as long as you are not bothered about your actual power and are just looking to work out whether you are getting fitter then using speed on a good fluid turbo is a lot better than nothing. If you use TrainerRoad then as long as you use the same turbo your training zones will be OK just don't believe the power figure TrainerRoad works out.

topcattim

twotyred wrote:

With a decent turbo using speed to measure power is maybe not accurate but its pretty precise and repeatable. Using the power calibration curve for my Kurt Kinetic Road machine estimated power is about 9% higher than actual power as measured by the power meter on the bike. However session to session the ratio of speed vs actual power is consistent so I'd say that as long as you are not bothered about your actual power and are just looking to work out whether you are getting fitter then using speed on a good fluid turbo is a lot better than nothing. If you use TrainerRoad then as long as you use the same turbo your training zones will be OK just don't believe the power figure TrainerRoad works out.

Thanks twotyred, very helpful. I'm much more likely to go for a turbo like that than to buy a power meter proper, second hand or not.

ut_och_cykla

My impression is that turbo trainers are sufficiently precise & repeatable too, but using a permanent set up , correctly inflated tyre , calibrated roller pressure etc. Certainly good enough for most peoples requirements anyway. AS to whether they accurately measure power - most people who have checked say they don't, but if they were accurate would they report it? :-)

amaferanga

twotyred wrote:

With a decent turbo using speed to measure power is maybe not accurate but its pretty precise and repeatable. Using the power calibration curve for my Kurt Kinetic Road machine estimated power is about 9% higher than actual power as measured by the power meter on the bike. However session to session the ratio of speed vs actual power is consistent so I'd say that as long as you are not bothered about your actual power and are just looking to work out whether you are getting fitter then using speed on a good fluid turbo is a lot better than nothing. If you use TrainerRoad then as long as you use the same turbo your training zones will be OK just don't believe the power figure TrainerRoad works out.

That's a fair point, but then you probably have one of the most repeatable turbo trainers available. My Cycleops Fluid 2 is also pretty good (though the power/speed relationship is affected by ambient temperature which could be a problem if like me you have your turbo in an unheated garage), but my Elite Supercrono Fluid turbo is woefully unrepeatable.

amaferanga

ut_och_cykla wrote:

My impression is that turbo trainers are sufficiently precise & repeatable too, but using a permanent set up , correctly inflated tyre , calibrated roller pressure etc. Certainly good enough for most peoples requirements anyway. AS to whether they accurately measure power - most people who have checked say they don't, but if they were accurate would they report it? :-)

Not many turbo trainers measure power and those that do cost more than just a power meter anyway.

bahzob

I am not sure there is much benefit in trying to work out power on a turbo if you have no way to measure it on the road.

Any estimate will be hit and miss, any changes will be hard to pin down.

And any estimate will be based on speed so it will be simpler to just use speed in any case.

If you conduct a few ramp tests and max effort intervals you should get a feel for what how far you can ride in given time period. You can use that as an objective basis for calculating training zones and measuring progress. For the purposes of improving turbo fitness this will serve nearly as well as having a power meter.

I do this dispite having a power meter. (I use Cycleops aluminium rollers * along with a Powertap.)

I use the Powertap mainly for data recording for WKO. For turbo work I mainly use it as a distance measure as I do intervals on distance/time rather than power. (It is just I find trying to do, say 10km in 15 mins is more motivating than holding a flat 300W for 15 mins.) One benefit of the rollers is that they are remarkably consistent across sessions so this still ends up producing good results in terms of power and progress.

* As it happens I have a Kurt Kinetic as well. I find the Cycleops rollers more reliable in terms of speed<>power. They do not have the set up issues that turbos have (in terms of getting tyre/roller friction correct every time) nor does the rear tyre heat up anyway near as much (which will also cause variation).

The ones I use are below. They are in your price range. IMO they are worth the extra money over cheaper plastic ones. They are in a different league in terms of build quality and ride very smooth. The resistance unit is non-contact magnetic so very reliable and not subject to wear, one reason power is so consistent.

http://www.google.co.uk/products/catalog?q=cycleops+aluminum+rollers&oe=utf-8&rls=org.mozilla:en-US:official&client=firefox-a&um=1&ie=UTF-8&cid=6219230198246605778&sa=X&ei=OQuFUN_ZC-qM0wXYjIHoBA&ved=0CEgQ8wIwAg

springtide9

Bigpikle wrote:

The basic 'pro' hub is still >£600 before being built to a wheel in most places, and the newer and better G3 is way more than that! They are still commanding silly prices on any forum and ebay - hard to see any quality rim or something that isnt years old and done '000's of miles for <£500.

FYI

I picked up a new Pro Rear Wheel for £580 from Wiggle. They had a "Wiggle sale" again a few weeks ago, and you could get the Pro Rear Wheel for £560 (Pro wireless Hub, Velocity 23 rim, DT Comp spokes)

I would be surprised of the Pro Rear wheels fetch more than £500 even if perfect and has had little use. As suggested, the older versions should be considerably cheaper.

springtide9

bahzob wrote:

The ones I use are below. They are in your price range. IMO they are worth the extra money over cheaper plastic ones. They are in a different league in terms of build quality and ride very smooth. The resistance unit is non-contact magnetic so very reliable and not subject to wear, one reason power is so consistent.

http://www.google.co.uk/products/catalog?q=cycleops+aluminum+rollers&oe=utf-8&rls=org.mozilla:en-US:official&client=firefox-a&um=1&ie=UTF-8&cid=6219230198246605778&sa=X&ei=OQuFUN_ZC-qM0wXYjIHoBA&ved=0CEgQ8wIwAg

How noisy are these rollers? I mean, could they be used in a room next to the sitting room, without the other parties screaming at you to keep the noise down while they watch TV? :-)
Or are these much better used as far away as possible from other householders because of the noise?

MarkAshton

The ONLY way to accurately and precisely measure power is with a specific power measuring device. Period.

Speed (or virtual power as some units call it), is wildly inaccurate.

I run a PM on the turbo. The power:speed relationship is never consistent. Sometimes the difference in speed for the same power is small, sometimes its big. Sometimes it drops for the same power, sometimes it increases for the same power. I have in the past been worried that even my PM was wrong, but I have used a bike with a SRM and PT and the results are consistent.

Too many variables to use speed (virtual power) as a substitute for power. It wouldn't be a problem if it was consistent, but in my experience, its not.

bahzob

"How noisy are these rollers? I mean, could they be used in a room next to the sitting room, without the other parties screaming at you to keep the noise down while they watch TV? :-)
Or are these much better used as far away as possible from other householders because of the noise?"

I use them in the garage with headphones on... However they are not that noisy, I'd say less so than my turbo and subjectively I would say its more acceptable, being more of a low pitched "thrum" than a higher pitched "screech".

It's difficult to say how they much noise they would transmit, I think it depends on the house as much as anything else. My advice would be try some and see what you think.

topcattim

Hmm, this is turning up lots more questions than I had anticipated

I can't/don't want to use rollers because in my warm up/down and in any longer middle-effort sessions, I tend to work at the same time as ride (sad, yes, but efficient, time-wise!). I don't fancy reading or watching videos on rollers, but thanks for the suggestion bahzob.

The general trend here seems to be that power can't be measured that accurately on a cheap turbo, but also that speed isn't a very good measure either. And heart rate isn't too good either, because of drift.

So what is the best way of measuring effort and maintaining effort through intervals - surely not RPE, as this is going to vary by temperature etc.

I kind of figured that power measurements on a turbo, while not accurate, were at least the least worst of all the other options?

Team4Luke

If your only concern is measuring your effort on the turbo and not to utilise that data for road use too, then use speed with a seperate bike computer, ie using a crank and chainstay sensor, Cateye Astrale for example, this is not affected by anything, drift or otherwise unlike power and HR, although cardiac drift isn't as bad as some like to make out.

MarkAshton

Team4Luke wrote:

then use speed with a seperate bike computer, ie using a crank and chainstay sensor, Cateye Astrale for example, this is not affected by anything, drift or otherwise

I am afraid that speed is very much affected by drift. Tyre pressure, fluid viscosity change, heat buildup increasing resistance...

topcattim

MarkAshton wrote:

Team4Luke wrote:

then use speed with a seperate bike computer, ie using a crank and chainstay sensor, Cateye Astrale for example, this is not affected by anything, drift or otherwise

I am afraid that speed is very much affected by drift. Tyre pressure, fluid viscosity change, heat buildup increasing resistance...

So, would turbo-estimated power be the least worst?

bigpikle

depends entirely on your choice of turbo....the KK Road Machines get a good rep for consistency as they use oil filled resistance units that are barely affected by temp change as the unit warms up in use, but they are still very affected by the clamp pressure on the rear wheel, tyre pressure, tyre wear etc etc just like all the others. They can probably be used well enough for good sessions though and highly rate mine.

FatTed

What about the Lemond Revolution, no tyre required
http://www.wiggle.co.uk/lemond-fitness- ... -cassette/
you could buy the power meter later if you required it
couple it with your Garmin for Heart rate and cadence.

MarkAshton

topcattim wrote:

MarkAshton wrote:

Team4Luke wrote:

then use speed with a seperate bike computer, ie using a crank and chainstay sensor, Cateye Astrale for example, this is not affected by anything, drift or otherwise

I am afraid that speed is very much affected by drift. Tyre pressure, fluid viscosity change, heat buildup increasing resistance...

So, would turbo-estimated power be the least worst?

You need to answer this question:

"What am I training for and how much am I bothered (i.e. prepared to spend) if it **might** not be as productive as it can be"

Power can only be accurately measured using a PM. In my experience, even a PT drifts with certain conditions on the turbo. I have heard a Quarq can also drift but the SRM's seems solid. Now... If you really really care about being ultra precise, then spend the £1.7k and get a SRM. If you just want to get fitter, use a cheap speedo to use speed to keep you in the rough ball park area (this goes back to the statement above how it **might** not be AS productive... you could be going a bit to hard or a bit too easy at times, but PE should also be a good indicator as well..)

You dont need to spend mega bucks on turbotraining.

My set-up consisted of a basic magnetic turbo (tacx sirrus - £90), cheap speedo (£20) and a few plans I found on the internet (http://www.turbotraining.co.uk - Free). I saw good gains

amaferanga

topcattim wrote:

MarkAshton wrote:

Team4Luke wrote:

then use speed with a seperate bike computer, ie using a crank and chainstay sensor, Cateye Astrale for example, this is not affected by anything, drift or otherwise

I am afraid that speed is very much affected by drift. Tyre pressure, fluid viscosity change, heat buildup increasing resistance...

So, would turbo-estimated power be the least worst?

These power estimates are based on speed so drift in speed means drift in power guesstimate.

Tom Butcher

MarkAshton wrote:

Team4Luke wrote:

then use speed with a seperate bike computer, ie using a crank and chainstay sensor, Cateye Astrale for example, this is not affected by anything, drift or otherwise

I am afraid that speed is very much affected by drift. Tyre pressure, fluid viscosity change, heat buildup increasing resistance...

Tyre pressure is fairly easy to check, not all turbos use fluid - I'm not convinced there is much heat build up to worry about. No system is 100% accurate - I think realistically you can get a consistent measure using rear wheel speed.

As far as turbo power measurement goes - I haven't compared mine to a power meter but I feel that in a given gear at a given tyre pressure/clamp pressure it is pretty consistent. OK so there are quite a few caveats in there but still a useful measurement in my opinion.

twotyred

I am afraid that speed is very much affected by drift. Tyre pressure, fluid viscosity change, heat buildup increasing resistance

It depends on the turbo trainer. As I said before mine (KK Road Machine) has a stable relationship between speed and power as measured by the on-bike power meter even from session to session and I take no particular precautions such as making sure tyre pressure and the force of tyre to turbo is constant. For short intervals up to 5 mins I think on a decent fluid turbo that speed is a better indicator of effort than HR. For me at least my HR is only begining settle and plateau at the end of a 5 min interval so is essentially useless as an indicator of effort.