Kinetic Road Machine power accuracy

whoawhoa said:

I've been looking at the Kurt Kinetic Road Machine trainer for the winter. For an extra 50 bucks they have a speedometer that claims to accurately estimate the power based on speed of the roller. They claim that they are "powertap calibrated" and very accurate except for accelerations. This sound good? Has anyone ridden on one with a powertap and compared numbers? It's much cheaper than any powermeter or a computrainer.

They generate an estimate of power based on roller speed and assumptions of frontal area, drag coefficient, rider weight, etc. They can even be "customized" to a particular owner's body characteristics.

A friend of mine had one and had a very difficult time getting his readout to match www.analyticcycling.com power values for a range of speeds. I haven't seen a comparison vs. a on-bike PM.

frenchyge said:

They generate an estimate of power based on roller speed and assumptions of frontal area, drag coefficient, rider weight, etc. They can even be "customized" to a particular owner's body characteristics.

A friend of mine had one and had a very difficult time getting his readout to match www.analyticcycling.com power values for a range of speeds. I haven't seen a comparison vs. a on-bike PM.

I think we're confused-I meant it estimates power when riding the trainer (so only based on speed of the trainer-frontal area, etc wouldn't come into play, right).

whoawhoa said:

I think we're confused-I meant it estimates power when riding the trainer (so only based on speed of the trainer-frontal area, etc wouldn't come into play, right).

Not confused, I'm saying the same thing. Frontal area, etc come into play because they are programmed into the unit to set the power vs. speed curve for the computer. The device reads wheel speed *only* (no resistance or torque). The power is then *estimated* from the default parameters and measured speed, rather than measuring power directly from the resistance unit.

In other words, the computer is programmed to know that a rider of XX size should require 200w to do 20mph on a flat road. Ride the trainer at 20mph and it will show 200w. Drain all the fluid out of the unit and ride 20mph --> still shows 200w.

Anyway, back to the accuracy, my friend even found an online tool where the rider parameters (frontal area, weight, etc.) could be changed in the computer unit, and was still unable to get the readout to match the power curve he expected. That said, I think precision is more important than accuracy for power readings, especially if you don't have an on-bike unit to compare to.

frenchyge said:

Not confused, I'm saying the same thing. Frontal area, etc come into play because they are programmed into the unit to set the power vs. speed curve for the computer. The device reads wheel speed *only* (no resistance or torque). The power is then *estimated* from the default parameters and measured speed, rather than measuring power directly from the resistance unit.

In other words, the computer is programmed to know that a rider of XX size should require 200w to do 20mph on a flat road. Ride the trainer at 20mph and it will show 200w. Drain all the fluid out of the unit and ride 20mph --> still shows 200w.

Anyway, back to the accuracy, my friend even found an online tool where the rider parameters (frontal area, weight, etc.) could be changed in the computer unit, and was still unable to get the readout to match the power curve he expected. That said, I think precision is more important than accuracy for power readings, especially if you don't have an on-bike unit to compare to.

Here's a quote from the the kurt website: "The Kinetic PC operates on the principle that the Road Machine's resistance is speed sensitive and ONLY speed sensitive. Some people don't buy into that, but if you think about it, the only factor that has any impact on the resistance unit is how fast the roller is spinning - independent of rider weight, cadence, crank length, etc. After riding one of our trainers, you are, no doubt, aware of a relationship between speed and effort. The faster you go, the harder it is to pedal. The beauty of the Kinetic Road Machine is that we have tweaked and studied this relationship to a further extent than any other company."
kurtkinetic.comcomputer tech.phpOpen ↗ thats the link to read the whole thing.

frenchyge said:

Anyway, back to the accuracy, my friend even found an online tool where the rider parameters (frontal area, weight, etc.) could be changed in the computer unit, and was still unable to get the readout to match the power curve he expected. That said, I think precision is more important than accuracy for power readings, especially if you don't have an on-bike unit to compare to.

That doesn't make any sense. Kurt Kinetic Road Machine's resistance unit, as is the case with all fluid resistance units on trainers, a closed system that's dependent only on the rotation speed of the roller and the viscous resistance response of the fluid in the unit. Putting aside variations in the fluid's viscosity due to temperature, the force with which it resists the roller's motion depends solely on the speed with which the spinner wings within the resistance unit are moving through the fluid. As a result, at any sustained wheel speed, the power required to spin the resistance unit is the same irrespective of body position, frontal area, etc.

The only way in which frontal area and other aerodynamic factors would come into play is if you wanted to correlate speeds measured on the trainer to outdoor riding, but that's not very useful (in my humble opinion). It sounds like the original poster wants to treat the trainer as a purely power-driven tool (i.e. ride at X Watts, not simulate Y mph of outdoor riding). For that purpose, the Kurt Kinetic readout should be sufficient, within the tolerances of their curve fit of the resistance unit's response.

I use an even more lower tech approach of having a laminated chart showing the wheel speeds required for a range of power outputs and use them to drive my workouts...

Berend

Ok, I gotcha. The power curve is set to the trainer and not the rider. I was thinking of the AC Custom Pro trainer where they set the resistance unit to match the rider. Sorry. :o

It says +/-3% deviation from a Powertap, and they claim that their fluid resistance doesn't change with temperature (although they still put on a big heat sink, for some reason 😉 ). A change in the fluid would be the only reason the accuracy would get out of whack once it is calibrated to the unit.

squidwranglr said:

It sounds like the original poster wants to treat the trainer as a purely power-driven tool (i.e. ride at X Watts, not simulate Y mph of outdoor riding). For that purpose, the Kurt Kinetic readout should be sufficient, within the tolerances of their curve fit of the resistance unit's response.

Agree. I was misunderstanding the device and intent.

squidwranglr said:

I use an even more lower tech approach of having a laminated chart showing the wheel speeds required for a range of power outputs and use them to drive my workouts...

I started one of those when I first got my Powertap. The problem is that my Mag trainer heats up so much during a ride that the sheet is worthless at any temperature except the one at which it was created. If Kinetic's claims are true then I wouldn't expect that to be an issue with their fluid trainers.

squidwranglr said:

That doesn't make any sense. Kurt Kinetic Road Machine's resistance unit, as is the case with all fluid resistance units on trainers, a closed system that's dependent only on the rotation speed of the roller and the viscous resistance response of the fluid in the unit. Putting aside variations in the fluid's viscosity due to temperature, the force with which it resists the roller's motion depends solely on the speed with which the spinner wings within the resistance unit are moving through the fluid. As a result, at any sustained wheel speed, the power required to spin the resistance unit is the same irrespective of body position, frontal area, etc.

The only way in which frontal area and other aerodynamic factors would come into play is if you wanted to correlate speeds measured on the trainer to outdoor riding, but that's not very useful (in my humble opinion). It sounds like the original poster wants to treat the trainer as a purely power-driven tool (i.e. ride at X Watts, not simulate Y mph of outdoor riding). For that purpose, the Kurt Kinetic readout should be sufficient, within the tolerances of their curve fit of the resistance unit's response.

I use an even more lower tech approach of having a laminated chart showing the wheel speeds required for a range of power outputs and use them to drive my workouts...

Berend

Which trainer to you do this with? I do this with a minoura I have now, but I doubt it's accurate.

frenchyge said:

A change in the fluid would be the only reason the accuracy would get out of whack once it is calibrated to the unit.

That, and 1) the force with which the roller presses on the tire, 2) the tire pressure, 3) the tire model/brand, and 4) the tire's age.

acoggan said:

That, and 1) the force with which the roller presses on the tire, 2) the tire pressure, 3) the tire model/brand, and 4) the tire's age.

1, 2, and 3 are easily controlled. Is it fair to assume the 4th will cause fairly small differences?

It's probably as accurate as the iBike...

whoawhoa said:

I've been looking at the Kurt Kinetic Road Machine trainer for the winter. For an extra 50 bucks they have a speedometer that claims to accurately estimate the power based on speed of the roller. They claim that they are "powertap calibrated" and very accurate except for accelerations. This sound good? Has anyone ridden on one with a powertap and compared numbers? It's much cheaper than any powermeter or a computrainer.

The Kurt power odometer or whatever it is called is fairly accurate. I have a friend with one who later bought a PT and they were pretty close. Not 100% on but within 4-5% most of the time when pedaling a steady pace.

BikeInMN said:

The Kurt power odometer or whatever it is called is fairly accurate. I have a friend with one who later bought a PT and they were pretty close. Not 100% on but within 4-5% most of the time when pedaling a steady pace.

Thanks. Exactly the type of answer I was looking for.

whoawhoa said:

1, 2, and 3 are easily controlled. Is it fair to assume the 4th will cause fairly small differences?

Indeed, I find that as long as I'm using the same bike with the same wheel and tire at the same pressure and with the rear QR clamps set in same position, the coast-down calibration on my Velodyne is highly reproducible. However, many trainers don't allow you to fix the position of the roller relative to rear axle, and/or people use relatively crude procedures such as "X turns after the roller contacts the tire" to try to keep this constant. Consequently, rolling resistance and therefore the overall power requirement can vary significantly.

acoggan said:

Indeed, I find that as long as I'm using the same bike with the same wheel and tire at the same pressure and with the rear QR clamps set in same position, the coast-down calibration on my Velodyne is highly reproducible. However, many trainers don't allow you to fix the position of the roller relative to rear axle, and/or people use relatively crude procedures such as "X turns after the roller contacts the tire" to try to keep this constant. Consequently, rolling resistance and therefore the overall power requirement can vary significantly.

Would it be smart to do my own sort-of coast down test every time I road the trainer? Just accelerate to a certain speed, stop pedaling, and time how long before the wheel stops spinning, then adjust to a set time?

whoawhoa said:

Would it be smart to do my own sort-of coast down test every time I road the trainer? Just accelerate to a certain speed, stop pedaling, and time how long before the wheel stops spinning, then adjust to a set time?

That would probably help a lot, but whether you really want to go to all that much trouble every time is sort of up to you. You may find that simply keeping everything else as constant as possible is good enough for day-in and day-out training purposes. It would, though, be a good idea to perform such a calibration prior to any sort of fitness testing you might do.

I agree with that statement. I have a kurt trainer, and like it alot. It is very close to road resistance, but varies based on "x # of cranks" so the watts are probably accurate within some range of fudge factor. Crank it down to keep it from slipping during intervals and it seems harder than the road...or maybe it just sucks riding on my back porch. What would be a great idea would be to make a frame/trainer that would eliminate the back wheel, and be chain driven off the chainring. Maybe mounted right below wher the front wheel is.Pre programed for a 52 or 53 watts would be exact based on rpm..i think Thanks BP

acoggan said:

Indeed, I find that as long as I'm using the same bike with the same wheel and tire at the same pressure and with the rear QR clamps set in same position, the coast-down calibration on my Velodyne is highly reproducible. However, many trainers don't allow you to fix the position of the roller relative to rear axle, and/or people use relatively crude procedures such as "X turns after the roller contacts the tire" to try to keep this constant. Consequently, rolling resistance and therefore the overall power requirement can vary significantly.

Okay, I've almost decided to go ahead and buy the kurt with the power computer, for a few reasons:

1. Much cheaper than a powertap.

2. I only ride my road bike for about half my annual hours, so I wouldn't be able to track tss and stuff anyway.

3. I will need a new trainer soon anyway.

Biggest downside is no np, as far as I know. Does anyone know of a way to save the data and download it into cycling peaks?

Also, any other opinions on accuracy? Billsworld, don't you have a pt and one of these trainers now?

whoawhoa said:

Okay, I've almost decided to go ahead and buy the kurt with the power computer, for a few reasons:

T.

Biggest downside is no np, as far as I know. Does anyone know of a way to save the data and download it into cycling peaks?

Also, any other opinions on accuracy? Billsworld, don't you have a pt and one of these trainers now?

I have the computer that comes with the Kurt. I have my track bike on it . 48x14.....16mph =160 watts 17mph=180watts 18mph=204watts. I cant do hard sprints on it without slipping. I had my road bike on it with the power tap. My "guess" is that it is fairly accurate or even a bit low on true watts as long as your not slipping. I have a set of tubulars with power tap on the way from www.wheelbuilder.com for the track bike. I will leave the kurt computer next to it for a bit and Ill give you feedback. Analytacal sells that trainer. He might have a clue on getting info to peaks That computer is a little cheepo computer , so I doubt it. Last thing, I got mine from Analytical with the larger fly wheel. I think frenchguy has a Kurt with a pt, and velomanct had it too.......your a motivated dude if you can put half your annual miles on that thing😄