More power = more speed?

[swampy1970]

So, I may have a little issue with the powertap but before I give the Saris folks a call I thought I'd throw this out for comment.

What I've noticed on my PT (2.4 SL, latest firmware) is that if I ride in one gear on the trainer I get a situation where speed matches cadence (ie you pedal faster the speed goes up) yet there are odd times, over periods of 5 to 10 minutes where for a given speed, power will drop 30 to 40 watts. Cadence, speed, heart rate and percieved effort remain the same. To compensate for this "loss" I could pedal faster and consequently end up going faster for a "given power output."

One way I've noticed to 'rectify' this situation is to stop pedalling for 20 to 30 seconds... not something you'd want to do every once in a while, especially during a test.

I haven't tried swapping out any of the batteries in either the computer or hub yet. Not sure if that would help.

At first I thought there could be something weird and wonderful going on with my trainer. I made sure all parts were clean and that magnet was firmly held in place for a given resistance setting... nothing unusual or untoward looking there.

Any ideas from experience PT folk?

 

[rmur17]

So, I may have a little issue with the powertap but before I give the Saris folks a call I thought I'd throw this out for comment.

What I've noticed on my PT (2.4 SL, latest firmware) is that if I ride in one gear on the trainer I get a situation where speed matches cadence (ie you pedal faster the speed goes up) yet there are odd times, over periods of 5 to 10 minutes where for a given speed, power will drop 30 to 40 watts. Cadence, speed, heart rate and percieved effort remain the same. To compensate for this "loss" I could pedal faster and consequently end up going faster for a "given power output."

One way I've noticed to 'rectify' this situation is to stop pedalling for 20 to 30 seconds... not something you'd want to do every once in a while, especially during a test.

I haven't tried swapping out any of the batteries in either the computer or hub yet. Not sure if that would help.

At first I thought there could be something weird and wonderful going on with my trainer. I made sure all parts were clean and that magnet was firmly held in place for a given resistance setting... nothing unusual or untoward looking there.

Any ideas from experience PT folk?

Is this clearly discernible from the downloaded file or from (and I'm not saying you're wrong) what you read on the PT display?

My 1st thoughts are on the well-known 'aliasing' phenomena that occurs with the PT. Fixed reporting rate of 1.26 seconds and a cadence that matches or doesn't match that. It looks bad on the display but invariably average power over a reasonaly long interval evens things out. Another common 'solution' is to employ 3-5 sample averaging in the CPU to lessen the jumpiness -- but the problem still reamains to some degree.

In contrast, SRM reports average power per complete crank cycle and, ceteris paribus, that's generally much smoother than the PT.

OTOH, could be something wrong with your PT or the wireless interface ..

 

[rob of the og]

So, I may have a little issue with the powertap but before I give the Saris folks a call I thought I'd throw this out for comment.

What I've noticed on my PT (2.4 SL, latest firmware) is that if I ride in one gear on the trainer I get a situation where speed matches cadence (ie you pedal faster the speed goes up) yet there are odd times, over periods of 5 to 10 minutes where for a given speed, power will drop 30 to 40 watts. Cadence, speed, heart rate and percieved effort remain the same. To compensate for this "loss" I could pedal faster and consequently end up going faster for a "given power output."

One way I've noticed to 'rectify' this situation is to stop pedalling for 20 to 30 seconds... not something you'd want to do every once in a while, especially during a test.

I haven't tried swapping out any of the batteries in either the computer or hub yet. Not sure if that would help.

At first I thought there could be something weird and wonderful going on with my trainer. I made sure all parts were clean and that magnet was firmly held in place for a given resistance setting... nothing unusual or untoward looking there.

Any ideas from experience PT folk?

I'd still be blaming the turbo trainer. It will behave slightly differently at different temperatures. On my iMagic (electro-magnetic) the power readings from the trainer vary as the roller heats up and cools down, and the makers are aware and advise to calibrate the machine after a long warm-up period.

I don't look at the speedo (PT is set for power, cadence and time) but certainly the gear that I use will 'drift' for a constant power output across a session. The fact that 30 seconds of cooling down time for the roller fixes the "problem" indicates to me where the "problem" is coming from.

 

[swampy1970]

I'd still be blaming the turbo trainer. It will behave slightly differently at different temperatures. On my iMagic (electro-magnetic) the power readings from the trainer vary as the roller heats up and cools down, and the makers are aware and advise to calibrate the machine after a long warm-up period.

I don't look at the speedo (PT is set for power, cadence and time) but certainly the gear that I use will 'drift' for a constant power output across a session. The fact that 30 seconds of cooling down time for the roller fixes the "problem" indicates to me where the "problem" is coming from.

The trainer has nothing fancy too it and I have checked for heat issues - but it's just a plain Jane steel disk with a magnet. There's nothing too heat up. Two flywheels and a steel disk:

 

[swampy1970]

Is this clearly discernible from the downloaded file or from (and I'm not saying you're wrong) what you read on the PT display?

My 1st thoughts are on the well-known 'aliasing' phenomena that occurs with the PT. Fixed reporting rate of 1.26 seconds and a cadence that matches or doesn't match that. It looks bad on the display but invariably average power over a reasonaly long interval evens things out. Another common 'solution' is to employ 3-5 sample averaging in the CPU to lessen the jumpiness -- but the problem still reamains to some degree.

In contrast, SRM reports average power per complete crank cycle and, ceteris paribus, that's generally much smoother than the PT.

OTOH, could be something wrong with your PT or the wireless interface ..

It's discernable on both the display and the file. On the file (using the software that comes with the PT) it's more noticable as there's no longer a nice correlation between power and speed - one goes down and the other goes up.

At first I was just thinking that I was getting tired and my poor legs just couldn't keep the effort going despite being 10 to 15bpm from where I'd expect to be near threshold - then I noticed that I was pedalling faster in the same gear and was putting out less power.

Unfortunately, Comcast photos doesn't give photos a usable file extension so I can't post from there. I'll get an account with some photo place that I can link screenshots from when I get home.

 

[daveryanwyoming]

The trainer has nothing fancy ....just a plain Jane steel disk with a magnet. There's nothing too heat up. Two flywheels and a steel disk:

Actually, magnetic braking like your trainer relies on eddy currents induced in the disk near the magnets. Eddy current braking systems do heat up (the braking force you feel and work against is converting your efforts to heat dissapated in the disk) and when they heat up their braking changes. It's why you have to warm up a CompuTrainer prior to roll down calibration, the braking resistance changes as the rotors heat up.

I have no idea whether your trainer's resistance curve is changing enough to account for the power vs. speed changes you're seeing, but mag trainers aren't exempt from braking force drift due to heating.

-Dave

 

[frenchyge]

I'd still be blaming the turbo trainer.

That's where I am initially. Especially since a 20-30 second cool down period restores the resistance to normal.

Does the problem occur on lower-power rides, or only on higher-power testing? Each trainer has some ability to dissipate heat (ie, a certain number of watts) and remain stable, but if the power input exceeds that ability then temps rise uncontrollably.

Re: nothing to heat up..... next time it happens, jump off the bike and touch your tongue to some metal part on the left side of that resistance unit.

(just kidding... don't)

 

[rmur17]

It's discernable on both the display and the file. On the file (using the software that comes with the PT) it's more noticable as there's no longer a nice correlation between power and speed - one goes down and the other goes up.

At first I was just thinking that I was getting tired and my poor legs just couldn't keep the effort going despite being 10 to 15bpm from where I'd expect to be near threshold - then I noticed that I was pedalling faster in the same gear and was putting out less power.

Unfortunately, Comcast photos doesn't give photos a usable file extension so I can't post from there. I'll get an account with some photo place that I can link screenshots from when I get home.

okay it doesn't sound like aliasing then. I got the impression it happened rapidly at a very specific cadence etc and that threw me onto a tangent.

It is quite gradual then? I believe most, if not all, electrically braked ergometers suffer from some degree of drift due to heat buildup. I've experienced modest drift in ergo mode with my CT -- added a small external cooling fan to virtually eliminate this. But the drift was pretty slow over say tens of minutes vs. tens of seconds.

 

[swampy1970]

Actually, magnetic braking like your trainer relies on eddy currents induced in the disk near the magnets. Eddy current braking systems do heat up (the braking force you feel and work against is converting your efforts to heat dissapated in the disk) and when they heat up their braking changes. It's why you have to warm up a CompuTrainer prior to roll down calibration, the braking resistance changes as the rotors heat up.

I have no idea whether your trainer's resistance curve is changing enough to account for the power vs. speed changes you're seeing, but mag trainers aren't exempt from braking force drift due to heating.

-Dave

It doesnt get that warm though...

... I could try putting a small fan near it just too see.

 

[daveryanwyoming]

It doesnt get that warm though.....

It may not seem very warm, but if you ride the trainer at 200 watts for roughly 35 minutes its dissipating enough energy to bring a liter of water from just above freezing to a full boil. All that heat has to go somewhere....

I really have no idea whether you're experiencing trainer drift or if your PM has an issue but just don't think any trainer is immune to changes in resistance as you ride it.

-Dave

 

[Piotr]

FWIW, power drift is real! Lately, I've been forced to reset the torque during some indoor rides, because the upward drift at certain speed is undisputable. Why and how it happens is still unclear to me, but it happens during non-stop isopower riding and resetting the torque fixes it. Until I can resolve this issue I can't do an FTP test. (or maybe ).

 

[dkrenik]

It doesnt get that warm though...

... I could try putting a small fan near it just too see.

Eh, I dunno. I ride Kreitler rollers with a "Killer" headwind unit for resistance. I definitely notice a change throughout the workout as I need to pedal faster (or at an increased speed) to maintain the same wattage (0-offset on my SRM is steady). I figure the belts must be warming up.

I don't think that it takes much in the way of heat build up/dissipation to make a discernible difference in speed.

Dave

 

[swampy1970]

So I put a fan blowing directly on the roller just too see if there was any effect - and it seems as though I'm still getting the same issue. The fan, despite being pretty small does move a lot of air and it was pretty darned chilly...

As you can see I changed gear right at 28 minutes - about a minute after the 'problem' occoured and mysteriously resolved itselfs at around 32.5 minutes.

Any ideas?

Despite being new it looks as though I'm gonna need some new batteries in the headunit real soon. Hmmm...

 

[dkrenik]

Huh, that's kinda weird. My first guess would be either poltergeists or alien intervention and I'll defer to those more knowledgeable in the mechanics of trainers and witchcraft for comment.

Other than that "incident" ~28 min's or so your power tracks to speed pretty directly.

Dave

 

[swampy1970]

Huh, that's kinda weird. My first guess would be either poltergeists or alien intervention and I'll defer to those more knowledgeable in the mechanics of trainers and witchcraft for comment.

Other than that "incident" ~28 min's or so your power tracks to speed pretty directly.

Dave

Kinda weird that both power and torque 'head south' for about 5 minutes before returing to normal...