Power meter inconsistent on turbo trainer

martinheadon

I recently bought a Stages power meter for training. I'm aware that you can ride at the same power and get wildly different speeds on the road, but I'm getting weirdly variable output on my turbo trainer.

As you can see from the attached screenshot, the first 20 minute interval was at 21.2mph and 238 watts. But even though the speed dropped in the second 20 minute interval to 20.4mph, the power was up to 244.



Anyone know what's going on? Is this to be expected or is it a problem I need to solve? I can't think of any variables that might be affecting things.

Alex_Simmons/RST

Not sure you have a problem, other than using two means of comparing power that have an inherit level of inaccuracy and inconsistency.

But by and large, this look pretty typical considering the variability factors involved.

martinheadon

Alex_Simmons/RST wrote:

But by and large, this look pretty typical considering the variability factors involved.

Thanks for the response. I think this is the bit I'm struggling with - what are the variability factors if you're on a turbo trainer? Is it not as controlled an environment as I'm thinking?

RChung

martinheadon wrote:

Stages power meter [...]
I can't think of any variables that might be affecting things.

Hmmm. I can.

Alex_Simmons/RST

martinheadon wrote:

Alex_Simmons/RST wrote:

But by and large, this look pretty typical considering the variability factors involved.

Thanks for the response. I think this is the bit I'm struggling with - what are the variability factors if you're on a turbo trainer? Is it not as controlled an environment as I'm thinking?

Well:
- trainer speed v actual resistance will vary during a session depending on the type of trainer and how the resistance is provided to the bike/rider. e.g. if a tyre is pressed against a roller, then as the tyre warms up/cools down the resistance at the same wheel speed varies. If the trainer uses something like a fluid resistance system, then its resistance will vary depending on the fluid temperature

- trainer speed v resistance will also vary from session to session depending on how firmly the tyre is pressed against the roller, tyre pressure - things like that

- variable power output asymmetry - which is normal - and varies from day to day and also with absolute and relative power, cadence and fatigue amongst other things. Since you are using a Stages power meter, well these naturally variable factors will result in reported power varying from actual power (since the Stages assumes power = 2 x left crank power, which is often not a valid assumption)

martinheadon

Alex_Simmons/RST wrote:

martinheadon wrote:

Alex_Simmons/RST wrote:

But by and large, this look pretty typical considering the variability factors involved.

Thanks for the response. I think this is the bit I'm struggling with - what are the variability factors if you're on a turbo trainer? Is it not as controlled an environment as I'm thinking?

Well:
- trainer speed v actual resistance will vary during a session depending on the type of trainer and how the resistance is provided to the bike/rider. e.g. if a tyre is pressed against a roller, then as the tyre warms up/cools down the resistance at the same wheel speed varies. If the trainer uses something like a fluid resistance system, then its resistance will vary depending on the fluid temperature

Thank you, that all makes sense. I do use a fluid resistance roller, the temperature change in that hadn't occurred to me. I'm a bit of a novice at this, so this is really helpful.

RChung

There's also torque zero drift. So, in summary, there are four things that could have changed:
1. The trainer's resistance
2. The tire/roller interface
3. The power meter's measurements
4. The balance between your L and R legs

One of the oddities from your graph is that the first 5 minute warm-up seems to read low relative to the other 5-minute intervals, but the subsequent 20 minute work interval seems high relative to the other 20 minute work interval. That is, the 1st interval is "low" relative to the 3rd interval but the 2nd interval is high relative to the 4th. They're going in different directions. This is suggestive (but not dispositive) that it's neither the trainer nor the tire/roller interface. That leaves either a drift in the torque zero or else asymmetry between your legs over time. I'd check the torque after the first warm-up but before the first work interval, and again after the last work interval. If the torque is unchanged then the most likely cause is asymmetry. You can't do anything about that with the Stages.

jgsi

But the 'watts' are the 'watts' though arent they?
Surely the Op just needs to disregard turbo speed and just look at the power numbers.. or am I just being too simplistic?

amaferanga

JGSI wrote:

But the 'watts' are the 'watts' though arent they?
Surely the Op just needs to disregard turbo speed and just look at the power numbers.. or am I just being too simplistic?

The Watts are only the Watts if there's little or no ZO drift (possible) and if the OP has fixed asymmetry (possible, but not necessarily). The turbo speed is just odd and inconsistent with the power figures if the turbo is behaving like a typical fluid trainer. If anything you'd probably expect a higher speed for the same power on the 2nd 20min interval when the fluid is fully up to temperature and so less viscous, not a lower speed for higher power.

It looks like the Stages is over-reporting power on the 2nd interval (or under-reporting power on the first interval).

supermurph09

For comparison, this was a recent session I did. Tacx Flow, Stages PM.

RChung

Supermurph09 wrote:

For comparison, this was a recent session I did. Tacx Flow, Stages PM.

Alex_Simmons/RST

Oscar the Grouch. Nice