so i compared my quarq and my stages left-only (on different bikes) on the same indoor trainer (tacx flux). the test was a ramp from 150w to 300w in 18w steps, 1min at each step (i.e. the first half of the TR ramp test). trainer controlled by its own power, not using external power meter (recorded externally). the average power over each 1min step matches up well between the quarq and the stages.
above is quarq, below is stages. stages gets the average power over 1min correct but is super jumpy. is this expected?
If the averages match well then I wouldn’t worry about it.
A difference that big looks likely to be caused by the smoothing/lack of done by whatever devices are recording the power numbers
Are both recorded at the same rate?
what?! no. Look at the graphs. One is right one is wrong. Get to the bottom of it and throw the bad one off a bridge.
A lot will ha e to do with smoothing as said above. But I also think it makes sense that the one-sided crank-based PM is more jumpy, because it only sees power for half a stroke. The Quarq in the spyder probably gets a more even power application, because of the chain tension on the chainring.
I think either the Stages is not working correctly or the device isn’t recording the power meter correctly. I have a spider PM, hub PM, and a Stages just for fun (it was $99 on closeout, so why not) and I don’t think I’ve ever had a ride where the power looked like that. I’d replace the battery, check the Stages app for firmware, unpair/pair with the recording device, or try recording the Stages on a different device.
I know a handful of people with Stages “power meters” and none are good. I’d never in a million years use one. Spend a few hundred more and get something real.
I have two (on different bikes) and both are fine.
Think again. Power is reported every second. Cadence is probably over 80. So every second has a different proportion of downstroke. So the power numbers are expected to be jumpy, even for somebody with very stable power output.
A lot depends on what smoothing you set to rhe display. In the case above, I’d assume it’s the same, so the difference is real.
There will be smoothing the is in-build to the PM, and there’s nothing you can do about it. It’s likely different companies deal with that differently.
There’s always threads on here where people complain about “jumpy” power, because it’s the first time they’ve seen it with the smoothing turned off.
Not surprised and congratulations. You can’t sell millions without some working fine.
I had a similar issue when I was using Stages LR a couple of years ago. Power graphics looked really jumpy and PowerMatch really didn’t feel right with them. It always felt like there wasn’t enough resistance coming from my trainer. When I eventually gave up on them and changed to Assioma Duos, the difference was like night and day. PowerMatch works much better with the Assiomas and my trainer provides a much more realistic feel.
Assiomas top, Stages bottom.
Some inside and outside threshold efforts around 250W from this year (2021):
Stages dual-sided OUTSIDE:
Stages dual-sided INSIDE TR POWERMATCH (Dec 2018)
Same power smoothness on all of those, at 15-21W standard deviation.
Ramp test Stages single-sided (left only) with PowerMatch (Sept 2018):
Ramp test Stages dual-sided with PowerMatch (March 2019):
Never had a problem with my power meters, as you can see above the ‘smoothness’ is consistent, power graphs look similar, and smoothness is really dictated by my ability to deliver smooth power. These are the 3 power meters:
- Stages Ultegra 6800 left-only
- Stages Ultegra 8000 dual-sided
- Quarq PM on SRAM Force crank
Power graphs also look like those from two gym bikes: Stages SC3 indoor bike (single-sided) and FreeMotion S11.8.
What does that have to do with things? I have 100+ rides with 3 PMs on the same bike, a Stages left only, P2MNG Eco and Powertap G3 hub and my data never looked this bad on my Stages. 30rpm and 100rpm, the Stages tracks similarly to the other PMs. Something is wrong with that power meter.
The Stages PMs aren’t bad, necessarily, but they’d be one of my last choices to measure my work rate and do other PM intensive tasks.
It explains that the jumpiness is to be expected. I have no idea about your experience with all the PMs you have. But the majority of people apply most power in a very small part of their pedal downstroke. Unless they cycle at exactly 60rpm, this means that every second the power reported from the last second will jump up and down. And since the stages reports every second, the jumpiness in the graph is very normal, and doesn’t mean the PM is broken.
As a side note, that does make for an interesting (and easy) experiment. Ride as smooth as possible at a certain cadence, and see if the “jumpiness” changes. If your theory is correct, it should be smoother, the closer you get your cadence to the sampling rate. (Even though my instinct would be it’d be better to oversample, ie ride at 120 rpm, and that eg 90rpm would also give a fairly smooth curve - better than eg 73 rpm).
You do realize that power meters average the power measured during a fixed interval and reports it to the head unit on a per second basis, typically. Or, at least on Garmin, you aren’t really recording at more than 1hz.
Do you want me to make a recording for you with 4 simultaneously recorded power meters (Stages included) and show you that the Stages power should NOT look that jumpy. You can go look at DCR data, too.
The power meter OP has is not functioning properly.
Yes that was my point. It averages over 1 second. Not over one pedalstroke.
No need. I have a stages myself.