Yes, scarcasim when someone is trying to help and making a valid point … That will help! 
1 Like
100% agreed.
BTW, my support ticket is now 15 days old with zero reply from Stages. Is my bike usable? Yes. Can I ‘hack’ it to work acceptably? Yes. Is this an acceptable product or service at the price point paid? No.
1 Like
Right, and I have no problems seeing how that is frustrating.
But with no industry standard way of testing, it becomes a bit hard to say what’s right and what’s wrong. To my understanding, both Powermeters in this case showed the same trend, the same left-right balance. So clearly they are working, there’s just an offset in their outputs. Even DCRainmaker to my knowledge have just chosen one brand, and testing everything else against it.
What would be more interesting to me, is if the OP had several powermeters from the same brand & model, and they were all different!
1 Like
Ok, I’m late to this party. More on this quote in a second…
As for the OP’s data/testing, having finally read through 56 moderately detailed replies, I think he’s done everything right and spot-on to what anyone else should do. It seems clear that it’s not a firmware/crank length/balance issue (since the balance is largely on-spec with the other sources). Thus, it seems like it’s likely a Favero factory calibration issue. Which, isn’t necessarily unheard of actually.
In terms of how I test, I outline that all here*. But, I don’t have one single unit I use. I have multiple units, and all of my day to day stuff that I use is just from regular retail shops. In fact, even my original Favero Assiomas for my in-depth review was bought at retail (because that was before Favero learned how to provide media loaners).
I go through numerous pedals, against numerous power meters. Each of my bikes is equipped with 2+ power meters, plus whatever I’m testing. So most of them have Quarq cranksets, and then some also have dual-sided 4iiii’s built atop those. Some have PowerTap G3 wheels. And I have other power meters floating around I can swap in/out (older 4iiiii ones that aren’t subject to the Shimano issues, Stages ones, SRM, etc…). I’ve got 4-5 sets of Garmin Vector 3/Rally pedals I use, 3 sets of PowerTap P1/P2 pedals, 3 sets of Favero Assioma pedals, and so on. I’ve bought waaay to many pedals, but I don’t like having to swap constantly between bikes with different projects underway.
On the pedal side, I mix and match across almost all trainer/bike reviews, including Favero, SRM, Garmin, and now Wahoo’s will get into the mix too. Using weights or a dyno is a good starting point for validation of the basics, but approximately 0% of power meters fail on that these days (in testing). They fail on human pedaling styles, temp/humidity, or rough road/terrain conditions. None of which automated methods have figured out how to easily replicate (temp can be done to a degree in automation, but it’s messy).
Within most of my trainer/smart bike tests (especially smart bikes), I’ll vary between a few different sets of pedals - especially if I see any issues that cause for pause. Else, since I’ve got a wide range of ‘known good’ pedals, if things are matching up there as expected across my tests, I’m not usually going to swap a bunch for fun.
Ultimately, I’m interested in seeing how the replacement pedals work.
*How to: Troubleshooting Power Meter and Trainer Accuracy Issues | DC Rainmaker
5 Likes
Thanks for the detailed reply.
One quick question, where is the data processed in the typical powermeter? I imagine, there must be so much noise, particularly in offroad applications. Is it done in the power meter unit, the head unit or somewhere else?
Always within the power meter. It’s the processing of that noise that ultimately separates the good from the bad. Or, the newbies from the veterans. Or, the useless from the usable.
It’s all about the algorithms.
3 Likes
The problem with power, is that it is not trivial to know if it is accurate or not. Especially not when you’re talking about human powered “engines”, heck power in the auto industry is ambiguous at best. Where most tuners only agree on power numbers set on a specific dynamometer at a specific time, and weather conditions etc.
But to your point, the industry should agree on a test setup, but I would not hold my breath. I don’t think they have an agreed upon standard on something as simple as frame tolerances for pressfit bottom brackets…
2 Likes
I’m just being practical and my expectations for the bike industry are generally not high. Unfortunately, it’s always a race to the bottom in pricing and volume in the bike industry.
They market all this high tech stuff to us. They give us the impression that these $1000-6000 framesets are made with the same engineering as million dollar carbon fiber wings or helicopter blades. They are obviously not.
One could buy brands like SRM. They’ve made their reputation on being a quasi scientific instrument and with that comes higher cost and maybe having to send it to the factory periodically for calibration service. Most people though want a cheaper option that they never have to send in for service.
And they crash, get viruses, have bugs, spyware, glitches, etc. and we all live with it … unfortunately.
Seriously, I wonder if part of the issue is that there is no standard on how to measure power on a bicycle. A pedal is measuring the pressure on a pedal, left sided measures the bend in the left crank arm, a spider power meter measures crank, a hub power meter (NLA) measures output at the hub. If one’s chain is dirty, then you lose some unmeasured watts in the drivetrain.
Maybe these companies are correct insofar as their power meters are 1-2% accurate for what they are measuring. But pedal vs left crank vs spider are all measuring different things.
I think if one wants four matching power meters then maybe you have to buy all the same model/brand? Maybe one needs to pony up for SRM that built their reputation as being a scientific instrument? Or, use something like the Assiomas that allow you to adjust the power to match your other power meter?
My 0.02 - I bought 3 Quarqs and a calibrated weight, and use the app to do static calibrations every year or two. (I use the Qalvin legacy app as mine are older).
The Quarq design is as good or better than any other crank spider PM on the market according to Keith Wakeham.
IMHO, a lot of single sided and pedal based power meters are “for amusement only”. If you can’t calibrate it yourself, how do you know you have a good one? As someone else already said, you might as well use the HR-based or Strava estimates for power.
I’m unclear what actions you’re expecting. Boycott? Protests? Power meter burning? My Assiomas seem to be fine. I’ve compared them a few times to my elite direto and they’re very close. Never more than a few watts off when I compared a few different intervals and overall.
Im just wondering what you mean by “demand better” are you upset people aren’t more upset?
1 Like
The problem is, that SRM’s simply aren’t as accurate as people think they are - notably because with the exception of the SRM EXACT (which has it’s own issues), and the SRM X (which is good), they didn’t until just recently start doing temperature compensation. Thus, if you rode up a mountain/climb without stopping, it’d be off. In theory, you could post-correct for that if you wanted to in SRM’s software. But nobody in the real-world would.
Their latest SRM Origin editions that were announced this past fall do have that - and they themselves outline the benefit of it - “The integrated temperature sensor allows for maximum accuracy in changing conditions.” Which, ultimately, just makes it like every other power meter on the market that have had that for the better part of a decade.
In terms of this thread, I wouldn’t really let a single unit as is the case here, seemingly dictate an entire market. Nobody is saying 10% inaccuracy is acceptable. But aside from this instance of a unit that’s clearly bad, nobody is seeing that day-to-day.
I don’t know how Favero tests their units on the assembly line. Some companies test X% of units, some test 100% of units, Some test 100% of units but only in a certain portion of the tests.
Either way, there are many of us (including power meter makers) that would like to see power meter and trainer accuracy standards. However, as I outlined above, an indoors lab is really just the very tip of that iceberg. It’s why folks such as GPLAMA and myself do both indoor and outdoor tests, across a variety of conditions.
2 Likes
Two power meters, both claiming 1ish% accurracy, being 10-15% different, and you claim both are working?
All you care about is if power meters from the same manufacturer report the same values??
I am not claiming that I know which one is right - I think it is a very reasonable claim to make that at least one of OP’s power meters is not ‘working as advertised’ using any reasonable definition of ‘working’.
Stated another way, I think any definition of ‘working as advertised’ that claims both of his power meters are ‘working as advertised and in spec’ is not a reasonable definition of working.
Think you’re taking this a wee bit out of contest?
I said since both power graphs are near identical , just offset, then clearly the strain gauges are working.
For OP or any other person with several different power meters, yes that is a problem.
For a person with 1 powermeter the absolute value shown in the display is not critical, as long as it is repeatable. I train inside with fluid trainer & virtual power. I don’t care what the “real” number is, as long as it is repeatable. and I can set my FTP and train accordingly.
2 Likes
I can completely relate with the OP’s comments, although my experience seems to be in the reverse. I can consistently produce 25-30 more watts (with an FTP around 310), when I’m on my assioma’s than when I am on the SB20. This is not an inside/outside issue. I’m using the assioma’s on a spin bike at the gym at work, ant the SB20 at home.
One power meter would be ideal, but often time it just doesn’t work that well. I can’t be moving the power meter for every workout, I just don’t have the time. We need power meters to hit the accuracy numbers the claim. It really causes a lot of challenges. It makes me beg TR for two FTP’s, because so many tools don’t work quite right (especially with AT and AI FTP)
2 Likes
It’s definitely a pain.
I have a pair of Assiomas which I had been using for awhile on my road bike which not sits on my trainer. At some point I forgot to charge my Assiomas so did a workout with my direto and watts felt about right. That’s when I decided I’d do a couple workouts and compare the Assiomas and the direto. Low and behold they were very close. So now I’ve been using the direto as my power source indoors.
Outside on my gravel bike I had a left only stages crankarm. Being left only I’m sure it was not entirely accurate but on gravel I don’t care as much and numbers seemed fairly close to my RPE.
Anyway. Black Friday I got the rally xc200s as I did want dual sided. So decided I should test those against my direto. Well yesterday doing that there was a ~20 watt difference across entire workout and about the same for each interval. So now I have to sort out why. I haven’t calibrated them direto in awhile so maybe that’s different. I torqued the pedals correctly and reset angles and calibrated it.
But the only reason I know this is I have multiple power meters. I agree a common standard would be nice but if I had one power meter like many people it makes little difference. In my case, I could just use the xc200s for everything. It gave me a nice wattage boost. 
I’m not sure if you’re aware, but you can adjust the Assiomas to be closer to the SB20 in the app. (I’m not saying I don’t think there should be a standard, just that this would keep you from needing 2 FTPs)
I did more digging and testing with another set of pedals. Also talked to some engineer friends and other people having similar problems. Our verdict is:
SB20 cranks have a problem of firmware handling the parasitic loads and also figuring out where exactly the load is coming from.
4 Likes
Sigma sports currently have the Assima Duo’s available at £554… wondering if i’m likely to see the price drop significantly lower than this or if it is time to purchase them? The other option would be Garmin Vector 3 Dual Sided at £474!
yeah, I’ve been meaning to do the hanging weight test on the Assioma’s and also to put them on the SB20 to do a real comparison, but I just never have the time for that kind of monkeying around.