Favero Assioma Duos / Saris H3

I have a pair of Assioma Duo’s, about 5 years old, with about 50k miles on them. I just got a new Saris H3 Smart Trainer. There is a significant discrepancy in the power readings to the tune of 30+ watts, consistently over multiple rides on different days. I get that they wouldn’t be totally harmonious but it’s my understanding that typically the pedals might read higher and/or “spikier” power because they are closer to the generating source and more sensitive to input. However, this is not the case. The pedals are reading the lower power numbers. Everything has been calibrated, etc. properly on every ride. My question is whether power pedals can wear out or degrade over time?

The power displayed in app during a workout will default to the power meter, not to the trainer. But the trainer is providing the resistance. In effect- regardless of which reading is correct- I’m having to compensate for that 30+ watt discrepancy. For example, an interval may be for 200 watts. The trainer reads 200 watts but the pedals are only reading 170 and that’s what the app displays. This forces up the resistance on the trainer so that the pedals will read 200. But then trainer is then reading 230. A few watts off would be no big deal to me. I don’t care so much about the absolute power and which one is correct per se, but consistency is important. This is to the tune of 10-20% off at times. Currently, I’ve just turned off the pedal reading so that the trainer can just do it’s thing. It seems to respond faster to changes this way. I’ve had a month off the bike from injury so I expected a loss of fitness and compensated by adjusting my FTP way down. I’ve also moved from mostly outdoor rides to indoor rides so it’s somewhat difficult to assess which device is off, though the pedal reading feels more out of whack, ego aside.

When you pair a power meter and smart controlled trainer, TR will default to using PowerMatch:

As such, the delta’s that exist between the devices are largely irrelevant, because TR is setting your power targets based upon the data of the power meter. This is because they assume (and most people expect) that the power meter is the “source of truth” because it is presumed to be used inside and outside. So, despite any discrepancy between the units, the power meter consistency is what matters and why TR maintains that.

  • Even when people have power meters that are the same, changing trainers can have a real and functional impact on your ‘FTP’, assuming you determined it with the prior trainer in use. This is because things like flywheel performance, along with gearing and/or trainer modes in use can all affect the final value of ‘FTP’ determined with each trainer.

  • As such, my blanket recommendation is to retest your FTP with the new trainer. Ignoring the power data variance, it is possible that your FTP will vary even when measuring using the power meter pedals.

  • Any device can become inaccurate over short or long periods of time.
    • When in real doubt, checking against other known “good” power devices can work.

    • With some power meters, like your pedals, you can also do the static weight evaluation test to see if the power meter is generally functioning properly. A bit of google searching on your power meter will show you the process and requirements.

Hey @mcneese.chad – don’t want to hi-jack the OPs thread, but a related question.

What if you’re wanting to use power pedals for outdoor riding and rely on the H3 for indoor riding? In other words, you don’t want to move your pedals around between rides.

Do you just live with the discrepancy of power outputs an impact to AI FTP, etc?

Obviously, it would be great if they matched or both were within spec (even if each is at the opposite end of the tolerance range with one HI vs other LO), but as we see this type of split more often than we’d all like.

If you ask your question with the TR realm of reps, they say not to worry about it… that AT and the surrounding world within TR’s service will handle it. Personally, as someone with more power devices than a sane person should have (8 power meters, 4 smart trainers) I have no idea how TR can possibly deal with the variation that exists.

Even in a “simple” example we see here that is common (one PM, one trainer), the delta seems likely to impact the TR experience. It will also depend a fair bit on the distribution of rides (more time on one than the other), as well as all the usual variables that may make things like the setting of power targets and even AIFTPD all subject to some questionable gaps in power data values.

Some power meters at least, offer the option to shift the values up or down to match other known sources. 4iiii & I think Assioma can be altered to more closely follow things like smart trainers. But this may be less than ideal on a number of levels.

Most basic is that power meters and smart trainers may not all have the same delta from one end of the power spectrum to the other. A delta might be 5% at one end while 10% on the other, with whatever between them. So using the offset mention above may be good at one power level, but lead to differences at other points.

To be honest, it’s all a massive mess and short of some very basic initial testing between my power meter pedals and smart trainers, I have largely ignored it myself. It’s not that it doesn’t matter, because it might and probably does, but doing anything meaningful about it is not simple. The testing we see done in too many cases here is elementary and often flawed. It lacks detail and/or thoroughness that is appropriate to make meaningful claims about these delta’s. I have no doubt they exist, but the level of confidence people show in some of there testing oversteps the reality of the process & detail they applied.

I admit that this is a bit negative slant, but we get firm claims based upon questionable data more often than not. It takes real planning, effort and implementation to get decent test data that is appropriate to make claims and take any action. And even when we get good data, there is not always a practical solution to apply.

Such is the state of “power” in the cycling world. It’s messy and far more complicated than most device makers and app suppliers might want to recognize. A bit of the “Wild West” I think is how Shane Miller has mentioned at times, and I agree with that basic assessment.

Sorry this is more of a rant than an offering of solution, but there is not much useful light that I see in this space.


fwiw, i also have a pair of 4 yr old duos. i recently did the static weight calibration, and spent some time verifying the test weight across a few sets of scales to 3 decimal places. the test revealed no adjustment was required.
i then ran both my pedals and 5 yr old tacx neo on bluetooth and clicking on “paired devices” i can see both readings sets. ive also unpaired the pedals and checked the readings on my headunit. they were within 1 or 2 watts over a few minutes. i also checked a new set of duos are all were the same, within a couple of watts, max. oddly, in time past, the neo ran about 9 watts lower than the duos but that seems to be gone.
i can only suggest checking your L/R balance is normal for you. a recent knee issue has thrown mine off a bit, thus possibly reducing my overall power.
if i had a 30 watt swing in data, one of those devices would have to be king.

The only thing you can do is finding 3rd power meter. My normal testing setup is: pedals (I own 2 Assioma duos), crank powermeter (Ronde, Inpeak, Powerbox, 4iiii) and thrusted trainer (Neo, Kickr V5, Direto XR, Saris H3). Adding 3rd source of truth will help you.

1 Like