I think mine are very accurate. I only see about 1-2 watts difference with my Kickr when I'm on Zwift and using my Assioma linked to my Garmin 530 to record the workout. On outside rides I'm not seeing any obvious variations across similar rides and heart rate. I don't have any other means of comparing power measurement.

Figured some data would help. This is all from 60+ min workouts / races on zwift.

The power isn’t even consistent between zwift and pedals

I calibrate the pedals before every ride, firmware is up to date. Is there anything else I should be doing?

I have a friends Kickr, I might try against that tomorrow

First Q: How are you comparing? Are you just glancing at the screens or are you capturing data on 2 devices and comparing the FIT files? I did several trials where I had multiple intervals on TrainerRoad & on my Garmin with 2 devices. I currently capture my Tacx power on TR and Assioma power on Garmin and compare the overall averages.

Mine are very close. The pedals measure about 3W-5W above my Tacx Neo 2 at ~15 MPH flywheel speeds and >> 200W. This seems about right, given some drivetrain losses.

However... the Tacx Neo and Neo 2 become very inaccurate at flywheel speeds above 25 MPH. So, of you are on the big chainring and a fast cog, trust the Assioma power.

I comparing FIT files over 60+ min rides indoor on zwift. Most of those are races, the latest was an erg mode workout

Zwift connects to computer and Pedals connect via head unit.

I did a ton of aero testing last year and if there's one thing that will show a lack of accuracy or precision it's that.
My goto PM was a Quarq that I would regularly test with a hanging weight.

But I always had my Assiomas well and they were always within a watt or two of the Quarq.

Since November i have been riding a Kickr with the Assomias and a Computrainer with the Assiomas. All 3 within a few watts of each other. I only see the Kickr be off at lower power (150w or less).

Your data post came just after I asked. ;)

Since those are Zwift simulations, I assume that you have many sections at faster speeds. When I am above 25 MPH virtual speed on my Tacx Neo 2, it reads at least 20W low compared to my Assioma pedals.

If you get a chance to compare to a KICKR, I am very curious. Also, try an experiment where you compare at low versus high flywheel speeds on the Neo.

This is how I compare/test mine. I keep my Assioma's on my gravel bike and not on the trainer. However, every once in awhile, including all ramp tests, I move them over to my trainer. I then have the Neo 2 connect to TrainerRoad and the Assioma's connected to a Wahoo Bolt. I hit lap at each interval change* and after the fact compare the overall averages. They are always within 1-2 watts of each other. That's more than good enough for me.

*I have to get my wife come out in the garage for the last 2-3 interval changes when doing a ramp test.

Averages over 60 minutes are a pretty crude way to assess agreement or lack of it.

We share 99% of our DNA with chimps and bonobos. Sometimes it's not how close things are on average, it's exactly when and where they're different.

Ah yes !!! And that's why aero testing is such a great way to see differences in a powermeter.
I can at "a flick of the switch" change from one PM source to another during analysis.
Chung curves never lie :-)

Mine are typically 2.5% higher than my kickr which is in line with expected deviations from measuring at the pedals versus the hub. It's consistent when comparing fit files.

---

Human Person

My wife and I each have a set of Duos, and they seem to be pretty consistent. I tested my Duos against my trainer (Hammer) and my wife's (Kickr Core), and the numbers matched very closely. I tested her Duos in the same way and those numbers matched as well.

Who knows about absolute accuracy, but our Duos are consistent with our trainer, so we can use a consistent power level across our training devices.

The Duos read about 20 watts lower compared to my prior Powertap P1S (single sided) pedals. But again, I'm more concerned with getting consistent readings between my power pedals and smart trainer than anything else.

I wish mine were that close .. I don’t even want to use my assiomas because 15w difference is so material

I averaged 25.7mph during my interval workout today, on Fuego Flats. I was on “big” front ring and middle cassette IIRC. But I think it’s a compact and 12-28 cassette

Maybe that’s why it was so far off today

Will try again tomorrow

I am sure that speed & Tacx high-speed inaccuracy is the problem. Your Assiomas are probably correct.

Here's a test you could try...

• Setup your Tacx in ERG mode so you are on a flat power and let the Tacx control ERG power
• Set a power target at something like 225W or 250W
• Ride some short-ish intervals (a couple minutes each) alternating between big chainring & small chainring
• Pick the rear gearing so that you flip between about 15 MPH and 25 MPH when you change the chainring
  
• Keep cadence constant
• Capture the intervals on two devices, one for each power source
• Then, compare the power differences for each interval

I guarantee that the Assioma will be alternating up and down with the chainring changes, while the Tacx power will be steady. And, the Assiomas should be very close to the Tacx when on the small CR.

So interesting - never thought about this

Quick clarification / dumb question

When I swap to the small CR, and pedal same cadence and same power, my speed shouldn’t change right?

How do I drop to 15mph

Your speed should change. ERG mode will dynamically increase or decrease the resistance to maintain the target power.

You can't really get a 10 MPH drop if you keep cadence steady, but it is OK to vary cadence a little. According to the BikeCalc.com website, if you have a 52/36 CR and are in the 14 cog at 95 RPM, you will be around 26 MPH. Drop the CR and reduce cadence to 85, and that will get you to 16 MPH.

If you set this up and really want to play with the extremes, compare the differences at same Tacx power between a 52-11 and then the 36-25 at 90 RPM. That will take you from 33.5 MPH down to 10.2 MPH. That should show a huge wattage difference swing between the Assioma and Tacx.

I tested at 8 difference cadences / speeds / gears for 1 min each at between 150-250w

My Assiomas and Neo were never within 7w of each other 3%+ off. And for almost all intervals I was 15-20w - about the same I was on the intervals I did this morning

I’m going to try in a race tomorrow to see if it’s any better, and then try the Assioma + Kickr .

If you tested low and high virtual flywheel speeds, and the Duos were never close, then you probably have a problem with the pedals. That sucks. But, the Tacx def has accuracy errors at high flywheel speeds too.

KICKR is probably better for Zwift because of the Neo’s accuracy issues at speed. That said, the big caveat is gplama said that a recent firmware update improved power accuracy on the 2T. I asked him if it rolled down to the Neo and Neo 2, and he said he was not sure. I have not ridden mine since the update because of corona. I will do a retest eventually.

Edit: I just found this about the new update. It pretty much validated the early power errors. Of course, it is bad news for your pedals if you are already on 0.34. https://tacxfaqx.com/...-2-linearity-update/ (And the other beautiful dynamic to Tacx’s test... they were using Duo pedals.)

It’s strange - even at speed (all my Neo vs Vectors were done in races so >25mph) I never noticed Neo have problems at high speeds.

My issue just seems different. The pedals are always 15-20w higher wattage than Neo

A really strange thing - look at how choppy the zwift power curve (recording via ANT+ to MacBook) is vs the Assiomas (recording via Ant+ to garmin 810)

https://imgur.com/a/OLuw1XI

I just updated to the 0.34 firmware today, so I may do a test tomorrow to help provide another look. But, I do suspect a problem with your pedals. The one funky thing to me is that my Tacx power has been way smoother than my pedals. That again casts doubt on your pedals.

What cool did you use for the comparison? Is that DCR’s?

Zwift analyzer

Just feels like something is off.

I’ll try to do my race test tomorrow, then compare Kickr to Assioma, and then maybe call Tacx

Drat. I don’t Zwift. But I will figure out a way to chart it.

I have been enjoying my outside rides, but I feel compelled in the name of science.

Lemme do testing on my side first ...

I’ll replicate that 100rpm freewheel test tomorrow on all gears to see if my power drifts.

Then I’ll test vs the Kickr

Then you can do science =)

Cool. Whatever you want me to test, let me know. I’ll do it. I’m quarantined, after all. I need some good reason to play with my toys.