BikeErg Power Meter Accuracy

[jaradgiese]

I recently added Garmin Vector 3 power meter pedals to my BikeErg. I set the crank length to 170mm as I found in the BikeErg specs, calibrated both the BikeErg and the Vector 3 pedals, but have found after hundreds of kilometers that the Vector 3 power meter consistently reads ~15-19 watts higher than the BikeErg. My question is in regards to the accuracy of the BikeErg wattage that is displayed? Has anyone else added a power meter to their BikeErg, and if so how did the watts compare? I ride a lot on Zwift and would be happy to find I am slightly stronger than my BikeErg based FTP suggested, but at the same time I really don't know if that is the case and am interested to know which is more accurate, and why the readings would differ so much between them.

[jamesg]

The C2 system measures the work done by the flywheel fan pumping air, but not the inertial losses due to our moving body mass.

Could be strain gauge systems in pedals can account for this, if they measure the forces acting there.

[Nicola]

I did the same test with Favero ASSIOMA pedals and had same results (15-20w difference)
I sent an email to support that told me that C2 measure power generated by the fan and it doesn't take into account the power losses.
The 2% error claimed by C2 is related to machine to machine (consistency between ERGs).
I suggest to use your Garmin pedals when training in Zwift

[jaradgiese]

Thank you James and Nicola! This power difference makes sense if measuring off of the flywheel instead of the direct strain at the pedals. It is reassuring how consistent the difference has been, which was why it was hard to reconcile as to which would be more correct, great that they are both correct - just in different ways!

I will switch my Zwift cycling to the pedals, which is a nice free boost to my FTP! I have been using the North Pole Engineering 'Cable' fob to transmit power/cadence/heart-rate to Zwift since my Bluetooth connection is to the ErgData app on my phone for the online C2 log. Now I can setup the Cable device to just read from my SkiErg and Rower, which will simplify what has become a tangle of apps and devices to make this all work across various applications and machines to log as much data as possible (Garmin, Zwift, Sufferfest, Strava, yikes!) Thanks again for the responses, this is exactly what I was hoping to discover!

[Marbuel]

It would be great if Concept2 could apply certain correction so that the BikeErg will show the actual Watts.

[NavigationHazard]

There is no one "true" way to measure "actual" cycling wattage against theoretical rider-output values. All known power meters are subject to at least some degree of measurement error. In general, variation from manufacturer to manufacturer seems to be greater than variation in individual units.

See e.g. https://road.cc/content/feature/258997- ... wer-meters

And https://www.thieme-connect.de/products/ ... 043-102945, which alternatively is linked to in the article above.

[Marbuel]

I am confident that reproducibility is high using flywheel. The measurements are precise. Same effort gives you the same number of watts. But the measurements might not be accurate. Meaning that the number is off by certain concrete and always similar amount. This could be due to loss between pedals and the flywheel. The difference could thus be a function of power or cadence or drag. Seems like this is easily 25W, which is a lot. Typical power meter has 1-2% error!

The only way to know would be to mount power meter in the pedals and compare, as the some did. And ideally also different power meters, both legs, different efforts, cadance, drag etc.

Basically, BikeErg measures power output at the flywheel but most power meters measure power input on pedals. So the only way to get comparable data from your normal bike and BikeErg is to install clip in pedals with power meters.

It sucks that the crank arm cannot be exchanged on BikeErg. It's not a standard bike arm. There are some good and cheap crank arm power meters. I don't want to have clip in pedals on BikeErg. And there is no pedal power meter with normal flat pedals.

So ... You have to live with this issue and accept that all the stats you are getting from BikeErg are not comparable to bike smart trainers or power meters. Would be great if one could somehow set a compensation on PM5 ... or at least change the crank arm!!

For example all the stats that Garmin shows are thus off! We just don't really know by how much... But anything more than 10 watt is already quite a lot and at higher efforts this could be much more.

I understand why Concept2 did the power measurements the way they did ... It's now consistent with their other machines ... But for cyclists it is not ideal.

Best,
Marek

[LloydyH85]

Hello everyone!

I tested my Bike Erg this morning using the Assioma Duos, My readings seem to be pretty consistant with what most of you have found, Typically under by 15-20 watts. It seemed to me that it was worse at lower wattage than when i was pushing close to ftp as it did seem like it stabilised itself after some time. Its not a huge issue for me as i don't race much on the platform but would probably put the pedals on for Time Trials or FTP Tests etc Not Ideal but i do love the bike itself! Its built like a tank and you recieve the amazing support from the guys at Concept 2. What i am curious about is if the new flywheel (Retrofit Model) Makes any difference in these results? I am guessing it woudn't but thought it was worth mentioning. I am using one of the early bike ergs with the orginal flywheel! Secondly i don't know if any of you follow DC Rainmaker but he has a tool here (https://www.dcrainmaker.com/analyzer) Where you can record two power meter efforts and compare them side by side with Graphs etc I believe Golden cheater does the same thing! Might be interesting to do a series of tests at low and high output and just see what the data shows!

[Cant Climb]

Thanks for posting your findings.

[LloydyH85]

Although i had observed up to a 20 watt difference at times, After further testing i am going to say its closer to 10-15 watts most of the time!

[flatbread]

Now the rabbit hole to go down is to test your Assiomas against a properly calibrated SRM.

FWIW, pedal-based PMs ten to read a bit higher than crank-based ones (SRM, Quarq, Powe2Max) which tended to read a bit higher than my favorite, the old hub-based PowerTap (I still have three rear wheels that I will ride until they die, then catch up with crank-based).

So, the Bike Erg might be closer than you think.

[johnlvs2run]

Filippo Ganna's 4k world record is 4:01.934, which is 1582 watts.
The 4k record on the BE is Joshua Bugajski at 5:47.5, which is 535 watts.

Kevin Metcalfe recently broke the age 55 world hour record on the track at 49.121 kilometers,
which is a 4k pace of (3600 / 49.121 x 4) 4:53.1, which is 889 watts, way higher than the 4k best on a BE.

I would accept BE record watts being 80 to 90 percent of the watts for a world class cyclist.
But 34 percent?? And only 60% of the watts that an age 55 cyclist can do for an hour on the track??
I'm highly sceptical that the BE watts calculation is anywhere close to the watts produced on road and track bikes.

[Nomath]

20 Watt was also the discrepancy that S. Boyas and co-authors found between the power measured by the C2 rowing ergometer and the power calculated from sensors for handle force and handle speed. See the study Power Responses of a Rowing Ergometer : Mechanical Sensors vs Concept2 Measurement System, published in the Internation Journal of Sports Medicine (2006).
The figure below is taken from this study. The calculated discrepancy, 19.6W, is roughly independent of the applied power.
The discrepancy is likely due to losses in the chain drive : inertia of the chain, and friction and bungee losses.

[flatbread]

Filippo Ganna's 4k world record is 4:01.934, which is 1582 watts.
The 4k record on the BE is Joshua Bugajski at 5:47.5, which is 535 watts.

Kevin Metcalfe recently broke the age 55 world hour record on the track at 49.121 kilometers,
which is a 4k pace of (3600 / 49.121 x 4) 4:53.1, which is 889 watts, way higher than the 4k best on a BE.

I would accept BE record watts being 80 to 90 percent of the watts for a world class cyclist.
But 34 percent?? And only 60% of the watts that an age 55 cyclist can do for an hour on the track??
I'm highly sceptical that the BE watts calculation is anywhere close to the watts produced on road and track bikes.

Kevin Metcalfe did not hold 889 watts for an hour. No one can do that. His FTP is more like 350.

Likewise, Ganna did not do 1500w for 4 minutes. 1500w is a world-tour level 5sec sprint (for a light rider -- say 65kg). Ganna's WR was about 600w.

https://www.youtube.com/watch?v=6FfRlKuuO0A&t=384s

[Nomath]

Filippo Ganna's 4k world record is 4:01.934, which is 1582 watts.
....

Can you explain how you calculated the power delivered by Ganna ?
It seems improbably high to me for a 4 min effort.
Have a look at the figure below that was cited in a recent review on cycling power (authors F. Malizia and B. Blocken ; reference in the top of the figure).
The dot-dash line is the estimated maximum human power with optimum mechanism. Compatitive rowing records line up roughly with cycling records.