Stages SB20 and possible (un)intentional cheating

Dear Zwifters, I have seen some detailed arguments about the accuracy of Stages SB20 smart bikes. Lots of users show very high or low readings.
A lot of powermeters or trainers have many problems.

But the important distinction is: This can happen on all of the SB20 units. What causes the problem in the first place is how wide (or narrow) a rider is pedaling. Due to the way SB20 cranks work riders leg span can change the power reported.

Some examples:


As some of us know it is possible to adjust these pedal powermeters to match the Stages SB20.
A lot of users also using SB20 without any dual recording.
For most it is their only or first powermeter.

It does not take much consideration to see where this is going. I am almost exclusively racing online and would not care about most of the obvious cheating or hardware problems but this is different due to unintentional side of it. And these units are sold in thousands.

I do not know what can be done or if anything should be done but I believe it is better to know this problem exists.


There is a thread on the trainerroad forum where someone has done alot of work on the accuracy of the stages bike and they seem to think it’s the way the crank/pedals are set up that is causing the issue with power readings.
Worth having a read of that if you haven’t already… but it pretty much matches your post.

I have an SB20 how do I pedal on the cranks to increase power readings :thinking::joy:.

that’s a new one to me… thanks for posting

The multiple crank length options cause an inconsistent reading on the power meter by the looks of the TR thread.
They also mention if you then change the pressure of how you pedal this then amplifies the issue, ie pushing out, pedalling on the toe etc…

Head down to post 83 or 84.

There is another thread on there aswell…

This is something I’ve been complaining about for a while. The SB20 cranks are junk. Something about them is exaggerating a design flaw in the pods themselves that wasn’t evident when bonded to other crankarms. I asked Keith Wakeham about this and he seems to think the sheer mass/stiffness of the crank might be contributing to non-rotational forces not being cancelled out and effectively amplified by the crank design. After all, the stiffer the crank, the higher the resolution needed when measuring ever smaller amounts of strain.

I have tried this for myself with 3 pairs of Garmin Vector 3/Rally dual-sided pedals and 1 pair of Assioma Duos. All of the pedals match up nicely when compared to other power sources (Hammer, KICKR, two DZeros, etc.) If I squeeze my knees/shins inward, I can get the cranks to report lower than my pedals. If I push my knees outboard, the cranks report ever higher numbers. When I get really torquey as my power output increases, the cranks report higher…again because under that much load there are lateral and twisting forces in the system. There also definitely is something to the bear-claw design. It effectively creates an asymmetric crankarm in some positions, and the bores at the ends for 170mm and 165mm cranks will contribute to higher twisting forces (which should be cancelled out, but aren’t.)

On top of all this, the cranks have several more issues:
With fresh batteries installed, voltage tested at 3.3V, the cranks will report higher power values. With partially discharged batteries around 2.8-2.9V, you will notice the reported power is much lower. Either some kind of compensation needs to be done via a calibration table a la temperature compensation or the Stages Gen3 hardware needs better voltage regulation.

Single-leg pedaling also has dumb behavior. I can induce 1800W instant with just my left leg for example. None of my other dual-sided PMs fail this hard.

Also L/R balance is completely wrong if the cranks are allowed to auto-zero while positioned at an off angle.

I really hope Stages fixes this with updated Gen4 cranks or something and sends them out to existing SB20 owners for free.

Edit the quote fucked up…

I set crank length to 165mm but due to being a narrow rider I set cleats fully outwards to keep my Q factor as small as possible. I would guess this would mean I would be peddling on the centre to inside of the pedal thus potentially a lower power reading. A teammate recently got an SB20 and compared them to his Assioma pedals while dual recording the fit files were pretty much identical so accuracy is very good as long as your peddling dynamics are good.

Thanks for that link. It was a good read while pedaling easy before my intervals.
There is a video right at the end of that TR post :smile:
I decided to test this while doing my intervals:

Normally my pedals track very close to the Stages. A normal result in ZP would be around 1-1,5%.
But when I pushed my legs out: bingo. A free 20-30 watts. One problem is pushing out is so hard on legs. I was cramping after a minute so could not go the full 5 minute mark.
I got a good 5% out of this so people getting 10+% are probably on wider q-factor pedals.

In the other sb20 thread over on TR GP Llama has said he will try to raise this with Stages and get feedback.

1 Like

Cat A. Here I come :slight_smile: