Iāve been using a Stages SB20 trainer at home for over a year. I love it, jumping in the saddle three or four times a week to tackle a TrainerRoad workout. The idea behind TrainerRoad is that you aim to hit a specific power output for each āintervalā within a ride, whilst keeping your feet spinning at an optimal cadence of between 85 and 95 revolutions per minute. For a while, Iāve been noticing that towards the end of a ride I may end up in a higher gear than I started, despite being asked to hit the same measured power output as an interval at the start. I donāt understand whatās going on.
Today I tackled Low Saddle, a 1h45m ride that includes a single 1h33m āintervalā at a steady power output. My completed ride is shown in the diagram below. The top of the blue shape represents the power target that TrainerRoad wanted me to hit, the yellow line is my measured power output and the red line is my heart rate.
My completed Low Saddle TrainerRoad ride
Once I finish a ride it gets automatically uploaded to Strava, which provides further analysis. A ride like Low Saddle is perfect for illustrating the issue. In the charts below you can see that my power output stays roughly the same throughout the ride whereas my āspeedā drifts from 25km/h to 29km/h and my cadence from 95rpm to 105rpm. This is a massive difference that I can really feel over the course of a ride. Iām glad it isnāt my mind playing tricks on me.
Power, Heart Rate, Speed and Cadence from the Strava analysis of my ride
The SB20 trainer has a power meter on each crank. Iām not sure whether the problem is that:
- The measured power output from the power meters falls off during a ride, forcing me to push harder to keep the same measured power output, or
- The resistance on the trainer eases up over the course of the ride, meaning that I need to turn the cranks faster to maintain the same power output.
Does anyone know if there is something I can do to find out?
Update 12 February 2026
Reader Mike emailed me to say that he had the same issue. Stages went out of business but have now been acquired by Giant, so he emailed their customer support team. This was their response, which Mike agreed to let me share here:
Iād be happy to begin the troubleshooting process now to see if we can resolve your issue. This is a classic case of “Resistance Drift,” a phenomenon specific to electromagnetic brake systems under sustained thermal load. The SB20 uses a large electromagnetic coil to create resistance against the flywheel. As copper gets hotter, its electrical resistance increases. This makes it harder for the Lower PCB to push the same amount of current through the coil. If the current drops due to heat, the magnetic field weakens. A weaker magnetic field means less resistance at the flywheel, which is why your cadence “runs away” from you (spinning 10+ RPM higher) even though you haven’t changed your effort or gear.
1) Check your bike’s firmware is on the latest version.Ā Update firmware on your Stages SB20 smart bike
2) Ā Increase Airflow. Try to have a high-powered fan aimed not just at yourself, but at the front-right side of the bike shroud. There are vents near the flywheel; getting cool air into those vents helps the coil dissipate heat and maintains a more consistent magnetic field.
3 If the drift is becoming a hurdle for your training, ERG mode is the technical solution. In ERG, the bike’s “brain” constantly monitors your power output. If the resistance starts to “ease up” due to heat, the bike will automatically increase the current to the magnets to keep you exactly at your target wattage, regardless of your cadence. See.Ā Recommended Settings – Zwift
4) It problem continues, then theĀ Lower PCBĀ most likely needs to be replaced.
Please let me know if any of the above steps helped solve the problem, or if you would like a one time 30% discount code towards the purchase of a lower PCB board.
Thanks so much for this, Mike. It makes a lot of sense. I might investigate activating ERG mode, but Iām a little nervous about how that would feel versus using the gears as I do now.
@adoran2 That does sound exhausting. That’s a pretty good cadence increase just to stay at the same power output. All I can think to try would be some experiments at different outputs to see if the drift is consistent — which involves a lot of work and time plus uncertainty!
@alans The heart rate increase shows that itās definitely harder work for me by the end of the ride, which is no bad thing I guess? But that harder work might just be a higher cadence. The trouble with trying different power outputs is that it still wonāt prove what the cause is.
@adoran2 Right, it’s not necessarily bad, but it’s definitely a different kind of ride than ‘just’ endurance. I was thinking with the power output thing that maybe it could help isolate the problem or give you more data that you could take to tech support, which sort of feels like the ultimate destination for this.
@alans Itās in that weird spot where itās not enough of a problem to pursue it too much. Maybe I should just toughen up and live with it š
I have got the same problem the left hand power seems show less power in time .they sent me another pedal(power meter) and still no change.i just pulled the battery out of the left power meter which keeps it stable
Heart rate creep can be fatigue/heat stress of your body. Might be that the power meter/trainer is also being affected by warm up. A pedal based power meters to have a secondary recording would be the best way to check if you’re able to get access to some.
Thanks Nick. I havenāt looked recently but I think power meters are expensive, so it would be a big investment to add those to the setup. Iāve just been continuing to live with it for now, expecting to be in a bigger gear by the end of a 90 minute ride, say, to keep the same cadence and power output as at the start.