I rejoined Zwift a couple of months ago after getting a Zwift cog (14-tooth). Since then, when doing Zwift workouts - in ERG mode, at lower target power (less than 200W) and when the cadence is between 80 - 90 RPM - the cadence randomly jumps between 4 to 10 RPM very briefly, usually for less than a second but occasionally for a few seconds, before dropping down. The Tacx NEO 2 is connected via Bluetooth to Zwift running on a MacBook Pro. Initially I thought perhaps Zwift was the culprit so I also paired the Tacx NEO 2 via ANT+ to my Edge 840. I was able to confirm the cadence jumps occur simultaneously in both Zwift and the Edge 840. The cadence jumps do not seem to affect the power so right now it’s just more of an annoyance because whenever it happens, Zwift tells me that I am outside the target cadence.
I complained to Garmin about this issue and also mentioned the fact that the three-year old NEO 2 had been emitting a smell whenever i hit over 500W. Garmin sent me a refurbished replacement for the NEO 2 and unfortunately, it also exhibits the same cadence issue.
The NEO 2 is on the latest firmware (2.70) and Zwift race mode is enabled.
Here’s the link to a comparison between Zwift (with Tacx NEO 2 providing the power and cadence data) and Quarq DFour power meter. Zwift has the purple data line while Quarq has the blue data line. If you zoom in the cadence chart from 49 minutes till the end, you will see the jumps.
Any Tacx NEO 2 owner also experiencing this issue?
Additional info: Chainring is 36T.
My hunch - and it is just a hunch - is that this is less a fault in your trainer and more just a limitation of trainer-based cadence detection.
I don’t know the full details, but I assume any trainer can only detect cadence by trying to spot a cyclical pattern in the power being put through the hub. I.e. the trainer sees a spike in power as your right foot comes down, a bit of a dead zone at the top/bottom of the stroke, then another spike as the left foot comes down etc.
There are situations when that pattern can become a bit difficult for the trainer to detect. If you had mastered souplesse and were pedalling in circles with a perfectly uniform power output throughout the stroke, there would be no pattern for the trainer to detect. I suspect this is what is happening - at low power outputs, you focus on form and smooth pedalling, and in doing so you make the cadence signal hard for the trainer to detect.
I would also question how much it matters - unless deliberately doing cadence drills, I would mostly just trust your natural cadence, and ignore small fluctuations in the reported cadence. If you do want accurate cadence data, I think you could also just configure Zwift to take cadence from your Quarq power meter rather than your trainer.
Prior to being replaced by the refurbished unit from Garmin, I had used the NEO 2 trainer for over three years: the entire off season between 2023 - 2024 on Zwift, the entire off season between 2024 - 2025 on MyWhoosh (I had cadence drop issue and I suspect this was due to the really buggy MyWhoosh app because the drop was occurring at seemingly fixed intervals), and recently back on Zwift. The random cadence spike issue did not appear until after I updated the Tacx firmware to support Zwift virtual shifting. Because the exact same issue is happening on the replacement NEO 2 Garmin sent me, i suspect this might have something to do with the firmware. I unfortunately immediately updated the firmware on this replacement NEO 2 prior to doing a workout so I do not have a before and after firmware update comparison. The next thing I am going to try is to disable Zwift racing mode on the NEO 2 and see if that makes any difference.