Tacx Neo 2T with fluctuating power numbers

A light bike doesn't replace good fitness.

Moderator: Moderator Team

Post Reply
User avatar
Sjoerd
Posts: 321
Joined: Fri Jan 15, 2010 7:22 pm
Location: Les Pays Bas

by Sjoerd

We have two Neo's in the house: my Neo (1st generation) and my girlfriend's Neo 2T. When my girlfriend uses the 2T (free riding in Zwift) I've noticed that the power changes very often, sometimes multiple times per second, even with 3-second-average enabled. The fluctuation in power is there when the terrain is flat. My original Neo is much, much more stable and the difference is such that I am suspecting something is wrong with the 2T, but going by the information on the interweb it's not a common issue.

Anybody out there with a similar experience?

TobinHatesYou
Posts: 12443
Joined: Mon Jul 24, 2017 12:02 pm

by TobinHatesYou

Record a video of it.

by Weenie


Visit starbike.com Online Retailer for HighEnd cycling components
Great Prices ✓    Broad Selection ✓    Worldwide Delivery ✓

www.starbike.com



User avatar
mvnsnd
Posts: 78
Joined: Mon Mar 21, 2016 8:16 pm

by mvnsnd

Just a few questions:
How are you connecting to the software? Bluetooth or ANT+. Are there any other connections to the same trainer?
What hardware is connected? Laptop, phone, Apple TV, etc
Trainer difficulty setting in Zwift?
Have you tried any other trainer software or head unit, and does that behave the same way?

User avatar
ms6073
Posts: 4288
Joined: Fri Jul 23, 2004 8:24 pm
Location: Houston, Texas

by ms6073

Sjoerd wrote:
Tue Nov 15, 2022 10:46 am
We have two Neo's in the house: my Neo (1st generation) and my girlfriend's Neo 2T. When my girlfriend uses the 2T (free riding in Zwift) I've noticed that the power changes very often, sometimes multiple times per second, even with 3-second-average enabled. The fluctuation in power is there when the terrain is flat. My original Neo is much, much more stable and the difference is such that I am suspecting something is wrong with the 2T, but going by the information on the interweb it's not a common issue.

Anybody out there with a similar experience?
Don't use Zwift, so is this a Zwift option? I thought smoothing was to set how Zwift interpreted power output from the Neo to help smooth the resistance the trainer was providing, not to smooth the output diplayed on the sceen.
- Michael
"People should stop expecting normal from me... seriously, we all know it's never going to happen"

User avatar
Sjoerd
Posts: 321
Joined: Fri Jan 15, 2010 7:22 pm
Location: Les Pays Bas

by Sjoerd

Thnx for the replies so far. Having a very busy week but I might record a video to illustrate my point. I thought I did some thorough testing, but your replies made me realise that I might have left some options open.

I am using Zwift on a desktop pc, connected by BlueTooth. ANT+ gives the same results. Both my Neo as the 2T use this same setup, but of course never at the same time. I use a Wahoo Elemnt that is connected to the trainer, my girlfriend doesn't, but I have tried her 2T with my headunit connected and it doesn't make any difference.

User avatar
ms6073
Posts: 4288
Joined: Fri Jul 23, 2004 8:24 pm
Location: Houston, Texas

by ms6073

Sjoerd wrote:
Wed Nov 16, 2022 3:58 pm
I use a Wahoo Elemnt that is connected to the trainer, my girlfriend doesn't, but I have tried her 2T with my headunit connected and it doesn't make any difference.
When my girlfriend uses the 2T (free riding in Zwift) I've noticed that the power changes very often, sometimes multiple times per second, even with 3-second-average enabled.
I thought I read somewhere once that the Neo series of trainers calculate power data well over 1000 times per-second, but regardless of the power source, monitoring real time power data is obviously going to be extremely stochastic and many computer head units will lag. Also, I thought that Zwift, like Rouvy did not offer an option to display 'smoothed' power data. So where exactly are you seeing instantaneous power data displayed when you expect smoothed data/averaged data? Is it on the screen on Zwift or the cycling computer that is connected to the trainer?
- Michael
"People should stop expecting normal from me... seriously, we all know it's never going to happen"

Omiar
Posts: 385
Joined: Wed Dec 14, 2011 8:20 pm

by Omiar

How much does it fluctuate? Is rpm steady?

+/- 15W is totally fine, can even be a bit more and also in ERG mode. Over 30W is too much I would say.
NEO needs to have good rpm signal, usually problematic with short cranks and/or MTB cranks (too wide).

Rider input (torque) is not constant. NEO firmware doesn't do power smoothing. As there is no physical flywheel, there is also nothing to smooth the power output physically.
3s average power should match closely with any decent pedal/crank powermeter.
Cannondale SystemSix R8170
Trek Checkpoint SL5 MY2022

User avatar
Sjoerd
Posts: 321
Joined: Fri Jan 15, 2010 7:22 pm
Location: Les Pays Bas

by Sjoerd

I think the fluctuation is about 15 watts with steady RPM. I have no other power measuring equipment for comparison, only the two Neo's.

There's just a significant difference between the Neo and the Neo 2T and that's what got me thinking there might be something off. On the Neo original I get a change in power numbers around every second in Zwift. My guess is I am seeing the 3-second-average every second. The 3-second-average is a Zwift option btw.

The 2T is much more erratic, sometimes it changes (albeit small changes) about four times a second and I don't feel like I am putting in less or more power. Maybe the 2T just sends data more frequently to Zwift than an orignal Neo? I'll try and visit the local Mantel store this weekend, they have a test setup there and I can test their version for comparison.

@ms6073: I am talking about the numbers Zwift displays on screen.

Omiar
Posts: 385
Joined: Wed Dec 14, 2011 8:20 pm

by Omiar

Sounds normal for NEO2.
Cannondale SystemSix R8170
Trek Checkpoint SL5 MY2022

User avatar
ms6073
Posts: 4288
Joined: Fri Jul 23, 2004 8:24 pm
Location: Houston, Texas

by ms6073

Sjoerd wrote:
Thu Nov 17, 2022 2:51 pm
The 2T is much more erratic, sometimes it changes (albeit small changes) about four times a second and I don't feel like I am putting in less or more power.

@ms6073: I am talking about the numbers Zwift displays on screen.
One of the principal changes from the orginal Neo and the Neo 2T was the addition of stronger magnets (see quote below from DCRainmaker's review which also helps improve on the trainers ability to more radpidly adjust resistance in reaction to changes in the torque applied to the pedals. As others have said, its an anoyance but one that helps make the trainer more accurate but they have also keyed on an issue with shorter crank arms (< 170 mm) that effects the accuracy of how the cadence is measured and I would think might possibly contribute to a somewhat more eratic power output.
Now – as for the changes, like they teach you in grade school, it’s what’s inside that matters most. And in this case what matters most is magnets. The company has made a pile of changes internally around the magnets. It’s magnet mania. Here’s what’s different inside at a geeky level:

– Stronger Magnets: This gives it the higher torque that eliminates the slip for the vast majority of riders (including myself)
– Thicker Wiring: This reduces the heat output and improves efficiency
– Changing Wiring Topology: This reduces vibrations but also reduces the sensitivity of manufacturing tolerances
– Skewing of Magnets: This reduces vibrations and sounds
– New Magnet Holder: This helps in manufacturing by increasing position accuracy, but also reduces vibrations
- Michael
"People should stop expecting normal from me... seriously, we all know it's never going to happen"

TobinHatesYou
Posts: 12443
Joined: Mon Jul 24, 2017 12:02 pm

by TobinHatesYou

Smart trainers do not use cadence to calculate power. It’s all based on flywheel acceleration, speed and braking power (resistance.)

I would also bet that Zwift, like any other recording device/app, does not record at greater than 1-second intervals. The power meter or trainer may report at greater than 1Hz, but if you look at any FIT file, data is only collected every 1 second.

User avatar
ms6073
Posts: 4288
Joined: Fri Jul 23, 2004 8:24 pm
Location: Houston, Texas

by ms6073

TobinHatesYou wrote:
Fri Nov 18, 2022 12:35 am
The power meter or trainer may report at greater than 1Hz, but if you look at any FIT file, data is only collected every 1 second.
I think we have Garmin to thank for that as the companies head units default to smart recording and for many devices, the other option is 1-second recording. Garmin Edge series of computers receive and display data at a faster rate but still limited to 1-second interval in the FIT file.
- Michael
"People should stop expecting normal from me... seriously, we all know it's never going to happen"

by Weenie


Visit starbike.com Online Retailer for HighEnd cycling components
Great Prices ✓    Broad Selection ✓    Worldwide Delivery ✓

www.starbike.com



Post Reply