Equipment calibration - Tacx I-magic and Fortius

According to the TR website and the TR app, my Tacx I-Magic trainer is supported. This is a dumb trainer and is wired to a PC running Fortius software. It works well, except that the power displayed by Fortius and the power displayed by the TR app are different. I know some difference is expected, but I would expect (maybe wrongly) that the difference would be roughly constant. So, if Fortius shows x% over TR at low power, that percentage to be similar at higher powers. This is far from what happens in my setup and I want to think this is due to the settings.

Quoting rough figures, at low power, 108W on TR, Fortius shows 120W. At 135W, both meters are around 5W from each other. By the time I’m pushing 250W on TR, Fortius is showing 230W and my last sample point is 330W on TR, which results in 300W on Fortius. So, as I increase power at a similar cadence, the deltas differ: -12W, 0W, +20, +30W.

This is very annoying because I am not even sure what impact this has on training. Am I even training at the right power for SS, VO2max, etc.?

The specifics. The TR app shows the I-Magic for virtual power and within it it has settings from “Level -4 (Easiest)” to “Level +7 (Hardest)”. I presume this relates to the slope settings in Fortius as I have to set the slope to +0.4 to be able to push enough power on the biggest gear. However, the slope setting in fortius goes from +20 to -20 in 0.2 increments and I don’t know how this relates to the 11 Levels in TR. After playing around with all the levels I have it set to “Level 1” as this gives me the closest match (which is not that close as you can see above). I really want to think I’m doing something wrong and there is a setting that will get the power from both sources much closer or with a more constant difference.

Any recommendations greatly appreciated.

Do you have an independent source of power measurement or are you using “virtual power” with a speed sensor?

For what it is worth, I have a Tacx Flow T1682 which is basically the same as the I-magic, I think. Tacx used to sell an upgrade kit which replaced the yellow head unit for my turbo with the computer-wired version, anyway.

My findings are similar to yours. I have a Stages power meter on my bike. At about 150W the Flow and TR show about the same power. Below 150W the Flow reads high, above 150W the Flow reads low.

1 Like

But that shouldn’t be the case, right? If TR supports the specific trainer is because it has the power curve for that trainer so both TR and Tacx should be following quite closely one another. At least this is what I am hoping and that by getting the right settings in both I can get to something more reasonable.

Well, TR have a power curve for that trainer because a few years ago someone sat with a PowerTap (or whatever) and measured power at a certain combination of resistance/ speed to produce a curve.

That doesn’t have to match what Tacx thinks the power is, and displays on the head unit/software.

That’s interesting at it means that the power TR shows would be more accurate than what Fortius shows?

Do you know how to reconcile the 11 Levels in TR with any settings in Fortius? that would be very useful to know.

Well yes I’m more inclined to believe what my Stages shows me than what Tacx is displaying.

The Stages is measuring strain at the crank but Tacx are calculating based on the resistance and how fast they see the flywheel spinning, but there are many assumptions there that could affect that calculation that it’s pretty much a guess.

Afraid I can’t help map the levels to the Fortius unfortunately. My unit has controllable resistance from -4 to +9.

I believe that with virtual power they have a curve for my trainer with resistance +4, but I have had a power meter for a few years now so don’t remember exactly.

Does TR match fairly accurately the power measurements from your Stages meter? I don’t mean exact numbers but maybe a fairly constant offset? If so, would you mind sharing what Level you have selected in virtual power in the TR app and what resistance setttings you have on the trainer software?

TrainerRoad uses the data from my Stages as the source of power, so yes, they match as they are the same thing.

1 Like

Hi @Wouter_Tacx_Fortius,
I’ve had a look at your code and I understand you need Python to run it? Whilst I have a look at how I can use it I have a question, which is how this thread started. I now understand a little better how the trainer works so I can be more specific. I understand internally, Fortius has around 14 resistance levels. From a video about Antifier, I can see that when it works, and you show debug data, you can see what resistance level is being used. I would like to think there is some relation between these levels and the slope you select or some other variables. If I could understand how Fortius selects the resistance levels, that would sort my problem. At the moment I have TR running on Level 1, but I don’t think that’s the right resistance level for my setup and presume the virtual power values shown by TR are way off. Do you know how I can find out what resistance level I should chose?

Does it depend on user weight, slope %, power, other?

Hi mon269

Installing python, the required modules and activating FortiusANT is not difficult, I will check whether the wiki is clear enough. (github … FortiusANT/wiki)

I know that antfier works with 14 levels in it’s powercurve definition and I expect that that is based upon another trainer than the Fortius.

The Fortius does NOT have 14 levels, that’s for sure. The Fortius works with resistance, which is a number from 0…32000. More on resistance can be found in the wiki (github … FortiusANT/wiki/The-FortiusANT’s-powercurve).

Also, the resistance can be negative (-32000 … 0) which causes a negative torque: the trainer is actually driving the bikewheel. As if free-riding from the mountain.

The difference between “the 14 levels” and the way those levels are implemented is the reason that the powercurve does not really work for the Fortius. Since antifier’s John does not respond I’m not 100% sure about this, but this conviction made me rewrite the program and publish FortiusANT. I did not find a succes story for a well-working powercurve for the Fortius, but perhaps did not search thorough enough.

Hope this helps

Please post further questions on the github FortiusANT page, since it appears I’m not allowed to put links to github here.

Well, since today you do not need anymore; a windows executable is published