Support for Tacx older smart trainers

(paha SBG [ZHU]) #141

Ok, I see.  Basically I’d need a power meter.  I’m considering buying one on Amazon, making the test and sending it back. :slight_smile:

(Domen Rogina) #142


I have one questions about power data in power_calc_factors.txt. What power is meant here? From trainer or from power-meter?


If this is a power from trainer, I thinking about following test.

If I calibrate my trainer over Tacx TTS4 software and then ride 20km/h at every grades(slopes) from -5 to 10%, I can get the avg. of Watts for each grade. Maybe this will be good alternative until someone get done the power_curve with power-meter.

For better results I can measure speed from Garmin Speed/Cadence sensor (ant+).

What do you think? Is it possible? :slight_smile:
(For trainer Fortius)

(Stephen Dalling) #143

John, the antifier app that you have developed looks amazing. I had almost given up hope of using my tacx imagic with zwift.
Not sure if I have missed this in the thread above, but will this app allow the ride in zwift to be realistic, with the resistance changing as the terrain varies?
I will need to invest in an Ant+ dongle to attach to my XP PC which has the imagic currently connected. Will I also need an Ant+ dongle attached to my Win7 laptop, which will be running Zwift?

(John Taylor) #144

>power data in power_calc_factors

This contains factors for a linear equation which expresses power as a function of wheel speed for each of the resistance levels. This has been derived from real world power data in the case of the iflow. We’re just waiting for a Fortius power curve- this should happen soon so don’t worry too much at the moment :slight_smile:

>will this app allow the ride in zwift to be realistic, with the resistance changing as the terrain varies

Yes, and also in erg mode for other trainers. Have a look here


(Stephen Dalling) #145

Thanks John. I notice that the app is compatible with the 1932 and 1942 head units. I have an older T1902 lime green head unit. Will this be compatible?

(John Taylor) #146

No idea! could you post how it goes on GitHub? I’m sure we can make it work :slight_smile:

(John Taylor) #147

I’ve uploaded a new build as I’ve found a new setting on Zwift that affects how gradients feel, and how Antifier works:

A note on gradients- Antifier will set the resistance of the trainer according to the data it receives from the app. Zwift has a setting “trainer difficulty” which alters the gradient sent from the gradient ahown on the screen. In its default setting (middle of the slider), the gradient sent to Antifier is approximately 50% of the slope on the screen. In this case a screen 5% slope shows on Antifier as a 2.5% slope. If the slider is moved all the way to the left (off), then Zwift will only send a gradient of 0% regardless of screen slope. If the slider is all the way to the right, then Zwift sends a gradient double that which is shown on the screen, i.e. a 5% slope on the Zwift screen will show in Antifier as 10%. Adjust this slider as required- about 3/4 of the way to the right will result in Zwift sending the same slope it shows on the screen. This will not affect calculated power values sent to an app from Antifier.   

BTW my trainer cave is slowly turning into a sweaty server cave so thanks for that :slight_smile:

(Domen Rogina) #148

John, I did the test with TTS4 :slight_smile:

Here are the results:

Now how to correctly convert those data to power_calc_factors.txt format? :slight_smile:

(John Taylor) #149

I don’t think there’s enough data there. Power is a product of revolutions (wheel speed) and torque(resistance). Even if we knew speed we need at least a second data point to accurately predict power at other wheel speeds. Do you have a power meter?

(Domen Rogina) #150

No, I don’t have power meter.
In this test, I calibrated trainer via Tacx Trainter Software 4. Then I ride 20km/h (data from garmin Speed/Cadence sensor) and watching Watts from TTS4 software (Watts from trainer).

Then I merge the data and did the chart above where the watts are avg. in 30sec period …


(John Taylor) #151

Thanks Domen. Ill put together a power curve based on that. We’ll just assume 0 watts at 0kph ( obviously!) It may be out at higher or lower than 20kph though as the linear relationship only starts above a certain speed and it’s impossible to predict the rate of rise. In addition the power numbers from tacx are only based on a value from the trainer. The trainer doesn’t contain a torque meter so this in itself is only a rough estimation

(paha SBG [ZHU]) #152

John, do you know someone who could get a power meter and make the curve for the fortius?  If not, I can ask in local facebook groups if someone had one for a few days.  

(Jo-) #153

John. you note a new build has been uploaded, could you please supply link to it. Im scratching and clawing my way through this, so any help for a 60something guy is appreciated … 

Also, where is that setting in ZWIFT that you note… 

Thanks as always Joe (aka Pappy) 

(John Taylor) #154

(John Taylor) #155

Build is in the github site as usual. Setting is under menu > settings :wink: it comes up if you move the mouse when riding

(Jo-) #156

…John , Im a knucklehead … So I do this … Delete old .exe…Put this ……in the folder instead … All the other files remain the same . ??  

(John Taylor) #157

Delete the whole directory and all the files in it. Download the package file for the architecture of your choice then run the self extractor. You need to update all the files

(Domen Rogina) #158

John thank you.
please upload the data so that we can test that power curve :slight_smile:

(John Taylor) #159

power_calc_factors_fortius.txt uploaded to GitHub. Download into same directory as antifier and let me know!

(Domen Rogina) #160

Thank you.

I am not at home this weekend, so I will test this on Monday. :slight_smile: