Support for Tacx older smart trainers

(Kjetil Kirkholt) #181

The delay is most notiable when you ride in a race, drafting is very difficult. My T1932 head unit is from December 2009 and sn starts with 3212. But based on your experience i will borrow a head unit to test your theory.

(Robert Klinkhammer[TriFury]) #182

James Budge:   I have the T1932 with a serial number that starts with 3210…

Håkan Samuelsson:   I was looking at a python3 port of anitifier as well.  I have the test program running as python3.  However I had issues installing the python3 version of one of the modules, at least on macOS.   However, in my opinion, I think teasing apart the centralized polling loop in to separate threads is more important for performance and reliability.

Kjetil Kirkholt:   As for a possible cause for the delay, there may be an issue in how antifier handles lost or delayed messages.  Rather than using the previously stored values, it sets these missing values to 0.  This could cause what appears to be a skip or lag.  


(John Taylor) #183

Thanks all- I’m happy how it works here with my setup. I’d be more than happy to add others to the github project to develop and refine it further for other trainers and OS’s. I can’t do Fortius on a Mac because a. I don’t have a Mac and b. I don’t have a Fortius ! :slight_smile: I’m not really anything more than a hobby coder and find I need to test extensively to get things working!

Btw imagic = iflow :slight_smile:

(paha SBG [ZHU]) #184

Regarding to the serial numbers.  This is my head unit.  Is it the PD (which starts with 12) or the SN?

(paha SBG [ZHU]) #185

Robert Klinkhammer (MRC): could you help us creating the power curve?  You’d help so many people here.  Thanks.

(Håkan Samuelsson) #186

I had a look at mine and again, it is some time back I had the issue with TTS, but, if I remember right that issue was related to the serial number starting with 3212… I now have a 3211… and no issues that I have noticed.

It was just a thought and may have nothing to do with the lag issue here, but if Kjetil is able to find a different head unit with a 3211 or maybe even a 3210 serial we can at least tick that off the list for lag. :wink:

(Håkan Samuelsson) #187

I agree, I don’t think python3 is an issue as long as people running in to the 

[shell]$ sudo python
File “”, line 454
if debug:print “get ant stick”
SyntaxError: Missing parentheses in call to ‘print’

issue knows it is caused by running it under python3 and should be possible to solve by running it under python2 similar to the solution in the following post on github:

Python3 not supported post


(Kjetil Kirkholt) #188

Hi All.
Just want to share what im seeing when riding and would like feedback if any of you have seen the same.

In addition to the delay i have written about i experience that running the 64 bit antifier.exe the slope gets fixed to a value after 5 seconds and dont change. When running the it works better for a while, but experience that slope will lock to a value for a while and stop following Zwift slope, and then after a while start following Zwift again. This happened to me today 4 times on my 50min ride. I look at antifier while i ride to follow the updates and when it locks up, antifier dont receive any slope updates from zwift even if slope change in zwift.

It would be very nice to get feedback if you have or have not seen this type of behaviour.

(Bernard Walther) #189

Good news Fortius users, I planned to go to a friend who has a power meter to create the power curve. Normally this evening or tomorrow.

I will send it to John as soon I have it.


(John Taylor) #190

First power meter generated Fortius curve submitted

Download and replace any that you have so far and select Fortius as your power curve - let us know how it feels :slight_smile:

(Fabi An) #191


finally my second ant stick arrived today but unfortunately i was not capable of running antifier. When i select the Fortius power curve and locate the hardware, everything is fine. (Trainer detected, Power curve Fortius, Using suunto dongle and power factor = 1). After clicking “start runoff” i get the command of “cycle above 40kph then stop”. But with this resistance it is not possible for me. (at around 35 the tyre starts slipping and the resistance is really high) If i start cycling immediately and then click start runoff, i get a rolldown time (around 4.5s) and the green led on the T1932 starts blinking but after starting the script, i get the error that it would “need 14 levels of resistance for power curve”. Zwift does not find the trainer either.

Could this be a problem with the brake of the fortius? Using the Tacx Software, i could ride normally and the erg mode was working too, but i could not calibrate the brake because of an error. Being not a programming expert i am quite helpless. Has someone had the same problem oder can help me please?

(Bernard Walther) #192

I tried the power cuve I sent to John this evening, and I get the same error message with the “need 14 levels…”, you need to use the old power curve file, this one doesn’t work.

After some modifications in the file, the file is clearly wrong. it’s impossible to use it

(Fabi An) #193

Hi Bernard, thanks for your fast answer. I tried the old one again and it did work. But the resistance to speed up to 40kph for the runoff test is incredible high. How can I change that?

I also tried the power.curve.exe but it didnt work. As you have already created this file perhaps you can help me. Do i have to start the power.curve.exe without the antifier or the antifier first and then the power.curve? After doing the runoff test in the power.curve (how do i know here when i reach the 40kph?) the calibration does not work. After sending the calibration request and a countdown from 10 to 0 I get the error “No calibration data received - try again”. With my Garmin calibration with the Quarq works without problems.

(Bernard Walther) #194

Fabi, you must  just run the Powercurve.exe and then follow the different steps, but do you have a powermeter ? The Powercurve.exe is an application to calibrate the trainer with the real power.

 If the resitance is to high you have certainly another problem somewhere else


(Fabi An) #195

So i just tested it again. There are still several problems but at least some features did work. I finally tried to create a new powercurve and verified it by checking the values from the powermeter on the Garmin. So I tried to hold steadily 180W on the powermeter. And this is what the Tacx was measuring:
Resistance Level 3: --> ~160W
Resistance Level 4: --> 180W (quite the same with the powermeter)
Resistance Level 5: --> 240W (Tacx shows much higher value)
Resistance Level 6: --> 100W

Can someone explain to me how i can manually adjust the power curve values? Then i would try to adjust the values and controll with the powermeter till they correspond approximately.

At least the Tacx is variying the resistance according to the slope in Zwift. I hope that we will solve the other problems.

(John Taylor) #196

all in readme.txt

(Christian Bothén) #197

I’ve been riding a couple of rides with antifier and it works great. I am however experiencing problems with lagging speed, my problem is probably more related to Zwift than antifier though. If I start pedaling at 20 km/h (both Zwift and antifier showing the same value) and then increase my power output until the speed showed in antifier reaches 45 km/h the power in Zwift updates pretty fast and stabilizes after maybe 3-4 seconds. The Zwift speed however takes more than 10 seconds to stabilize at 45 km/h. This is when riding at a 0% grade in Zwift but I have the same problem at other grades too. I have my weight set to 83 kg and I use pretty standard gear in Zwift. The trainer difficulty bar is set to 3/4. I’m on an i-Flow and use the iMagic-curve in antifier.

(Jo-) #198

@ John Taylor … You have done great work and I have my Fortius working - So with my limited knowledge of how to fuss with notecards, as you can imagine Im reluctant to fuss with something that is working … That being said, please tell me if I am doing this correctly … 

  1. go to

     … >> Copy above text … 

  1. Open Antifier . Find power_calc_factors_fortius.txt . … Open and replace with above text. 

  2. Use  Antifier Normally 

… If I could ask for one more thing, could you post a link to the last and most updated version, Im kinda losing track here . 

(David Winter) #199

Newbie to all of this here. I have just got hold of a Tacx Flow Multiplayer T1901 (from label on unit). It has a head unit with 4 buttons and a digital readout and a cable that looks like a telephone connection. Can what you have developed here get me working with Zswift ?


(Kjetil Kirkholt) #200

I have the same unit and running Zwift. In addition you will need 2 ANT+ dongles. Good luck.