Tacx Neo all the way
I have a kickr core - when in the little ring changes are quick. In the big ring they are a little bit slower. For sprint workouts in erg mode I just put it in the smallest gear I have in the front and back and the response is super quick.
My experience as well, although Iāve just learned to live with it. I was going from KK Road Machine to Direto with no other experience of erg mode, and when the TR / Direto / 4iiii combo using PowerMatch felt to me almost unrideable in intervals less than 2 minutes, neither Elite nor TR were of much help ā writing it off to my inexperience with erg. Well, that too, but not solely. I sent TR files showing the weird meanders in power with a steady cadence, and got a suggestion to turn PowerMatch off. Marginally better, and Iāve grudgingly adapted to the fact that there is NO way on my setup in intervals less than 5 min to meet or exceed target power ā because of precisely the big delay Phil mentions. Itās always negative, although easily met or exceeded in resistance mode (like the KK). The average power for the interval is apparently calculated from where the blue block begins and ends, not where you start and stop delivering power. You may be delivering the power for the duration, but you start and end later, so it looks to TR like you fell short.
If Iād seen Philās post when first wrestling with Direto back then Iād have sent it back for something else. Meanwhile, it seems to me there is an easy software fix in TR: just make how early TR sends its on / off signal to the trainer adjustable by the user for the particular trainer ā instead of the 2-sec-fits-all. Likely thereās some good reason this isnāt done, but Iām curious why it isnāt. Other than the lag, the Direto has been reliable and consistent over almost 2 years. I wish my two go-to reviewers had picked up on this bug-feature early on ā maybe distracted by all the fuss about the āprinterā noise, which for me is a non-issue.
1 Like
Thanks - sorry you did not see my previous posts - I was all over the Elite forum with it.
I am not sure adjusting the delay per trainer would fix the issue for two reasons:
- It is the overall time it takes to respond that is the problem. For me it was in two parts: a slow response and then a slow ramp.
- Adjusting the delay in TR would be great for TR users. But you would still be exactly the same problem when you were in Zwift, Sufferfest or Eliteās own programmes (or any other) because it is a trainer problem - not a software problem. Getting TR to fudge something because the trainer is screwed up does not seems to be the sensible answer (unless you are a pure TR user). I started using Zwiftās interval sessions and other structured training programmes and they were almost unmanageable. it was only getting to Zwift and noticing the 2 second advance notice that made TR a bit more manageable. But it was still, fundamentally, the trainer that was the problem.
And for the avoidance of doubt I was using ANT+ protocols and never BT.
1 Like
Not just the trainer. The actual wireless transmission protocols have limits. For example most ANT+ protocols only require new data 1 time per second. If a power is increased just after a power signal it could be almost a second until the new power is even sent the trainer. Iād like it if TR would confirm how often it transmits power targets in ANT+ and BT. Maybe they do it better than every 1s and BT is better too.
Using the windows Tacx Desktop App that you get for free with a new trainer Iāve found the Neo 2 to be too quick. So quick I wonder if it is using a private transmission protocol. The iOS app is better.
Iāve had my Direto maybe a year and noticed the same issue but Iām not doing many 30 second intervals. When it takes letās say a max of 10 seconds to adjust to the new watts for a 5 minute interval Iām not worried about that. But it has been something Iāve noticed.
1 Like
I wonder if there is something going on with some Diretoās vs others. Mine was a pre-order unit and I havenāt had belt issues and I donāt see the commonly-reported long lag. I do see a delay in resistance changes but nothing close to what some folks post ā¦ up to 10 seconds!? With TR pushing the resistance change command early, I only see a delay of ~1.5-2 seconds from when the interval should start to when my power is at or above the prescribed level. That works well enough for any interval Iāve come across.
For anyone reporting a longer lag, consider if the lag is real (resistance is late and you donāt āfeelā the increase for 10 seconds) or only perceived (the displayed power data increase is delayed). To test this out, ride into a sprint interval with your eyes closed and count the # of seconds from the interval change beep to when you feel it. If you are measuring the delay by looking at workout data, it can be significantly impacted by the Elite power smoothing setting. You can set that to the minimum (1) by using the mobile app. Your data wonāt look as smooth, but youāll get unfiltered power measurements and will see a quicker rise in power. For what its worth, I ride in the small chain ring, middle of the cassetteā¦ slower flywheel speed may also have an impact on the reaction time.
1 Like
Yup thatās my experience with the Neo as well. It takes a couple of seconds max for the resistance to change.
@eerke Data rate for Bluetooth 4 (BLE) is significantly faster than that of ANT+, ~16x faster IIRC.
@mcneese.chad Sounds like you havenāt figured out that ANT+ issue w/your Neo2. Very odd!
When using my MacBook, I use ANT+. While BLE works, it will occasionally drop out. This could be signal interference from all of the neighbors WiFi and/or poor antennae design. ANT+ dongle w/out the USB extension is problematic. However, with a USB extension cable, ANT+ is rock solid.
If running Zwift, the above applies as well. ANT+ w/USB extension is solid. Everything else, not so much. If using my iPhone, BLE works great both TR and Zwift.
Lastly, if using my AppleTV for Zwift, BLE is great! So in my case/environment, perhaps it is is a MacBook specific BLE issue.
In theory both should work fine. However, often times, one will be better than the other due to environmental conditions, hardware (e.g., laptop, trainer, dongle, etcā¦), software (i.e., OSX, Windows, etcā¦). Experiment and go with the one that works!
It isnāt about the data rate during continuous data transfer. It is about how often that data is sent. The Bluetooth FTMS specification says āā¦typically once per secondā¦ā
Iām hoping they send it more often, but who knows.
Yeah, I need to get back to testing with my ANT+ connection on PC. I suspect the difference in trainers and their transceiver location may be an issue. I plan to try with different ANT+ USB sticks (I have 5 different ones to test) and a similar number of extensions.
I may just need to place the final location of the USB stick in a different spot than works for the other trainers. Iāve been using BLE on my old phone due to convenience, and not bothered to test again like I should 
I didnāt see anyone mention this so Iāll add to this thread. I use TrainerRoad for nearly all the athletes I coach and itās common that they get back to me after the first session saying it doesnāt change quickly enough largely irrespective of trainer. On troubleshooting I usually find they are using a monster gear which makes it harder for the trainer. This makes sense for trainers with flywheels because the faster the flywheel is spinning the greater the force required to alter the inertia. I now routinely instruct athletes to use the small chainring and the 4th cog down from the top.
7 Likes