Erg mode slow to react - Tacx Neo, Ebbetts Workout

@justmakeitwork I have a Tacx NEO as well. The issue you raise is true for 100% of rides where there is a substantial burst and the burst is short. I learned this when I was in SSB MV1 and decided to do Disaster-1 that included 10x 10sec 200% bursts and then later in full Disaster (that included the same sequence) as part of the TR user community workout day, and it’s been true in every workout with similar bursts ever since.

The TrainerRoad software seems to attempt to correct for this by starting the ramp a few seconds before the actual burst. And as one user complained, the software lets off the gas a few seconds before it supposed to further limiting the user benefit of the burst.

This issue was particularly “annoying” when I decided that the Ramp Test wasn’t a good FTP test for me and did my first 20min FTP test using the TR workout. As I know that you need a real ramp to get up to “FTP speed”, I imported the TR 20min FTP test into Workout Creator and added a 15sec ramp. This was before I understood TR’s software issue. What a mess of the workout it created! [I’ll skip the details].

So the workaround I discovered that works great was to create/re-create the workout in its entirety in Workout Creator (fabulous tool, not difficult to use). In the case of the 20min FTP test, I decided to implement Dr. Andrew Coggan’s FTP test exactly as prescribed (differs from what Coach Chad put together), including the ramp I described above and the in-test ramps prescribed by Dr. Coggan. It worked perfectly! The same approach works for shorter bursts as well. If quality bursts are important, you can recreate the burst in WC with a ramp (e.g. 10sec ramp of 40% to 150%) and then 12sec at 150% if that is the workout.

Tomorrow I am leading a ride within the S.F. Bay Area hosted by one of the TR developers, @daniel and plan to talk to him about this issue. I don’t know whether it is unique to Tacx NEO, but I seriously doubt it because one of the key features of the NEO design is that it uses electro magnets operating at 1000Hz - it is how they can uniquely simulate riding over any type of surface (bricks, wood, gravel, etc) in Zwift (Shane Miller has a video on this). I will let you know if I find out anything worth reporting.

@mcneese.chad I have made all of my workouts public so you can see additional examples of what @justmakeitwork encountered as well as the file for the Dr. Coggan FTP test to see a user workaround.

Important to keep in mind that those using PowerMatch with their PMs can overcome this issue. While the trainer may lag in its ability to adjust for the 5sec burst, the power you hit for those bursts is real and recorded.

If you’re relying on the PM from the trainer itself, then I can imagine this to be annoying. :slightly_frowning_face:


I can only concur - same issue here, also on Tacx Neo.

I don’t really think that manual edition and recreation these workouts to add a ramp up is the way to go. I feel that TR should have a smarter algorithm to handle that.

Thank you,

Can someone please share a direct link to a related ride?

I’d like to review a specific case, and compare to my Kickr results.


I’m not sure if it helps but this is my Ebbetts on a Tacx Vortex smart trainer and had the same issue

Thanks @GeorgeAnderson. I am guessing this is heavily smoothed data. I say this because the power data looks excessively smoothed, via some app setting, maybe on the Tacx app, like Wahoo’s option?

If so, it’s not a useful comparison here and looks too “perfect”. Here is a shot of one interval.

For comparison, here is one from my workout, Kickr17, ERG mode smoothing off, zoomed in to the first sprint of the 2nd set.

  1. The blue block is the normal power target we usually see.
  2. The green line the actual power target setting sent to the trainer.
  3. The yellow line is my power as reported from the Kickr.
  4. The white line is my cadence.
  • Note that the green line is an exact copy in power and duration of the blue block, but advanced on the timeline by 1 second.
  • This is done to get the actual power data (yellow) to match up to the target (blue) for workouts because all trainer have.some adjustment delay.
  • If a trainer takes a time to adjust up to a power target, it takes that same time to reset back. That is the reason TR effects time shifts the power target instructions (green) ahead of the target blocks (blue).

My method:

  • I did most of the workout in my 34t small ring and middle of an 11-26 cassette (stock with Kickr).
  • I did standing sprints in the 50t big ring, around 65-75 rpm cadence, effectively treating these like short breakaway sprints.
  • The power data (yellow) is bouncy and not pretty. It took me a bit to figure out the timing.
    But I went to the big ring about 5-10 seconds before the sprint.
  • I hit the pedals hard about 1-2 seconds before the timer hit the official start. And I powered through the interval until I felt the resistance drop.
  • I eased up and stood for 5-10 seconds after the sprint ended, then sat down and went track to normal cadence in the small ring.
  • If you look at the interval data for the sprints, I “missed” most of them and did not match the target.
  • If you look at the graph, you see lots of time around the sprint target, but its not perfectly aligned with the blue target. That alignment (or lack there of) is what created the data for the interval summary.
  • that’s important to remember, because.i sti got great power for most of the intervals, but they just aren’t matched and pretty to the second.
  • we are looking at a VERY short 5 second effort here. It’s not likely that any trainer can hit this perfect.
  • What matters is how we do in applying power for the right time and amount, even if it’s not pretty
  • I got all the intended stress and benefits of the power drill.

I’d like to see one of these with the Neo to see the zoomed in details.


Its a tacx vortex. 5sec smoothing via the TR iPhone app

1 Like

I updated some of my info above.

The 5 second smoothing in TR is supposed to only impact the power data numbers on screen during the workout.

It should not impact the data captured in the workout we are reviewing. I am guessing that is from the Tacx trainer somehow.

No it’s purely from the TR iPhone app

I know you have the workout from TR and your phone.

What I mean, is that there is some smoothing happening front the trainer side. Before the data even gets to your phone.

The data in your workout is not “direct” data because it is way to smoothed, likely from the trainer inherently or a possible setting within the Tacx app that might allow adjustment of that smoothing.

It may be similar to the ERG modes smoothing that you can enable or disable in the Wahoo app for Kickrs.

I don’t use the tacx app

No worries

Hi Chad! I saw the same thing happening with my Neo on today’s ride during Spanish Needle.

Check the link for my activity. Hopefully you can see it.

I also log my workouts with my Edge so I’m guessing this is not a Bluetooth issue since it’s connected via ANT+ to my Edge. Link below for the activity on Strava:

1 Like

Your TR account is private, so I can’t access the info.

You should be see it now.

1 Like

Your workout looks great to me. That has all the basic work applied and you will get all of get training benefit, IMHO.

1 Like

@mcneese.chad please take a look: Log In to TrainerRoad

as u can see, i hit the peak power numbers as per target but if u zoom in you’ll see it was very laggy during the sprints…

After reviewing nearly every sprint in your workout, well zoomed in, I see a max of 2 seconds offset from the green instruction change to the point the yellow power meets or exceeds the blue target.

Importantly, the yellow actual power line is only 1 second offset from the blue target. I just don’t see that as “very laggy”. It seems reasonably quick to me.

As I see in my intervals, you are getting the full benefit of time at or above the power target. I don’t see a problem here. We are looking at a very narrow window of time.

If you want better control, you may be best to swap to Resistance mode to have full control of the process.

1 Like

I actually noticed the same thing, in Erg mode on my Neo, the short (less than 15) seconds or so it lags too much.

Here is my last run on Ebbetts from a few weeks ago. Didn’t seem to have any issues on my Tacx Neo. To achieve the power in the short 5-10sec sprints I have to be in the big ring. What chainring are you guys using?

1 Like