Smart trainer and VO2Max workout problem

Trainerroad’s powersmoothing only affects how you see the numbers on the display (iPhone, PC or what you use), actual power and power recorded by TR is still on a 1 sec frequency.

I belive the Direto however smooths the powerdata sent to Trainerroad, so it looks artificially flat. If you can, turn of the smooting on the Direto (might be done through their own app), and only use ie. 3 sec smoothing in TR app.

This is also a good tip, and something I do.

1 Like

Same workout last week - Elite Drivo 1 - same problem. It’s the nature of the beast.

Tried various different ways to ‘hit’ the target power on those short intervals.

Eventually found that letting it drift up to the correct resistance and for the final 5 sec pushing a fair bit harder (as by the time the trainer corrected itself the interval is over) things evened out nicely. While it does over run the start of the rest interval, it also means that it is correcting itself in the low power valleys where (AFAIK) it has little impact on the workout total.

Speaking with a guy at Elite, he told me that the hardware need around three seconds to change to an arbitrary high resistance, but they have smooth this value to 5 seconds because from their test they found would be better.
He told me that in the future they are planning to permit to change this value from their application, but he don’t know if with the actual trainer also or only with the new that will come, in other word for the moment we can forgot this way.

P.S. How can I share here my workout?

Blockquote On my Kickr the resistance kicks in 1 second early,

My Kickr starts to ramp up at 3 seconds prior to the interval start, I always thought this was by design.

The trainer only does what it is told (by TR’s software).

What’s the overall actual IF for your workout compared to target? On workouts like Gendarme with lots of intervals, my Actual IF is usually about 1% lower than target due to these delays in hitting resistance.

In the big scheme of things, this does not matter. A missed workout for example, has a bigger impact I’m sure than 1% of IF.

1 Like

Yesterday I also did Taylor -1 using a smart trainer.
Target for me was 288 and i hit -5 to -7 on all the 42 (yes 42!!!) intervals.
I was happy with that.
One trick is to up your cadence about 5 seconds before the interval starts and keep on right through the end beep.

I really would not be bothered about not hitting a few watts on a 30 second or less session


This is my workout:
In the next few workout I will try to:

  • put to 1Sec both TrainerRoad and Direto smoothing
  • use a bigger gear to help the magnet of the resistance with the inertia.
    I will update my result.

You want to use a smaller, or easier gear (small ring in the front, large cog in the back) to decrease the time it takes the trainer to apply resistance. You may have meant a bigger cog in the back - in which case you’re on the right track


Mine does as well - 3 seconds before. I have had two Kickrs (2014 and now a 2018) and they both do the same so I think it is intentional.

I’m using a Neo and I notice a delay in the power getting up to target.

However, it feels to me like the resistance changes very quickly (and starts changing before the interval start is due - a deliberate TrainerRoad behaviour to work around trainer lag). I believe what a lot of people are ascribing to a delay in the power is actually an artifact of power smoothing. This smoothing will give the impression of lagging power.


You shared a link to the ride, but your account is “Private” and we can not access the ride.

now it’s public.

Yes bigger cog with smaller ring, I use semicompact 36/52 and 11/32 rear, the challenge here is to use the bigger cog that permit the right output from the trainer.

Looking at this, you may be overestimating the problem. If you grab one of the sets of intervals, you’re only missing the required intensity factor by 1% (see screenshot)

While it takes a second or 2 to get up to power in the work intervals, it can also take a second or 2 to get down to the right power for the recovery intervals. So these can cancel each other out.


For the interval you are correct, but in a VO2Max aren’t any of the “microinterval” more importat?
In this scenario the power error came from 5% to 7%, at the end of the workout have you do the right amount of work?.

Ok I see your point, thanks.

1 Like

I’m just getting to my first VO2max workouts on my new smart trainer (Kickr Core). On high power intervals, at least for me, I’m finding the trainer behaves better if I slow my cadence just a few rpm before the resistance hits then hit the power but not until I feel the resistance go up. The key is not anticipating the power and waiting to feel the hit before accelerating. If you start spinning too soon, it takes a decent chunk of a short interval to catch up. I find the “establish your cadence early” tip works great on longer FTP or lower intervals but not the short high power stuff.

Also, things may not be as bad as they seem. Remembering back to my dumb trainer days (a couple months ago . . .) it still took a second or two to shift and get up to power unless I was cheating and starting early and/or going way over the target for the first second or two (and if you are use to cheating that way, it will screw up your smart trainer intervals - see above). And, I am doing exactly the same workouts I did last year and in comparing them, overall I am generally putting in more KJs with the smart trainer than what I did on the same workout on a dumb trainer. There is always the odd interval where you don’t nail the target, dumb or smart, but over an entire workout, the smart trainer usually wins.


This was my experience on the dumb trainer as well, and then I got all kinds of “free” TSS/IF in the recovery valleys because I didn’t want to shift the big ring or the cassette four times every interval, so I just slowed cadence and shifted once or twice to recover. That reduction in cadence in the recovery valleys might be why Taylor felt easier than it should have, too.

I noticed when using Ant+ in TR my Kickr also starts ramping 3 seconds before the interval but when using Bluetooth in TR the Kickr starts ramping at the interval start (or within 1 second).

One issue is that the numbers in the interval list do not match the numbers in the right panel when you highlight an interval. It’s a known TR issue which they are apparently looking into. I’ve picked one of the intervals at random so you can see what I mean. In this one the numbers in the panel are pretty close (341W/345W) but in the list they are some way off (336W/352W). I don’t know however which are correct.