direto erg mode in zwift

Moderators: Elite Admin S, Elite Admin M

raphnou
Posts: 46
Joined: Wed Apr 11, 2018 9:23 am

Re: direto erg mode in zwift

Postby raphnou » Wed May 16, 2018 11:06 pm

Elite Admin S wrote:Hi guys

@raphnou, you're doing a training quite "extreme" (big change in power in few seconds).

Please do this two tests:

1. do the same training (with the same segments) in my e-training and let me know if the situation is the same

2. in Zwift, increase a little the segments (20-30s instead of 10s) and let me know if the situation improves

Then, let us know.
Regards



Hello Elite,

I wanted to do the same training with MyTraining and compare with Zwift the values, unfortunately MyETraining did not add the intervals in the TCX file, so it is unusable. Another problem that I discovered with MET, the arms fall to me ...

Very disappointed to go round in circles, sincerely if your software offer was up to scratch, certainly the experience would be better.
But that's too much, I abandon MYETraining definitively, I just keep it for calibration hoping that once paid, I hope we can still calibrate the Direto for free ...

So I continue on Zwift trying to keep me on a program, for the moment I use a smoothing at 5, but I find that the Direto still has difficulties to maintain the constant power for powers above 100W

For example with Cheetah, I see for a target power at 150W an average of 146W, and for a target power of 180W I have an average of 174W.
I'm going to try a freeride session tomorrow because finally I have very little ride with this mode I hope that everything will be fine ...

Could you just give us the characteristics of the Direto on its reactivity, so that we can modify the training of Zwift by taking into account this latency, because some training of Zwift have intervals of 10 seconds, thus going from 105W to 325W then keep the 325W for 10 seconds it's impossible to do with the Direto.
Moreover, if Zwift offers intervals of 10 seconds, that means that trainers are able to do it, is not it? But who can do that at Elite?

I will admit that I was discouraged and called the online store to claim my refund before the 30 days, but the shop did not want to know because I had used the Direto.

have a good day

athirayoub
Posts: 1
Joined: Thu Oct 11, 2018 4:24 pm

Re: direto erg mode in zwift

Postby athirayoub » Thu Oct 11, 2018 6:06 pm

I completed the same structured workout on my new Direto (default smoothing setting of 2) as I had done with my previous trainer (Tacx Flow Smart) 3 days before. ERG resistance and power output on the Direto was all over place compared to my Tacx Flow Smart (see image for power profile comparison), and all the over/under/random resistance punched the legs so bad that I barely made it through the entire workout (effort score also attached). It was also extremely frustrating not knowing what random resistance the trainer was going to apply next during a block of (what should've been) static power output.

90% of the indoor training I do is in ERG mode, so if I can't rely on the Direto to keep me in the power zone dictated by a structured workout, I'll return it for another Tacx... as they seem to have a better grasp on how ERG mode should work.

I'll try increasing power smoothing as suggested by some, but it seems like more of a work-around than a solution... and I've seen nothing from Elite confirming that this is something they're actively looking into addressing. C'mon Elite, you've built a great trainer with the Direto... don't let it go to waste because of (what I believe to be) some bad firmware preventing it from adjusting ERG resistance properly.

TrainerComparison.jpg
Same workout, Direto vs. Tacx Flow Smart
TrainerComparison.jpg (233.11 KiB) Viewed 804 times


Athir

User avatar
Elite Admin S
Posts: 1506
Joined: Tue Nov 08, 2016 10:18 am

Re: direto erg mode in zwift

Postby Elite Admin S » Thu Oct 18, 2018 12:32 pm

Hi Athir

About the Relative Effort value, it's something that is calculated by Garmin and I really don't know how it is calculated.
About the comparison you've posted, you have to consider that Tacx shows a calculated power, while Direto shows you a lecture of the real power you develop. So, it's absolutely normal that the values you see on the chart are apparently more stable than what you see on Direto.

Please read my answer on this topic, there are also charts and detailed explication (it's about Drivo, but it's the exact same concept):
viewtopic.php?f=35&t=78

Regards

danga
Posts: 10
Joined: Thu Jan 24, 2019 11:50 pm

Re: direto erg mode in zwift

Postby danga » Mon Feb 11, 2019 9:35 pm

Elite Admin S wrote:The power smoothin value is not a "time" option, it is the number of pedal strokes that will be averaged.
If you put it at 6, it means that the power shown will be the average power of the latest 6 pedal strokes.
The higher the value, the more averaged will be the power.


where is the value averaged? on the trainer? Or in the app?

barryscsmith
Posts: 1
Joined: Fri Mar 01, 2019 6:21 pm

Re: direto erg mode in zwift

Postby barryscsmith » Fri Mar 01, 2019 6:56 pm

Hi,

I purchased the Direto on Feb 13 and have about 20 hours on it since. I did the trainer calibration as per the instructions when I got it, and then again on Feb 24 to see if anything has 'broke in' (the sticker says 6403 and the calibration value was 6398), since I noticed that the delay from changing power intervals seems to have made the ERG mode even worse on my Direto. I've read previous posts about changing the power smoothing, and I realize that for short burst intervals (10 seconds or changing over 125watts) that the ERG mode won't be effective with the Direto due to the lag of adjusting the resistance. Today my workout involved going from 105w in the rest zone to 3mins at 175w. When it was time to do the first effort, the resistance wouldn't go over 135w. I tried keeping my cadence the same for the first 10 seconds of the interval, but it didn't go up from 135w. I tried slowing down my cadence to see if the trainer would react and increase the resistance but it didn't help, I tried speeding up my cadence to manually increase my output but the trainer wouldn't move. I tried changing gears so that the flywheel was moving faster but that didn't change anything either. This took about 1 minute and I couldn't get over 150w. I ended up pausing the workout and turning off the ERG mode so that I could manually hit my targets.

Definitely disappointed with how the ERG mode works on the Direto, it seems like it's a common enough issue that ELITE could do a firmware update to make the ERG mode more responsive, I understand that there's a balance between being accurate and smoothing out the power, but if the current solution is to put power smoothing at 10, then even at a cadence of 120, that would take 5 seconds to respond and half of the burst interval is over.

If ELITE advertises the Direto is Zwift certified, then make sure it works properly with Zwift - having the consumers do the testing/problem solving without releasing updates to firmware is being lazy.

Will not recommend to others until this is resolved, and will also update my review where I purchased the Direto so that others can be aware of my experience.

sjester86
Posts: 2
Joined: Tue Mar 12, 2019 6:58 pm

Re: direto erg mode in zwift

Postby sjester86 » Tue Mar 12, 2019 7:15 pm

Due to really two reasons:
1) Elite seems to be ignoring / not acknowledging this issue with ERG
2) Because of 1) we're given absolutely no feedback on fixing

I am likely going to return my newly-bought Direto II and go with something from Wahoo or Tac. A friend at work as a Wahoo KICKR and the comparison between his trainer and my trainer is night and day when it comes to responsiveness and stability in ERG.

That it takes 7-8 seconds to even ramp up, and still not even be stable, shows there's clear software issues in that power control loop going on. Maybe try a linear approach based on calibration rather than have these obviously not correct power control algorithms not working?

I am guessing there's some low order PID control loop control going on in the background controlling resistance when likely you could probably build a look-up table of power/resistance target to cadence that gets created during calibration and linearly optimize it with better results.

If I am a 150W @ 78 RPM, shouldn't the resistance that the magnetic arms should be set to be known and always known (guessing based on some distance value vs resistance, which is also an already known value)? Why are you trying to do complex power control loops without using some baseline value that the trainer should already know?


Return to “Direto”

Who is online

Users browsing this forum: No registered users and 2 guests