Feedback on Computrainer and Quarq

(Juan Victoria (xXx)) #1
  • (Not trying to sound to negative here… just an impression) Finally got the software to work completely after updating my video drivers; I have a feeling that this will prove to be a giant challenge (internal support and end user) in the future. General expectations from end users are that when you install software, it just works… not that you have to update a series of drivers to make the software work (that’s time and frustration and I guessing you don’t want that to be first impression of the sw…) every other workout specific software we have in our studio worked first try without updating drivers. Just something to keep in mind when this gets out into the wild… support will be a huge challenge. Now on to the good stuff…

Computrainer (power, speed) Quarq (cadence):
Software worked great… connection with the Computrainer was not an issue, graphics smooth… initial course is fun but short…

General Impressions

  • Resistance appeared to be spot on with other sw we use.
  • Speed seemed to be off. Not sure if this is general or Computrainer specific.
  • w/kg - nice feature but was a miss at how it’s being calculated as I don’t
    remember entering my weight and couldn’t find a place to enter it.
  • Course went into night time mode after 3rd lap… this would be nice if it was
    driven by local sunset times; it was mid-day when it changed
    Nice to haves
  • TSS… TSS real time calculation would be nice for those that focus on power
    for training. That’s all I look at these days…
  • Segment announcer… would be nice to know when a segment is coming up,
    maybe a double beep or something
  • Segment finishing… maybe another beep type when at 200meters and
    100meters to the end of a segment.
  • Getting into suspend mode when rolling down a hill… had to wait until I was
    on the flat… would be nice to end it whenever you want.

Quarq (power, cadence) Computrainer (speed)
Work for about 30 seconds and then the software gave up control of the computrainer head unit and was no longer capturing speed rendering it not functional.

Let me know what else I can test…

(Chris Ashley B) #2

Im on a computrainer as well… When do you think the speed is off?? I think on the flats its pretty close… for example if I was riding at 240 watts on the flats the speed seemed correct… I historically do all my workouts on spinscan mode on the new Computrainer software and it seems close to me… I know on the downhills zwift is giving us a huge push so to say… of course you need to keep pedaling or you stop… If I wanna go for the mtn jersey I typically lock into a big gear and mash it… Never really looked at the speed over the climb but like you said the resistance seems right…

(Juan Victoria (xXx)) #3

– Power seems to be off by bit but not sure of its power smoothing within the program. My quarq shows about 10 watts more on average
– How do you calibrate for Zwift on a computrainer
– Wheel stops on the downhill but he program doesn’t stop; it just slows a bit still going 30mph
– Speed on zwift if off compared to speed on the head unit of the computrainer. 80 rpm at top gear is 22mph on head unit; 24mph for same rpm and gear on Zwift.

Thoughts for improvements

  • should be able to differentiate a rider that is using a controlled resistance trainer (computrainer, wahoo, etc.) vs a general trainer with no resistance. It’s a much different ride… Specially when your doing the hill climb challenge, no resistance vs. resistance. Let me know if I’m thinking of his wrong…

(John Larson (LAPT)) #4

Here are my observations using my Computrainer:


Windows Vista 64 bit
Garmin mini style Ant+ USB adapter
Garmin HR strap
Garmin Speed/Cadence sensor
Computrainer (power only)

Everything seems to work well, except that the speed registered in the application does not match expectations set by cadence/gear and power. This behavior appears to manifest itself on negative grade parts of the course. In the small chain ring with a 39x16 gear, 80 RPM, -3.9% grade, speed registered over 40 MPH (almost like I was coasting on a downhill). Computrainers can’t really simulate coasting downhill, so I’m wondering if the software is trying to do some calculations where the displayed speed is based upon grade, weight, possible other considerations, and not disdplaying what is read from the sensors.

I will try again with the program set to use the Computrainer data for speed, instead of the Garmin speed/cadence sensor.

(Chris Ashley B) #5

Hey guys, 10 year computrainer user here… the speed the computrainer displays is not accurate to real world… The CT measures speed by counting load generator revolutions times the known circumference of the roller. Your computrainer is going be approximately 1MPH or 1.6KMH slower than real world speed… I am suspecting that there is speed emulation going on… as like you said we can’t get a free ride downhill on the computrainer… but it is fun winding it up to sprint watts on the downhill :slight_smile:
Im not if the zwift software knows this and is accounting for it or not…

(Greg Gibson-Haymarket) #6

This has been said many times in the forums already, but regarding weight - Zwift is figuring everyone at 75kg. There will come a time when you enter your own weight, but for now everyone is an equal 75kg.

(Eric C. (Zwift HQ)) #7

Just a quick note regarding drivers:

Zwift is a state-of-the-art piece of software that requires some fairly hefty computer specs. It uses a lot of modern technology to its fullest and, often times, older drivers do not support everything we’re trying to wring from the software.

It’s inevitable that, at some point, people will just have to either use a more contemporary machine (which is why we ask if you’ve purchased a computer made in the past 3 years) or that you update your hardware to its latest and greatest capabilities.

(Juan Victoria (xXx)) #8

Eric / Greg,

Thanks for the feedback… Completly understand your point about Hatdware/Drivers.

Let me know if there is anything specific you need testing in the computrainer / Powerbeam world…