any word on the ibike unit?

Discussion in 'Power Training' started by mmerchant, May 17, 2006.

  1. larrynipon

    larrynipon New Member

    Joined:
    Sep 17, 2003
    Messages:
    105
    Likes Received:
    0
    Not the problem. I just finished a 23 mile, varied ride. The units seems to suffer from "mount drift" throughout the ride. The iBike and the PT tracked perfectly together for about the first 5 minutes. I was in a significant headwind going out, and within 3 miles the readings were starting to drift apart. I recalibrated Tilt and Coast at that point, but the drift started again within 5-10 minutes.
    I checked the tilt at the end of the ride, and it had drifted over 5.5 degrees off its initial setting. The mount on my Cannondale suffers from the same issue, but I have it mounted on an itm extension instead of directly on the handlebar like the Tete. BOTH mounts appear to be rock solid and immovable. I checked the mount screw tightness when I returned, and they appeared to have held. Basically, I'm totally mystified. I'll post the comparative between the iBike and the Powertap shortly.

     


  2. Foothills Rider

    Foothills Rider New Member

    Joined:
    Sep 30, 2004
    Messages:
    23
    Likes Received:
    0
     
  3. snhrider

    snhrider New Member

    Joined:
    Sep 29, 2006
    Messages:
    2
    Likes Received:
    0
    Not familiar with the unit. But how this unit measure power ?
     
  4. larrynipon

    larrynipon New Member

    Joined:
    Sep 17, 2003
    Messages:
    105
    Likes Received:
    0
    www.ibikesports.com
    measures resistance, tilt, weight, speed, wind...internal cpu does the calculation. Your inputs are total weight of rider and bike, tire circumference, and a tilt and coast calibration
     
  5. nrhorwitz

    nrhorwitz New Member

    Joined:
    Sep 17, 2004
    Messages:
    70
    Likes Received:
    0
    I wanted to post something real quick about Ibike and SRM comparison on a local hill (A short one though). The comparison seems to be very good. Overall I'm impressed with the Ibike Unit and am planning on doing more tests.

    Check the comparison out.

    http://www.thresholdpower.com/images/ibike_SRM_comparison.gif

    Robin

     
  6. peterpen

    peterpen New Member

    Joined:
    Jul 29, 2003
    Messages:
    345
    Likes Received:
    0
    I don't think anyone doubts there are certain situations where the iBike will mimic results from a powermeter, for instance, a short hill. I'll be impressed when it can produce similar power estimations in less controlled situations - in a pack, over road surfaces of different qualities, in a crit -- in other words, real world racing and riding.
     
  7. nrhorwitz

    nrhorwitz New Member

    Joined:
    Sep 17, 2004
    Messages:
    70
    Likes Received:
    0
  8. Alex Simmons

    Alex Simmons Member

    Joined:
    Mar 12, 2006
    Messages:
    2,471
    Likes Received:
    20
    600 metres on a flat road with a tailwind in 1 minute at 400 Watts? Take the streamers off your handlebars :D

    What caused the big bump in ibike line in the middle of the effort?

    What about the whole ride?
     
  9. nrhorwitz

    nrhorwitz New Member

    Joined:
    Sep 17, 2004
    Messages:
    70
    Likes Received:
    0
    :p I'm a big guy unlike you all mountain goats

    Big bump in the middle of the effort - I suspect it was when i hit a bump and the watts went out for a second then it jumped back high when it started reading again.

    Whole ride: I'm more cautious about this one because the areas I ride from work to home can be very bumpy and there's a lot of time my watts are out (on the ibike). John has mentioned that there's a firmware upgrde to resolve that - I'll post the numbers once when we all get the firmware upgrade.

    Nevertheless, here it is:

    Average watts: 147 for SRM, 147 for Ibike but here's the whopper: 184 Norm Power for SRM and 261 Norm Power for Ibike. Anybody know why?
     
  10. rule62

    rule62 New Member

    Joined:
    Aug 16, 2004
    Messages:
    622
    Likes Received:
    0
    FYI...I ride with a PowerTap SL and have been testing out an iBike unit. My numbers would have almost exactly the same flavoring as do yours Robin, right down to begging the same question as to the higher NP average from the iBike unit. I consistently get that same higher average but just as consistently do not know why.
     
  11. RapDaddyo

    RapDaddyo Active Member

    Joined:
    May 17, 2005
    Messages:
    5,088
    Likes Received:
    41
    More variability around the average will produce this result (due to the 4th power weighting in the NP algorithm). The question is why is there more variability?
     
  12. acoggan

    acoggan Member

    Joined:
    Jul 4, 2003
    Messages:
    3,047
    Likes Received:
    9
    Ouch! That is not good, not good at all...
     
  13. acoggan

    acoggan Member

    Joined:
    Jul 4, 2003
    Messages:
    3,047
    Likes Received:
    9
    Another issue might be the frequency with which the iBike records data. Specifically, if it saves memory usage by "downsampling" from the original data set (e.g., like the PowerTap), then the 30 s rolling average can't do it's job properly.

    Nrhorwitz and rule62: if you send me your files at acoggan at earthlink dot net, I'd be happy to look at them to see if I can determine the source of the problem.
     
  14. RapDaddyo

    RapDaddyo Active Member

    Joined:
    May 17, 2005
    Messages:
    5,088
    Likes Received:
    41
    What is the iBike's sampling frequency? I'm really troubled by the variability issue (whatever causes it) because of the pretty extreme implications for variable power pacing (when applicable). The (erroneous) running NP computation would really put the shackles on a VP pacing strategy for a TT.
     
  15. ahaile

    ahaile New Member

    Joined:
    Aug 6, 2003
    Messages:
    63
    Likes Received:
    0
    A press clipping on the iBike website says 16 times per second. I don't know what kind of downsampling it does, but experientially, there's definitely some kind of time lag or averaging going on.

    From the iBike v. Powertap files I've looked at, the inflated NP usually comes from a small portion of the ride (i.e., it's not consistently variable throughout, but only under certain conditions) and from a particular behavior of the iBike and iBike software. When the iBike starts having trouble on rough roads, it tends to blank out except for a few spurious inflated blips in the middle of the trouble section. When the iBike app downloads the data, it linearly interpellates the missing data, but since it's doing that between the inflated blips, those blips mushroom from a few seconds of error to several. So if you had 4 seconds of outage, followed by one second of a reading 3 times your real power, followed by another 4 seconds of outage, the software turns that into a pyramidal 9 second interval. 1 second wouldn't screw with NP's 30 second rolling average, but 9 seconds do, and that's where error starts to accumulate.
     
  16. nrhorwitz

    nrhorwitz New Member

    Joined:
    Sep 17, 2004
    Messages:
    70
    Likes Received:
    0
    That does make a lot of sense because the roads between where I work and my home are rough and I'm seeing a lot of watts being dropped. I'll email the files to Andrew and see what happens.
     
  17. Alex Simmons

    Alex Simmons Member

    Joined:
    Mar 12, 2006
    Messages:
    2,471
    Likes Received:
    20
    Possibly rendering it ineffective for calculating TSS and of course making the Performance Manager concept a complete no go zone.... which is why I keep asking these posters to provide the whole of ride file info - the use of PMC is one of the most valuable aspects of training and racing with power, so determining if this is a once off or a chronic issue will be pretty important I'd have thought.
     
  18. ahaile

    ahaile New Member

    Joined:
    Aug 6, 2003
    Messages:
    63
    Likes Received:
    0
    Are you downloading your files through the iBike app and then into CyclingPeaks or straight into CyclingPeaks? The two handle dropouts differently. If you were going through the iBike app, try going straight into CyclingPeaks and see what your PNorm values come out as. You'll need at least build 44 of CyclingPeaks for that.
     
  19. nrhorwitz

    nrhorwitz New Member

    Joined:
    Sep 17, 2004
    Messages:
    70
    Likes Received:
    0
    Good tip. I will do that. (have the latest build)
     
  20. acoggan

    acoggan Member

    Joined:
    Jul 4, 2003
    Messages:
    3,047
    Likes Received:
    9
    Someone (Richard Wharton?) said on the iBike list (which is down at the moment) that you could identify interpolated values because they weren't integers - can you confirm? (I just finished looking at nhorowitz's files, and as you say the inflated normalized power value stems from one length spike in calculated power late in the ride. There don't seem to be any missing/interpolated values, however, with the excessively high power instead appearing to be due to the tilt sensor temporarily going crazy.)
     
Loading...
Loading...