Powertap: CPU numbers vary (a lot?)

Discussion in 'Power Training' started by DerJan, Mar 14, 2010.

  1. DerJan

    DerJan New Member

    Joined:
    Nov 16, 2009
    Messages:
    5
    Likes Received:
    0
    Hi,

    I'm quite new to powertraining, been using my Powertap for a week or four. The thing that has suprised me the most is how hard it is to keep the power number in a certain range. For example, when I ride a flat road, the powertap gives me numbers between 170 and 200, but there are no noticable changes in speed or wind. The display interval is set at 3 seconds. Is this normal?
     
    Tags:


  2. tonyzackery

    tonyzackery Well-Known Member

    Joined:
    Dec 23, 2006
    Messages:
    3,517
    Likes Received:
    46
    Very normal. You're not a machine, and as such, you can't produce power at the same number throughout the 360 degree pedal stroke. For the overwhelmingly majority of us mere mortals ( Jacques Anqutiel excluded - well, even he wasn't perfect), there are two very distinct power peaks in our pedal strokes - the "push down" phases.

    There is considerable debate raging on whether or not it is really necessary to attempt to pedal in circles. Is the cyclist that strives for or achieves this "perfection" more powerful on average than a "masher"? Who knows, and I really don't give a rip...but that's just me:D. ymmv...
     
  3. Alex Simmons

    Alex Simmons Member

    Joined:
    Mar 12, 2006
    Messages:
    2,471
    Likes Received:
    19
    It is normal for two reasons:

    1. Power production is by its very nature, a highly variable thing, in particular when riding on outdoor terrain, where every little variation in gradient and wind results in changes to the forces you naturally apply to the cranks.

    2. The Powertap itself, due to the time based sampling of torque data at the hub, artificially creates a little extra variability in the displayed and recorded power data from data point to data point due to variable number of whole downstrokes it records per 1sec or 1.26 second times sample (sample frequency depends on how old your PT is). This is known as aliasing. It's not a big issue as on average the data is accurate. Aliasing doesn't happen with event based sampling on meters such as SRM & Quarq which use a whole number of pedal strokes.

    Easiest thing to do is perhaps increase the display average to 5 seconds (but not the data sample rate) as that will dampen it a bit more but not remove it entirely.

    Pedaling within in a tight power range isn't really necessary but it is easier to do on indoor trainers/rollers.
     
  4. DerJan

    DerJan New Member

    Joined:
    Nov 16, 2009
    Messages:
    5
    Likes Received:
    0
    Ok guys, thanks. It's good to know that my PT 2.4 is working as it is supposed to. @Alex, it is a 2008 model, upgraded to ANT+. I'm guessing the time sample is 1.26.

    I'll be working through this forum once again then, cause now I want to find out how to train with powerzones, especially with a power number that goes up and down.
     
  5. smaryka

    smaryka Member

    Joined:
    Aug 18, 2009
    Messages:
    186
    Likes Received:
    8
    You'll get used to seeing the power numbers go up and down, after a while you'll get a feel for the steady numbers you're putting out when you're riding even though the numbers jump up and down in the moment.

    Regardless, the after-the-fact analysis of AP, NP, etc. will be a better measure of what you're capable of riding for any given time period. Then when you're out on the road, trying to maintain xyz watts, you just want to keep it in the range of xyz +/- ab% where ab relates to what you're doing (3 min intervals will have a tighter range than say 20 min intervals).

    I don't use a Powertap CPU (I use a Garmin 705 for a head unit) but I'm assuming you can get it to display lap average power? That's very handy for when you're doing steady-state intervals or a hill climb, something where you have a wattage you're aiming for. That way you can see what your average is for that interval, as well as the fluctuating instant power numbers you're putting out.
     
  6. daveryanwyoming

    daveryanwyoming Well-Known Member

    Joined:
    Oct 3, 2006
    Messages:
    3,857
    Likes Received:
    97
    A couple of thoughts:

    - Training with power is a bit different than training with heart rate if that's your background. Heart rate has a lot of built in averaging, for instance it takes roughly 7 minutes for my HR to come up to the average during a 20 minute training interval if I keep the power very steady. So it makes it a lot easier to keep HR in 'zones' as it won't instantly drop below a zone and if you get more than a couple of beats above a particular zone it means you've been working harder than desired for quite a while already and it makes sense to back off a bit.

    - [edit]Another key difference between HR and PWR based training is that in general HR 'zones' don't change much over time where changing power 'levels' is exactly the point of training. IOW, for someone with several months or more of steady riding they could do lab tests, time trials or maybe Threshold intervals at say 170 bpm HR. Keep training and there's no reason to expect that number to change much, sure you should over time go faster (put out more power) for that HR, it may or may not change due to fatigue, aging, nutrition, hydration or day to day variations but those changes aren't in general a reflection of training progress. The entire point of training is to raise your sustainable power for durations relevant to your chosen events. So that same cyclist might do lab testing or TTs or long Threshold intervals at say 220 watts after a few months of regular training but hopefully that number will climb over time with good training. So by their very nature power levels will change over time where HR zones tend not to change. So the way you view those numbers while out riding is different as you should allow for fitness improvements when power training and with it changes in target ranges.

    - Power responds to every pedal stroke and can jump up or down quite a bit as you've seen, especially if you take into account PT aliasing issues the other guys talked about. But the human body doesn't really react instantly to these second by second changes. The body responds to abrupt changes in intensity (power) relatively slowly which is why both HR and perceived exertion (RPE) take so long to respond. That basically means that for the purpose of training the aerobic or metabolic processes (up to VO2 Max work or L5 in Andy's system) brief surges or dips in power that are shorter than about 30 seconds don't really mean much by themselves. A couple of hard pedal strokes to crest a hill (not talking about a full out sprint) or coasting around a tight bend doesn't make or break an interval but low power sections longer than about 30 seconds or surges lasting much longer than that start to change things. So don't get hung up on the micro-details but keep the big picture goals in mind. Setting your PT display interval (not recording interval as Alex said) to 5 or 10 seconds can help you to avoid chasing your tail.

    - You'll see a lot of discussion on these boards and elsewhere about training 'descriptively' instead of 'prescriptively'. A lot of folks struggle with this one, especially if they're accustomed to holding very tight HR target zones when they train. The basic idea here is that you 'do what you can do' you target particular energy delivery systems mostly by choosing the duration of your intervals or rides, you use the PM to roughly check that you're not going crazy hard at the start of efforts such that you won't be able to finish them or getting complacent and taking it easier than necessary but you don't glue your eyes to the PM and back off when you're feeling great or hammer yourself into the ground when it just isn't gonna happen today. Then you go home and review your data (the 'descriptive' part) to see how you did, where you could improve on things like pacing and use that information to guide future sessions. It works in part because training isn't a one time deal. No single workout determines the success or failure of your program and the information from one session can be fed forward to future sessions to improve the quality of your workouts. The review of power data combined with tuning into your body (RPE) lets you focus on solid quality workouts even if they're a handful of watts above or below the ideal target range.

    - So with the 'descriptive' approach and the jumpy nature of displayed power data you learn to take the training level ranges a bit looser, listen to your body and not be a slave to the instantaneous power numbers on your PT display. It's more important to pay attention to how you feel for a given power level on a given workout day than it is to hit a particular number. That gives you important clues to improving fitness or perhaps excess fatigue and helps you know when to target different systems or bump up your target levels.

    Hope that helps,
    -Dave
     
  7. DerJan

    DerJan New Member

    Joined:
    Nov 16, 2009
    Messages:
    5
    Likes Received:
    0
    Smaryka, one of the challenges of powertraining with a Powertap is trying to find out exactly what it can and cannot do...the manual is helpful, but only a little bit. A few good searches on this forum or Google Wattage Groups will do the trick I hope.

    Dave, I think I'll add your reply to My Favourites...very interesting stuff. Indeed I've been HR training for about 8 years now. It has always been easier to keep my hr in the desired zones than it is now with Watts. The idea of not being too rigid with training in zones sounds good, as it prevents training from getting an activity with a mostly digital view. Cycling, after all, is the most fun when you get to do it outside.

    My next step will be the purchase of Hunter/Coggans book, I'm curious how I can use my commuting-trips (is this spelled right? I'm Dutch, not too good at English) in a well constructed training plan. This book is reffered to as The Bible, so I'm confident it will guide me:cool:
     
  8. tonyzackery

    tonyzackery Well-Known Member

    Joined:
    Dec 23, 2006
    Messages:
    3,517
    Likes Received:
    46
    One other thing to smooth out your power application without changing any display on your CPU is to increase your cadence, if possible/practicable. This will lower each power peak in your pedal stroke. You'll find cycling at relatively low rpms exacerbates the power peaks considerably...
     
  9. bwbike

    bwbike New Member

    Joined:
    Oct 26, 2009
    Messages:
    26
    Likes Received:
    0
    "My next step will be the purchase of Hunter/Coggans book"

    Yes, Training and Racing with a Power Meter is an outstanding learning & reference book. Everything you need to know to get started using a power meter is in this book.
     
  10. Alex Simmons

    Alex Simmons Member

    Joined:
    Mar 12, 2006
    Messages:
    2,471
    Likes Received:
    19
    Whether or not that smooths out actual power delivery is a moot point as:
    - it may actually make the PT display jumpier (or smoother) simply because of the impact the cadence has on the aliasing. You'd need to measure power with a crank based meter to actually know.
    - metabolically, it just isn't all that important anyway
     
  11. tonyzackery

    tonyzackery Well-Known Member

    Joined:
    Dec 23, 2006
    Messages:
    3,517
    Likes Received:
    46
    ^^Stop trying to argue with yourself, Alex:rolleyes:...
     
  12. Alex Simmons

    Alex Simmons Member

    Joined:
    Mar 12, 2006
    Messages:
    2,471
    Likes Received:
    19
    ?
     
  13. Bike N Ski

    Bike N Ski New Member

    Joined:
    May 4, 2005
    Messages:
    107
    Likes Received:
    0
    Very shortly it will be book(s).

    Training and Racing with a Power Meter 2nd Edition
     
Loading...
Loading...