Home » Cycling News, Products, Training Guide

Comparing power readings from PowerCal and PowerTap

23 July 2012 37 Comments
Print Friendly

Comparison test:

…and…

 

This weekend I wrote about the first ride PowerCal (see here).

IMG_0945_thumb6

Now, I have more data about these rides.

I did 4 rides of varying distances, on two different bikes, each with different profiles, and I compared the power readings by capturing them on two separate Garmin units pictured above.

On this weekend’s post, I made a numerical comparison between the rides, by comparing the power average, normalized power and TSS calculation… This time, I’m focusing on pattern comparison.

The objective was to identify the power pattern exhibited by PowerCal and PowerTap, and watch the differences.

Here are the results, PowerCal in red, PowerTap in blue:

image

image

image

image

The first and most obvious result is that the PowerCal power reading does indeed match very well with the Power reading from the PowerTap.

A very small lag can be observed, as well as the fact that the PowerCal power reading tends to be smoother than the reading from PowerTap.

The is also the obvious fact that PowerCal tends to show a power reading, even when there is coasting (going downhill). This is more obvious on the second and third graph as in the others there weren’t any significant downhill/coasting periods.

Notice how the red line (PowerCal) is above zero, when the blue line is pegged at zero. This is a downhill section where I was coasting. Other sectors where there is a downhill but I kept pedaling do not display this pattern. One incentive for you not to stop pedaling on the downhill !!!!

The table below compares the 4 rides, as measured by the two Garmin device, for each the PowerTap numbers of the first line, then second line the PowerCal numbers.

distance (km) climbed (m) duration kph bpm Avg Power NormPower TSS
Ride 1 = 82.03 585 3:17:28    24.93    144 182 203 233
Ride 1 = 82.24 586 3:31:31    23.33    144 167 198 219
Ride 2 = 37.12 373 1:42:32    21.72    136 193 215 126
Ride 2 = 37.26 397 1:43:51    21.53    136 173 190 100
Ride 3 = 44.35 692 2:08:46    20.67    144 190 227 176
Ride 3 = 44.71 551 2:13:53    20.04    144 194 217 169
Ride 4 = 49.00 171 1:51:06    26.46    141 187 206 126
Ride 4 = 48.97 192 1:51:02    26.46    141 179 190 107

 

There is a tendency for PowerCal to be on the conservative side when compared to the actual power number from PowerTap, except for ride#3, which had several pronounced segments of downhill/coasting.

This table shows a remarkable accuracy for PowerCal, a device that derives power from heart rate, using statistical methods alone. The algorithm created by CycleOps is truly spectacular achieving very impressive numbers and showing that a very simple device can achieve impressive results !!!!

 

Did you like this? Share it:

Facebook comments :

37 Comments »

  • rider said:

    The accuracy of Powercal looks surprisingly good. Did you do any kind of personal calibration prior to use or did you just start using it straight from the box?

    • rodrigo1226 said:

      No calibration required. Initially cycleops had plans to include calibration but the final product requires no calibration… Just open the box and use it as a heart rate strap, and it will broadcast the Heart Rate and Power using ANT+.
      The secret is in the algorithm used by PowerCal, which works astoundingly well for your average Power.

      • kgg said:

        Wouldn’t it at least need to know your weight? It’s hard to conceive of how HR could be used to calculate power for a 100 lb’er and a 190 lb’er without knowing weights. From the online quickstart guide, I see no info about entering rider weight.

        • Rodrigo Loureiro said:

          I went into it as skeptic as you were. How could the device distinguish between athletes?
          The reality is that it does. In the tests I made, it came out very accurate considering the limitations. Granted it is a sample of 1 Athlete, and as more people use it, I will update my review, but from informal feedback I’ve been receiving, the device does seem to have a consistent accuracy of 5-10%, better for long intervals, worse on short, hard intervals.
          Most importantly it does provide you with a personal watts estimation that is consistent and very accurate between your workouts, so, that actual accuracy is much better if you view it as a personal estimation relative to your workouts, rather than absolute power wattage.
          Anyway, my main conclusion still holds… it is MUCH better than heart rate alone, and for the price ($100), this product is a no-brainer.

          • kgg said:

            I’ve been using it for a couple of weeks now. I’m not persuaded that it’s a useful training tool – the power level just fluctuates so much, I don’t know how I could ride by it. On longer segments the power output is sometimes close, and sometimes 20% low, compared to what Strava or the Cycleops power calculator would estimate. Today I let my wife use it. It said she put out something like 237 watts for her hour-long ride. Sorry, no way. She’s barely 100 lbs dripping wet. Half of that might be pretty close. I had high hopes, but until we see more powercal vs powertap comparisons for a variety of riders, I wouldn’t recommend it. Maybe they shouldn’t have dropped the calibration feature?

          • Rodrigo Loureiro said:

            @kgg I agree the power level fluctuates too much to be useful for training when the objective is monitoring short(er) intervals. I do find it useful when the objective is just to have an NP/IF/TSS estimate at the end of the ride, or to help me not exceed specific values during the ride.

            I do agree it would be better to have more comparisons for wider variety of riders, and the only thing I can say is that iIt has been accurate in all my measurements, and some of the riders I know that have used it as well.
            Like you, I do wish they had retained the calibration option. That way, you could use the factory defaults if you wanted to just go-and-use, but if you want a better estimation then you’d go through a calibration phase.
            Interestingly enough, it is a similar path that iBike has taken, since they had a calibration phase in their earlier devices and have dropped the calibration for their newest devices. However, I have not personally ever used any iBike device.

  • tim said:

    great review and just what I was after. I have a powertap hub and have just picked up some cheap zipp 404 tubulars. I was then going to find a powertap hub to put in these so I have PT training wheels and race wheels (cheaper in the medium term than renting a PT zipp for races). Now I’m thinking I could use the hub for racing (IM races to ensure I don’t ‘burn matches’ through peak power output so would need the immediate watts feedback and also accuracy over a long time trial) and the powercal for training where estimation would be OK – from your experience so far with both, what do you think of that approach?

    • Rodrigo Loureiro said:

      I do like that approach and in fact, it is the approach I take. Most rides, I do like to use my PT wheels, when having the power information (especially instant real-time power) during the ride is important.
      However, when the only thing I care about is having an NP average and TSS estimate, then PowerCal is perfect for this, since it allows me to use bikes where I don’t have any power meter installed.
      With the newer Garmin device firmwares, I also found it useful to monitor the TSS/NP/IF during the ride, especially when I don’t want to exceed specific values, for example, in recovery rides, or ride before an event.
      I have found PowerCal very useful for that purpose and could not do that with a simple heart rate strap.

  • tim said:

    good to know, thanks. Back to the PT hub

    • Rodrigo Loureiro said:

      Yes, for any serious power training, you have to use a direct measure power device, like PT. I have several PT wheels, and will continue to use it. I do use the PowerCal on non-training rides just to get a TSS/NP estimate

  • rider said:

    The Cycleops PowerAgent (ver.7.5) is now compatible also with Powercal. What does that mean: is it possible to somehow configure the Powercal so that it can better estimate the power?

    http://www.cycleops.com/en/products/software.html

    • Rodrigo Loureiro said:

      The released version of PowerCal is not configurable, there is no way to calibrate it. Initial (pre-release) had calibration options but cycleops eventually dropped the calibration phase from the device

  • Wouter said:

    Hi Rodrigo,

    Still surprised by your information and really like postings. Is it compatible with:

    Garmin edge 500 and 800
    Bryton rider 40 and 50

    Sounds like a great product for riders who are not really into pro training, but wants more than only heart rate.

    Thanks

  • Matt said:

    Hi Wouter,

    I used PowerCal with my Edge 500 yesterday , it worked fine after syncing with the head unit.

    I had to manually pair the PowerCal with the Garmin however.

    I would suggest reading your device’s owners manual on how to do so.

    Being a diehard PowerTapper I too was skeptical but after trying the PC out I was pleasantly surprised by it’s performance on my mostly Z3 ride.

    As Rodrigo stated , the PC can be a great tool for someone looking to get into the power training game on the cheap.

    Good Luck!

  • Wouter said:

    Thanks for your reply. Generally looking at all that has been posted on the web it maybe dismatches the extreme athlete (armstrong) and the super under trained individuals.
    (people who dont quite match the algorythm) But for ‘normal riders’ let’s say 3 times a week riders who are doing 190 watts with a heartbeat of 158, it is quite accurate. I really like this concept and looking forward to do some riding with it!
    Thanks for all the information, really helpfull.

    • Team Rodrigo said:

      Thank you. I’ve been using it quite extensively for the past month and it still matches my observations made here. It’s good to know that people out there are seeing the same results

  • Mike Edwards said:

    My experience isn’t as good as posted here. Using Powertap and Powercal together for steady endurance rides – what Powercal should be best at – Powercal showed only 60% of average watts and kJ for the session compared to my recently recalibrated Powertap. That’s a huge discrepancy which makes it almost useless for me. Will contact Cycleops to see if any kind of calibration is actually possible. I’m an (ageing) 3rd Cat rider so not undertrained or super strong. I’m also pretty unimpressed.

    • Team Rodrigo said:

      Yes, I would agree that 60% discrepancy is too much and would render results useless for training purposes. Is the difference consistent across multiple rides? I’m coming to the conclusion that it would have been good for cycleops to allow calibration, as it would iron-out those types of discrepancies. Just yesterday, I was talking to a rider that experienced the same level of discrepancy. In his case, this was due to a very low hear rate (threshold at 139). Clearly, when there are extreme metabolic factors involved, PowerCal may present abnormal results.
      However, from all data points I have, these are the exception and not the rule.

      • Mike Edwards said:

        I’m getting resigned to the fact that I’m some kind of outlier, although I have no idea why: I have three rides compared so far, all of which show very similar results and discrepancies, all of which fit the description of what Powercal was designed for. I have no heart abnormalities, my threshold HR at 160 bpm is unremarkable at age 48. To their credit Powertap asked to analyse the files to investigate but no word back so far. Still can’t see any reason why it should fail so badly for me but not others.

        • Team Rodrigo said:

          I’d be very interested to know if you hear back from them. I agree that from your description there seems to be no reason for the discrepancy. On the other hand, since the discrepancy is consistent, and consistently comes in at 60%, it allows you to mannually calibrate your results if you go out on a ride with the PowerCal. This all helps the case for CycleOps to introduce calibration of PowerCal. I agree with their conclusion that it is not necessary for most users, but for those that (for whatever reason) fall outside the bell curve created by CycleOps, it would be a great addition. Keep me posted

          • Simon Melton said:

            Team Rodrigo, great reviews thank you. I bought one of these after reading several reviews, yours especially. What settings are you using to get your consistent results Watts, Watts-3s, Watts-10s or Watts-30s, and which software are you using to output your figures into ?. Cheers Simon

          • Rodrigo Loureiro said:

            Simon,
            I have not done any calibration of the powercal unit, just using it out of the box.
            I typically use SportTracks for ride output analysis, and that is what I used for writing my post. You can check it out at http://www.zonefivesoftware.com/sporttracks/ I have found it to be the best journalling and analysis software out there. As a free version It’s already very good, and it becomes excellent using the many available plugins.
            For an even more detailed power analysis I often use PerfPro or TrainingPeaks.

        • Simon Melton said:

          Mike, having just bought one of these today as an entry method into “power” based training, at 46 and very close to your profile i would be interested to know if you have had any response ?

  • Wouter said:

    Rodrigo,

    Did test it few days ago with a bryton rider 40 on a tacx. The average was identical to what i recognized from previous medical sport max test , so i was pleased!

    The only thing (first time with power) that was not really ok was the 1 second wattage ‘going around like banana’ 220-200-170-230-235-190 etc.

    So i customised two fields into 30 sec en 3 sec average. That was the only possibility to get some control of riding with a constant wattage.

    1.Is that normal when training with power ? Or do you also ‘customise’ your fields in order to get it under control.
    2.Is there a bigger fluctuance with the powercal compared to power tap?
    3.Last question: seems like the joule GPS is a nice combo with it ? Or you prefer a specific computer with powercal?

    Anyhow, tasted like more, cant wait to ride outside.

    • Rodrigo Loureiro said:

      When using PowerCal, I’ve found that that numbers for short intervals (1 to 5 s) not very useful. Powercal does have a somewhat higher variability than powertap, but whereas in Powertap it does reflect instantaneous changes in the force applied on the pedals, for Powercal, these instantaneous changes are not reflection of an instant increase of the force on the pedals.
      Because Powercal computes power, derived from heart rate, it does so by measuring increases and decreases in heart rate and heart rate variability. However, these measures have a lag to instantaneous effort.
      I’ve found it more useful, when using Powercal, to user 10s Avg power, or even 30s average power instead of instantaneous power or even 3s.
      On my Garmin Edge800, I monitor 10s power and ride average power. I found these to be the ost useful with powercal. For powertap, I use 3s and ride average.
      Finally, for usage with powercal, I found the Edge800 to be better than the joule. I say this because the instant and short interval power are not very accurate, and I found the Joule an excellent combo for powertap, where I do short interval workout… I do find the Joule to be the best tool for short power intervals, as I can have the page with current interval and past interval and allows me to scroll. Very good for indoor use.

  • Mike Edwards said:

    Simon, no response and I’m not very hopeful at this point either. By now I’ve found Powertap figures (average power, kCals, Intensity Factor) are around 1.5 times greater than the Joule/Powercal. That knowledge at least makes it usable but if I’m estimating to that extent I might just as well have saved a lot of money and stuck with a HRM. Or spent more and bought another Powertap.

  • Matt said:

    I think the most important question isn’t being answered, and what I am interested to hear is repeatability which has always been the real problem with the non-traditional power gadgets. Honestly, it doesn’t matter if the reading is off a certain percentage so long as the numbers are consistent whether you are measuring power using watts or smiley faces.

    • Mike Edwards said:

      Matt, I agree, but that’s really my point and the source of my disappointment. There isn’t repeatability: the figures I’m getting “are around 1.5 times greater” so sometimes 1.4 sometimes 1.7 and without any easy-to-see rationale. The Powercal gives me figures that differ from Powertap with seemingly random variation. You might need to give me a crash course in smiley face data collection.

      • Matt said:

        Smiley face data requires a glass of chocolate milk during data collection! Thanks for answering my question from your experience. That was a very important piece of info that has been missing from a lot of the reviews.

        One thing I am interested in with your reply is why would you use something like this when you already have a power tap giving you the data which would be available to you whether on a trainer or the road. I would expect that this would be better for a person that isn’t using power on a regular basis as opposed to someone that has expectation of the numbers they get in certain power zones.

        • Mike Edwards said:

          Matt, I have a bike that I keep abroad so ride infrequently and when I do it tends to be endurance rides as opposed to intervals. Powercal seemed the better option than spending that much more on another Powertap system. Shame it didn’t work out for me (there’s an understatement) but I seem to be an outlier in that regard. Am seeking solace in alcohol-sponsored smiley face data. Expect further unrepeatability as a result.

  • Mike said:

    I recently did a VO2MAX and lactate test as a benchmark to assess my fitness. It was the first time I had used a power meter and after getting the report I did some research on the relationship between heart rate and power. It seems that this is a generally linear and repeatable relationship but can become decoupled over a long training session. So for long steady training sessions I would not think the PowerCal is any more useful than a HRM if you know what your relationship between the two (power and heart rate) is.

    The PowerCal does not work well for intervals and when you are not pedalling you know there is no power. So this again asks the question as to what is the benefit over a HRM…there doesn’t seem to really be any!

    • Mike Edwards said:

      I’d say that the value of Powercal with long duration sessions lies in the fact that, as you say HR and power become decoupled: cardiac drift, dehydration, muscular fatigue cause HR to go up for the same watts. So if you train monitoring just HR then you will produce less power as the session continues. Powercal should allow you keep training at the intended intensity.
      That said, the damn thing doesn’t work for me and after 5 months I’ve now given up any hope of Powertap getting back to me as to why not.

  • Mike said:

    Mike the rate at which the decoupling happens is an indication of fitness level – what I have not seen anywhere in any of the reviews is any indication that this shows up with the PwerCal. The data is there – just needs someone to do the correlation. if someone can give me the HR and Power data then I will try and look at it…

  • Mike Edwards said:

    Yep, Mike, I see your point – if Powercal is based off HR then the power it shows will decline along with HR. There must be a cut off point below and above which Powercal is no longer useful- I can see that 20 minute threshold intervals could work, 20+ minute Tempo intervals likewise.
    Again, though, you’re conversing with the wrong person about this. My experience with Powercal was so wretched I couldn’t recommend it to anyone. Powertap is a different matter…

  • Mike said:

    Mike, I think if one doesn’t have a power meter then at least getting on a bike with a power meter and making a note of heart rate at 100, 150, 200, 250W etc. will yield a nice set of data that can be used to do power training on the cheap provided that you are reasonably fit. I think if you are doing long training sessions of say more than 2 hours then the decoupling will become an issue.

    When riding on the road you know when the power is zero or less than where it should be – I don’t need a power meter to tell me that!

    Short intervals might be a problem but for longer intervals of say 3mins or more then it should be fine.

  • RLP said:

    Great review and nice interchange of thoughts. I have measured Powercal against PT and find it a useful tool with limitations. First ride out of the box I thought it was garbage and started an email exchange with the company. My HR runs much lower than average for men my age both on the resting and max side. The readings that Powercal gave me were proportionately lower than PT as my HR. While I am not getting a correct number, I am getting consistent data which a years worth of data will show my fitness as the season goes on. For the money, it has value. When coupled with TrainingPeaks, you can get a fair picture of fitness.

  • alat gps murah said:

    alat gps murah

    Team R – Awesome Bike Ride Adventures » Blog Archive » Comparing power readings from PowerCal and PowerTap

Leave a comment!

Add your comment below, or trackback from your own site. You can also subscribe to these comments via RSS.

Be nice. Keep it clean. Stay on topic. No spam.

You can use these tags:
<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

This is a Gravatar-enabled weblog. To get your own globally-recognized-avatar, please register at Gravatar.


− four = 0


Notify me of followup comments via e-mail. You can also subscribe without commenting.