PowerTap Cadence vs. Dedicated Cadence Sensor [Spoiler - the Cadence Sensor is way better]
Last year Coach P put my power file from 2010 IMLP in to the crucible, I took away a few things from this.
One concern was cadence, according to my power file, I was spending a lot of time over 100RMP, actually a lot of time over 110RPM too and that didn't make a lot of sense to me. I feel comfortable in the 90-100RPM range and I do spin-ups in 110-115RPM. Fast leg speed is considered a good thing, but what I was seeing after I looked at a bunch of my rides in WKO was way out-of-line with reality. My cadences were way higher than what I was actually pedaling.
I use a PowerTap Elite+ with a Garmin Edge 705, and I have the 705 get the interpolated cadence from the PT. I believe the cadence a PT generates is based on some math and timing of downstrokes on pedal swings, but it was way off for me. Maybe it was the 705, either way, I needed to correct this.
I had a Garmin GSC-10 Speed/Cadence sensor, but I couldn't find a good way to mount it to the chainstay on my 2010 Specialized Transition. A couple of weeks ago I finally mounted a sensor to my bike and took it for a spin on my usual after work route and rode like I normally do.
Wow, what a difference - they graph below shows two rides of the same route and time with vastly different cadences.
To sum up, if you're having cadence accuracy issues with your PT, consider getting a real cadence sensor. It made a huge difference for me.
Comments
Nice to know this information. I have a couple questions.
1) I also have a specialized what sensor did you finally use and was there any issues or technique to mounting it.
2) I use the LYC still and probably won't be updating for awhile. Does you 705 capture the new sensor data, I assume it does, or is this another small unit that need to be mounted. Would this system work with a LYC of would it require an upgrade? I was considering a Joule when I do upgrade does anyone have information on how this unit would capture sensor data. Let just say I am trying to keep it as clean as possible.
Thanks
Gordon
@Gordon,
The Garmin Edge 705 allows you to get the interpolated Cadence from the PM which is what I was trying to avoid, so I put a "manual" cadence sensor on (it also does speed). It required a magnet being mounted to the crank arm and a unit mounted to the chain stay that counted crank revolutions as the magnet passes and then transmits this information to the head unit via ANT+.
I use the Garmin GSC-10 Speed/Cadence Sensor it's ANT+, but I do not know if it's compatible with other head units (LYC, Joule, etc.). I'll post a photo of how i mounted it. Nothing really special, the Specialized transition's chainstays have a rectangular cross-section, and it tapers as it gets closer to the dropout so it makes it mounting somewhat difficult.
Cary - I'm wondering if maybe something is wonky with your PT? Although that's unlikely if Matt saw the same thing.
@Craig,
I'd normally defer to you on this as I know you've researched Garmin power quite a bit (and I've linked to your Garmin Power Page on the forums several times), however I see these wacky numbers all the time on routes that are flat, hilly and everything in between. You should have seen the RPMs I put out at IMLP last year - in The Crucible, Coach P asked me if I left something on the course by perhaps spinning to much. Also, as you pointed out, others that have posted here have had the same issue.
Now, it's possible I have something set wrong on the Edge 705, but adding the GSC-10 to the mix put all of my numbers exactly where they are supposed to be. I can tolerate "close" enough for cadence, but these numbers I was getting were in outer space.
If I can correct this without having to add a GSC-10 to my road bike, I'd be very happy, so if you can shine any light on the subject I'd be stoked.
The difference between native PT cadence and the extra sensor ( I have the PT cadence sensor) in accuracy of the reading and stability is noticeable. Every so often when I stop I may hit the extra senor and move it just enough where it does not pick up, so the Joule defaults back to the PT for cadence. I just looked at one ride on my normal course on the flats. The first 2:00 cadence avg was 91 with a max of 107. This is normal for me. The last hour I had hit the sencor so I was back to the PT for cadence. I can immediately notice it based on the readings on the Joule. The data for that hour was an average of 106 with a max of 200 rpm. The last hour was a steady ride at my normal pace and cadence. I saw this issue from the start with the RPM on the PT compared to RPM on the computrainer. The difference was to big for my data driven head to be comfortable with, so I added the extra cadence sensor and have been happy.
Al seems to be having a better experience with his PT and Joule, so there may be variation between how well the hubs resolve rpm. But as I say that, that i think rpm is a key element in how watts are calculated by the device????
Power = torque X 2 pie X rpm
So if we have a fluctuating rpm we get a fluctuating power reading. Solaris must have this taken into account to be able to say that the power readings are +/- 1.5 % accuracy.
Ok , I swear my last post on this one
I han to understand whats up with the PT and candence.
Ok – for inquiring minds who want to know, go to this page for a good overview of how the common power systems on the market do what they do.
http://anonymous.coward.free.fr/wat...html#speed
The Power Tap calculates its cadence by analyzing torque variation received by the hub (if torque is maximized at the 3 o'clock and 9 o'clock positions of the crank, and minimized at 6 and 12 o'clock, one can use the variation in torque to count crank revolutions).
Now I can sleep tonight
Cadence - why do we need it?
(All N=1 here.)
From a instant feedback point-of-view, cadence is really only important when you're looking to change a cadence "habit". Perhaps when you start riding, you're a 65-75 RPM rider and you want to bump that up 10-20 RPMs, you keep an eye on your cadence and make pedaling adjustments until it becomes a neuromuscular habit. After that you probably won't check your instantaneous cadence unless you're feeling off one day or desire to make another change in leg speed.
Another matter however is cadence as a function of post-ride analysis. If you see a cadence like I have in the original image on the left, you'd think there was something wrong, so you'd want to fix that. Or perhaps you want to see how slow or fast you were pedaling on climbing sections etc.
So I just wanted to get cadence recorded correctly, not to check myself as I was going (although I was worried there for a minute that I was over spinning), but to check myself at key points of a ride (via quadrant analysis etc.). The dedicated Cadence sensor provided me with piece of mind that I was getting accurate data to make my post ride analysis.
It'd be great to hear from Craig on this as well.
Sorry for the delay in replying. I had half-heartedly tried to find an old thread on this subject a couple times and in my brain was convinced it was on ST. Turns out it was on the google wattage list.