PowerTap v. Zpower Comparison

(Clint Westhoff) #1

I decided to compare Zpower and my PowerTap since I have a KK Trainer. I recorded my PowerTap on my Garmin downloaded my Zwift .fit file and and then compared the two powertracks in SportsTrack. The red line is Zpower and the blue line is PowerTap power. On time I was offsync just a little bit when I tried to start my Garmin and Zwift in sync, but was pretty close.


(Kyle Polansky) #2

It’s pretty impressive how close the power numbers are. It seems that the ZPower was a little low at the beginning. Do you think this could be due to the trainer “warming up”? My trainer’s instruction manual said to ride for 5 minutes to allow the fluid time to adjust. I’m sure if you analyzed the ride from 5 minutes onward the data would match even better. Great test!

(Clint Westhoff) #3

I’m not sure why it was low at the beginning because what you are looking at has the first 10 minutes cut off (but maybe it still wasn’t warmed up as I was soft pedaling, then had to stop for a dog, then had to stop for a kid, so much of that 10 minutes was paused or going very slow).

I may do this again when I won’t get, or at least am less likely to get, interrupted - but I normally like riding my rollers instead of the KK.

I have experimented with virtual power on the KK before and it definitely needs to be “warmed up” to get the numbers closer as there is some temperature drift.

(Andrew Williams) #4

just from my own experiences riding rollers and observing the speed/power ratio…

If zpower is based on wheel speed (or some factor of it) then it will go up as your bearings warm up and get loose. With rollers the difference can be quite striking (especially since my workout room can get down to 45˚F at the start of a session). I’ve experimented with speed-cadence based power calculations versus the Quarq and seen it go from 20 watts under when cold to 30 watts over when warm, all for the same wattage on the Quarq.

In fact I have noticed the speed/cadence riders on Tour de Giro tend to start out moderately, and their massive attack runs don’t happen until later in the game. Those legitimate powerhouse riders who are putting up big numbers with a powermeter could attack right from the gun.

(N. Jekov) #5

The power curves do look quite nice indeed, but there seems to be some crazyness going on with ZPower/KK when wattage output is really high and high cadence numbers are involved. Sometimes you read they can hold around 1200 Watts a while, not even Kittel or Cavendish can do this.
Most of these rides were flagged and removed obviously, but e.g. look at Strava analysis from Col de Zwift leader, compare to his outdoor wattage output, then one can see something is not wright.
Probably it is easy to miscalibrate a VP Trainer with/without purpose.

(John Scott) #6

That is pretty impressive. From years of using am Elite Travel Trac trainer the lower power reading on the KK on your power spike makes sense. They do a good job of recording average power but they can’t get the full instantaneous power. Still the tracking was quite impressive.

(Thomas Wanek) #7

I would not be surprised if the “warm up” period on the trainer was responsible for the variance at the beginning of the chart. On my CycleOps Fluid2 there was (is) a very noticeable increase in resistance after about 10-12 minutes of riding, enough to have my heart rate jump about 10 bpm while holding the same speed. I always attributed it to the fluid changing viscosity as it warmed up.

(Josh Sakofsky) #8

do you have the speed data from those rides? i am curious, because it seems the people with zPower seem to have the edge in the competitions vs those with PMs/kickrs.

(Clint Westhoff) #9

@Josh - Not sure what you mean? My wheel speed, which is what the Garmin would have recorded is irrelevant. Avatar speed on the “island” would not be dependent upon “speed” but would be dependent upon power.

(Josh Sakofsky) #10

right, makes sense. i wish there was a way to compare speed on island while using PM to speed on island while using zPower. i guess only way would be to have access to two accounts, two computers and pair accordingly.

(Mike Moore) #11

I know this thread is a little old but I am trying to get the zpower number close to a real power number on Zwift.  When you did your test what tire were you using and how many turns of the resistance unit (full 360 degree) did you use to get the numbers to match up?  Thanks.  

(G Nguyen) #12

My setup is an old cycleops fluid 2, with a yellow power tap hub with wired shark fin sensor.  I’m using a garmin speed / cadence sensor but the on screen watts read 30W too low compared to the powertap, esp after the trainer warms up.  It was spot on before the warm up.  Is there calibration needed?