I have a Zwift Ride with KICKR CORE, which may have a problem. Despite the firmware of the KICKR CORE and my Garmin RS200 pedals being up-to-date, and both manually calibrated before each ride, I’m getting a significant difference between them, with the latter averaging -13%. One of them must be significantly wrong, or both of them… I’ve ordered a pair of Assioma Pro RS-2 pedals to further compare, but they won’t arrive until late next week.
Does anyone happen to know of other public dual recording datasets on ZwiftPower showing power measurements from a KICKR CORE vs Garmin RS200 pedals? I’m intrigued to see what difference other people may have recorded. There is a dataset on my ZwiftPower profile, under ‘Analysis’ (I’d share the direct link, but this platform is not allowing me to do so).
I have read that the Kickr Core can read higher than other power sources.
A friend has a KickrV5 at his house & a Kick Core at his vacation home, and says that the Core reads ~10W higher than the V5.
My Kickr V6 varies from my Assioma Duo pedals, but not always the same amount.
The easiest solution for accuracy & consistency is to use the “Control with Ant+ power meter” within the Kickr Core’s settings on the Wahoo app.
A small discrepancy between power measured at the pedal vs measured at the rear hub is expected and normal because some power will be lost through the chain.
One thing that’s often overlooked is the importance of using a torque wrench to install your pedal power meters. Every manufacturer will publish their torque specs, and this is important because if you undertorque or overtorque, you change the expected preload on the pedal spindle, which in turn affects how the strain gauge reads. Do you have a torque wrench?
Regarding that drivetrain: as your chain / cassette / chainring wear together - the power losses stack up over time. Putting a new chain on a clapped out cassette and chainring isn’t going to bring the efficiencies back to the like-new state. Be sure to look closely at the condition of your drivetrain while you’ve investigating.
Thanks for responding. I’d be happy with a small discrepancy, but not 13%. The drivetrain is in very good condition (thoroughly clean and relatively new).
Interesting point about torque and spindle preload. I see Garmin recommend a torque of 34 N-m when installing the RS200 pedals. They don’t make a big deal of it in the instructions, just a recommendation. Mine were not installed to such a high pressure; just a grunt or two with a manual wrench. I have a torque wrench that goes up to that pressure, but I’ll need a 15mm crowfoot adapter, which I’ve just ordered (Park Tool TWB-5).
This keeps the effective length of the lever arm the same as using a socket type extension (like an 8mm hex socket). This is an overlooked / less-known detail that that affects the wrench’s torque reading.
Since you’re going to this trouble, would you loop back and let us know if re-installing the pedals following Garmin’s instructions helped narrow the (expected) discrepancy?
One other good general practice I’d like to pass along: when you manually calibrate your power meter(s) - you should do that once they’ve reached operating temperature. An obvious case is if your pain cave is in an unheated basement or garage in a place where winters get very cold, but it’s still good practice in all climates.
The crowfoot arrived and I got around to another power metre comparison, this time tightening the RS200 pedals to 34 N-m when installing them on the cranks, as recommended by Garmin. I did an intervals session and wasn’t expecting to see much change, but was amazed to see far better alignment between the power measurements of the pedals and the KICKR CORE. The dataset is public on my Zwiftpower page and below is a screenshot. The pedals measured an average power difference of -3.7%, far better than the previous -13% during the same workout. I’ll continue to test, but am happy with this margin of difference. Many thanks for your advice regarding the importance of getting the torque right when installing the pedals on the cranks
Hi @Des_Callaghan
Thanks for looping back with such a detailed report! Following the manufacturer’s torque spec matters a great deal.
There are other details you have to pay attention to in the owners manual. Most of them won’t advise you about the correct orientation of a crowfoot extension, but that matters. Do they say you should grease the pedal threads or not to grease? That also matters in how the torque wrench reads applied force.
Looking at your graph, the 2 appear to be closer early and seem to deviate more towards the end.
This might indicate a temperature issue.
If it is warm where you Zwift, you might want to use a cooling fan on your trainer.
I don’t see that in the data, Tim. If you look at the high power intervals periods (about 350 w), the measurements align more-or-less perfectly throughout the session. Also, if measurement drift due to rising temperature occured with the KICKR CORE, we would expect a drop in power readings due to the decrease in resistance, as far as I understand it, both during the low and high power phases. That doesn’t occur. Please correct me if I’m wrong.
Interestingly, though, the KICKR CORE generally reads higher during the soft pedalling between the high intervals, which is especially noticeable during the cooldown at the end (see graph below). Both estimate cadence more-or-less exactly the same during this period. However, the pedals have much more variation in the data, and my impression is that they are providing more accurate measurements during the softest pedal strokes, but that is just a guess.
My first comment was just eyeballing the curve .
I went to the dual recording on ZP and zoomed in.
If you zoom in on the 8-10 min around your first interval, you power meters were about 2.65 % but then the 8-10 min around your last interval, it was 3.8 ish.
The 3rd interval, in isolation was was over 4% off.
Overall, the whole workout was 3.5%.
I’m not sure what can be determined from analysis of 1set and I’m not familiar with your set up.
Not trying to diss you either.
It just appeared to vary more towards the end.
That may not be the case.
Keep up your excellent efforts to improve your data.
Well, just as I thought the problem may be solved, here it is again. This is the same intervals workout, completed this morning. The pedals differ by an average of -9.4%, which is too much. The data are public on my Zwiftpower page. During the high power efforts (about 350 w), the measurements are closer, within 4 to 6% of each other. During the low power recovery phases (about 90 w), they are much further apart, within 20 to 25% of each other. I can’t see any obvious drift in the KICKR CORE data that would be associated with warming inside the trainer (the KICKR data remain consistently higher than the pedals, by a broadly similar margin). I have a new pair of Assioma pedals arriving at some point, so I’ll compare them with the KICKR when they arrive. Other than that, I’m not sure what else could be done.
I thought the general rule of thumb was that pedal/crank power meters should read ~3% higher than the turbo reading, due to drivetrain losses on a clean drivetrain.
Yes. DC Rainmaker suggests 2-4% loss through the average drivetrain on a trainer, which suggests my pedals and trainer are measuring further apart than their data suggest (the pedals are averaging 9.4% lower).
