I’m fairly new to indoor training but been riding with power meters outdoors for a couple of years. My outdoor FTP was about 265W but on my new H3 a ramp test is showing 230W. It is pre season and I had a lazy winter so I thought nothing of it.
Just tried Antelope as an outdoor workout and there’s no way I could keep my power numbers down in the range I needed based on a 230 FTP, got back after 2 hours with a NP of 260W and that didn’t seem too hard.
What could be causing this mismatch? I calibrate both the pedals and the H3 so I’d expect them to be higher. Any ideas?
Laziness really. I can’t be bothered to swap the pedals from the road bike to the trainer bike all the time. Think that’s the answer though as I prefer the numbers from the pedals
Both will likely never match. Do the spin down for the H3 with the Saris app or rouvy and call it good. I get you like one over the other but just be consistent with whatever you are using to measure power each workout.
I have the H3 btw and a 4iiii but after a couple of rides just use the H3 number as over the workout the NP worked out to be really close.
The real issue I have is that I tried to do an outdoor workout and my power targets were way off, 200W on the trainer is difficult, outdoors its barely resting. I think I’m just going to swap the pedals for a bit and use them all the time. I might try ramp tests with pedals and H3 to double check what the offset might be.
My Assioma pedals and H3 are within a 1-3% or so. I would recommend pairing pedals to your head unit, and trainer to trainerroad, and record both. You can use DC rainmaker’s analyzer to compare them, or also Golden Cheetah.
If both are working properly, they should be pretty close. My take from so many people having problems with this is that there is not really truth in the power meter and/or trainer accuracy claims. This idea that that it’s OK and expected for a $800-$1500 trainer to be off by 10+% is silly, and the manufacturers should be held to a higher standard. If power match is being used to deal with more than about a 4% (ie power meter and trainer max error) difference with a high end trainer, it is working around defective equipment. I find it quite sad that this situation is so readily accepted.
I’ve been riding a few workouts indoors with both. Over 90 minutes today my NP was 186W (H3) vs 193W(Vector). Bearing in mind margins of error and drive train losses between pedals and hub I reckon that’s not too bad.
I reckon a cooling fan and a bit more indoor riding will see my FTP get closer to my outdoor.
The vectors do react a lot quicker so any time I glanced at both the Garmin was higher than TR.
Thanks for the idea. If I can pluck up the energy I’ll play comparing in more detail but the NP is good enough to confirm its not miles out.
The DCR analyzer is great for getting into the details, but that’s not always needed.
For just quick ‘is this really off’, I end up just hitting lap on my wahoo connected to my pedals at the start of an interval, and compare that to the interval power. I was doing this a lot tracking down some weirdness with my kickr, but on the H3 things have always been pretty close.
Your NP was 3.7% different. For comparing trainer accuracy, I would look at average power, not normalized, as normalized is trying to take into account human physiology.
NP comparison is fraught with issues. You could have similar NP but very different results, because the errors at higher power levels could be offset by that at lower levels, etc. You are much better to do sustained steps at various power levels, and a few rapid changes. The former allow testing of steady-state measurements, the latter the time response. That’s why DCR uses a 30/30 workout as his mainstay test: it does both.
I like the idea of doing comparing laps on the head unit. I’ll find a SS workout and compare average power with a couple of the intervals. I’m not that bothered about the response times, I’m new to ERG riding so my legs react slower than any trainer would.
Or if I’m stuck inside much longer I’ll actually read the DCR articles…
