I received my kickr yesterday so looking forward to some experimenting today. Not sure I’ll keep it beyond the 30 day “no hassle” return date if I see similar issues.
all by email? I heard Wahoo is email support only these days, not sure if that is true.
In my past experiences, it took a phone call for them to admit anything. I’ve got a 2017 Kickr direct-drive and the drift over a workout is crazy. Factory spindowns help, for awhile. So I stopped using the Kickr’s power meter and now only use my crank-based power meter.
E-mail yes.
After doing a short erg workout last night I’d tentatively say my unit does not have an issue with drift in erg mode - lines up constantly well with my favero pedals.
It was only 30min workout - my next will be longer.
On a bit of a tangent - my trainerroad workout paused so I gave a time shift on the second part of the graph using the DC analyser - is there an easy way to correct this?
The only way I know is to remove the paused section in fitfiletools.com using the section remover, so you get a time shift in the other file aswell. But I have never been a fan of the section remover, I find it allmost impossible to cut out the correct section.
What I find easier, is to copy the file and use the section remover in fitfiletools to remove first half of one file and remove second half of the other file, then you import both in DCR-analyzer together with the TR-file and use time-offset on the second half-file.
Like I said - i need a longer workout but no indication of drift in erg mode yet for me.
I really tried to mess with it coasting/low cadence drills in the rest sections. I also coasted during the final set - that resuslted in a bit of a death spiral - but the power reported stayed the same as my faveros.
What i have noticed when comparing power from Kickr 2020 with my Favero Bepro Dual Pedals, if you you are not in the “optimal zone” of the kickr flywheel the power readings from the Kickr in comparison to the BePro Pedals seem inaccurate.
Example:
-
Kickr speed 20km/h, 90rpm, KICKR 3% LOWER than BEPROs (ok for me accurate wise)
-
Kickr speed 12km/h, 65rpm, KICKR 6% HIGHER than BEPROs (extreme case - in Zwift the flywheel speed seems to be higher even when climbing steep grades)
-
Kickr speed 40+km/h, 115rpm, KICKR 6% LOWER than BEPROs (default speed for me when doing flat rides in SIM MODE/Zwift)
So this isnt a problem when doing erg-mode as i chose the gearing. But for SIM MODE e.g. Zwift-Rides this results in lower power readings - compared to Bepro Pedals - the faster i go.
Can anyone reproduce this e.g. when going small gearing (34/26) and low cadence?
Files can be viewed here -
https://analyze.dcrainmaker.com/#/public/ddd35e42-4cb5-4b01-68f0-5f185e33ff8b
@LeanderM: I belive this is what they call linearity. @gpl posted about this earlier, but I couldnt find it now, anyways you can read more about it here:
https://tacxfaqx.com/knowledge-base/linearity-showdown/
This is why I sold my Neo2, it was supersensitive to cadence/gearing. At that point Tacx claimed it was my pedaling-style which was the problem, I belive this is a well known effect now. Neo2T handled this effect superbly though. Havent tested my KickrV5 yet, too many other problems 
@kevistraining: Thanks for sharing! Maybe you need longer workout, or maybe its only my Kickr that is the problem. It would also be nice if you shared the link to DCR, easier to read and zoom in then 
@daniebr So my next attempt at an ERG mode workout wasnt so hot - I seem to be seeing a similar issue.
https://analyze.dcrainmaker.com/#/public/3522e9db-4b62-41b5-5420-e11a69838091
It was no way near as accurate - 10% out compared to my favero pedals on the last set I managed to complete.
I’m going to have to choose some easier workouts if I dont use powermatch.
Not impressed so far - my Elite Zumo was better 
@kevistraining: If your Assioma read lower than the trainer, powermatch would make workout harder.
Weird that your second set is 10% off and first set is spot on.
I’ve messed up the naming on my fit files - it’s the assiomas that read higher.
Yeah I seem to have a “step” around 10mins.
CHecking in as I just received a V5 as warranty replacement for my KICKR2018 which turned into a paris roubaix simulator after 75 hours.
@daniebr what happened if anything when you finish the warm up and you just coast for 3-5 seconds during the workout in ERG mode? Does the KICKR get the opportunity to auto-calibrate to bring things in line and cut down on the drift, or did this not solve anything? Sorry if you covered this above but i was skimming the 50 or so posts to try and find the answer.
I’ve done a factory spindown and it seems better but not perfect.
Although the disparity I’m experiencing between my favero pedals and my kickr doesn’t actually seem to be related to erg mode as I see a similar thing happen during sim rides but it’s only obvious during sustained 300w plus efforts.
My pedals track consistently 2% higher initially (probably expected) but then I get a jump usually around 10mins when my pedals then consistently read 7% higher 
examples…
@RONDAL: According to posts over, the KickrV5 needs 2 seconds of coasting to auto-calibrate. I have tried this several times, but I have never been able to see any effect in my datasets.
As mentioned, I did a factory spindown before workout november 4th.
-
First ride after spindown nov4 showed less drift than before. PM reading +3% at start, and +4,5% at end, which gives a drift of only 1,5%.
Nov4: DC Rainmaker Analyzer -
The second ride after spindown looked pretty similar to the old ones again. PM reading +0% at start, and +4% at end, which gives a drift of 4%.
Nov5: DC Rainmaker Analyzer
I then did a second factory spindown right after workout november 5th, when trainer was warm.
-
1st ride after the second spindown showed less drift than before. PM reading -3% at start, and -1% at end, which gives a drift of only 2,0%.
Nov6: DC Rainmaker Analyzer -
2nd ride after the second spindown nov5 showed less drift than before. PM reading -2,5% at start, and -1% at end, which gives a drift of only 1,5%.
Nov7: DC Rainmaker Analyzer -
3rd ride after the second spindown nov5 showed less drift than before. PM reading -2,5% at start, and 0% at end, which gives a drift of only 2,5%.
Nov8: DC Rainmaker Analyzer -
4th ride after the second spindown nov5 showed less drift than before. PM reading -2,0% at start, and -1% at end, which gives a drift of only 1,0%. This was in Zwift, so no-ERG-mode.
Nov9: DC Rainmaker Analyzer -
5th ride after the second spindown nov5 showed less drift than before. PM reading -2,0% at start, and -1% at end, which gives a drift of only 1,0%.
Nov10: DC Rainmaker Analyzer
As you can see, the drift is now much less than it used to be, readings in sim-mode and ERG-mode also seems very similar, which makes sense. Most of the drift happens in the first 30-40 minutes, which means that after that, there is no more drift, things seems to stabilize. This is also very similiar to my earlier datasets from the Kickr18.
Im pretty satisfied now, if I substract 15-20min of warm-up, the drift between first and last interval is now about 1%.
Im not superimpressed by the accuracy, especially compared to my Neo2T. So Im still considering buying a Kickr Core, as I can sell the KickrV5(got a good deal) without loss, and my guess is that the Kickr Core accuracy should be as good as the one Im experiencing from the KickrV5.
There is one shop here in Norway which let you switch to another product if you are not happy after 30 days, that gives me the opportunity to switch for Direto XR if I dont like the Core. But if I end up with the Direto XR I dont save that much money, and lately I have been reading about people having trouble with huge resistance from Direto with Ant+, so Im not sure if I should risk that.
Can I take from your testing and analysis that allowing the KICKR to warm up properly and THEN doing a factor spin down is reducing the drift?
Yes, this also gave me similar results for ERG-mode and sim-mode, which also was different before.
According to Wahoo-support: “our developers are finishing up a firmware update that will clean up power readings. This should help resolve any residual drifts.”
-
Still amazes me that it takes this type of user testing and reporting to get to something that should have been shipped with the product when they released it. It’s not like the cases here are unique or unexpected.
-
They (meaning ALL trainer makers) can and should be testing these in the known steps and methods that are already well established. The fact that they are releasing these and either ignoring clear problems, or leaving testing to consumers is unfortunate at best and careless at worst.
-
We aren’t talking safety issues, but they are clearly not holding to their claimed accuracy and performance, which is not great when you consider how large and capable each of these companies really should be.
- I am probably wrong, but I thought the resistance units between V4 & V5 are different enough that they may well be experiencing “new” issues that were not present on the V4.
I would think this is well-known?
All eddy current brakes are subject to heat-induced drift. On a wheel-on trainer, the tire and roller have to be warmed up as well. Reaching a constant operating temperature for the system as a whole can take 30-45 minutes, or even longer. If you calibrate before things get hot, of course you will see significant drift over time.
The could be, I am not sure either way. It may well be that all they changed was the “calibration” aspect. Even if so (or especially if it’s the same), I still find it odd that they hadn’t had the chance to identify and fix this drift issue before shipping units.
Just seems they are not doing sufficient testing of known use to have missed this issue.