Login required to started new threads

Login required to post replies

Prev Next
Re: Controlling the Kickr via external power meter now possible [J_R] [ In reply to ]
Quote | Reply
Yes, you are correct on the HR and cadence.

CEO at TrainerRoad
Co-host of the Ask a Cycling Coach Podcast
Quote Reply
Re: Controlling the Kickr via external power meter now possible [WahooMurray] [ In reply to ]
Quote | Reply
So what does this Wahoo beta software update really do for anybody? It only helps out those people who 1) have a Wahoo 2) have a crank PM that they can use on their trainer bike 3) don't use TR or PerfPro - seems like that is a pretty small group of people. What about those people who don't have a crank based PM and/or expected the Wahoo to produce watts that were within an acceptable margin of error compared to other PM's? OK, so supposedly less drift and better spin down..............
Quote Reply
Re: Controlling the Kickr via external power meter now possible [ccassidy] [ In reply to ]
Quote | Reply
ccassidy wrote:
So what does this Wahoo beta software update really do for anybody? It only helps out those people who 1) have a Wahoo 2) have a crank PM that they can use on their trainer bike 3) don't use TR or PerfPro - seems like that is a pretty small group of people. What about those people who don't have a crank based PM and/or expected the Wahoo to produce watts that were within an acceptable margin of error compared to other PM's? OK, so supposedly less drift and better spin down..............

If you're on iOS with TR you'd be able to use their solution without having an extra ANT+ dongle.

CEO at TrainerRoad
Co-host of the Ask a Cycling Coach Podcast
Quote Reply
Re: Controlling the Kickr via external power meter now possible [ccassidy] [ In reply to ]
Quote | Reply
ccassidy wrote:
So what does this Wahoo beta software update really do for anybody? It only helps out those people who 1) have a Wahoo 2) have a crank PM that they can use on their trainer bike 3) don't use TR or PerfPro - seems like that is a pretty small group of people. What about those people who don't have a crank based PM and/or expected the Wahoo to produce watts that were within an acceptable margin of error compared to other PM's? OK, so supposedly less drift and better spin down..............


The new firmware now has temperature compensation and better spin-down calculations, everyone will benefit from these changes. Lots of people will also love being able to bridge there ANT+ based cadence sensors via Bluetooth.

Linking 3rd party power meters is not to correct the error in the KICKR, its to match the power readings. All power meters (including KICKR) have a margin of error and read power from different points in the system so we are simply offsetting the readings so they match. Unfortunately this isn't available for hub based power meters.

Hope that helps.
Quote Reply
Re: Controlling the Kickr via external power meter now possible [WahooMurray] [ In reply to ]
Quote | Reply
WahooMurray wrote:

The new firmware now has temperature compensation and better spin-down calculations, everyone will benefit from these changes.

Thanks for the update. I applied the firmware this afternoon before my workout.

Just a question regarding the spin down calculations. If one is using TR or PP for the workout with the power meter in control is it necessary to do a spin down? I stopped since power meter control and everything seems to track okay without a spin down. The Kickr resistance adjusts pretty well based on the power meter.
Quote Reply
Re: Controlling the Kickr via external power meter now possible [WahooMurray] [ In reply to ]
Quote | Reply
Looks like the password is "publicbeta", not "public beta" (no space).
Quote Reply
Re: Controlling the Kickr via external power meter now possible [Felt_Rider] [ In reply to ]
Quote | Reply
Felt_Rider wrote:
WahooMurray wrote:


The new firmware now has temperature compensation and better spin-down calculations, everyone will benefit from these changes.


Thanks for the update. I applied the firmware this afternoon before my workout.

Just a question regarding the spin down calculations. If one is using TR or PP for the workout with the power meter in control is it necessary to do a spin down? I stopped since power meter control and everything seems to track okay without a spin down. The Kickr resistance adjusts pretty well based on the power meter.


You wouldn't need to do a spin down on the KICKR but you should follow the manufactures instructions for calibration on your power meter, you may need to do a manual zero depending on the brand you have.
Quote Reply
Re: Controlling the Kickr via external power meter now possible [ccassidy] [ In reply to ]
Quote | Reply
ccassidy wrote:
So what does this Wahoo beta software update really do for anybody? ... OK, so supposedly less drift and better spin down..............

What else would you want it to provide? Obviously the improvement is yet to be demonstrated, but this is exactly what people on this thread have been asking for.
Quote Reply
Re: Controlling the Kickr via external power meter now possible [agg] [ In reply to ]
Quote | Reply
agg wrote:
Looks like the password is "publicbeta", not "public beta" (no space).

Thanks for that, I've edited the original post.
Quote Reply
Re: Controlling the Kickr via external power meter now possible [WahooMurray] [ In reply to ]
Quote | Reply
WahooMurray wrote:

Linking 3rd party power meters is not to correct the error in the KICKR, its to match the power readings. All power meters (including KICKR) have a margin of error and read power from different points in the system so we are simply offsetting the readings so they match. Unfortunately this isn't available for hub based power meters.

.

Murray,

Just for the record what is Wahoo's current claim in regards to the Kickr's margin of error? If I remember correctly back at the time of introduction, the claim was + or - 2% with the potential to eventually be within .5%. Do you feel that the 2% value is being met?

Thanks for joining the thread.

Hugh

Genetics load the gun, lifestyle pulls the trigger.
Quote Reply
Re: Controlling the Kickr via external power meter now possible [ccassidy] [ In reply to ]
Quote | Reply
ccassidy wrote:
So what does this Wahoo beta software update really do for anybody? It only helps out those people who 1) have a Wahoo 2) have a crank PM that they can use on their trainer bike 3) don't use TR or PerfPro - seems like that is a pretty small group of people. What about those people who don't have a crank based PM and/or expected the Wahoo to produce watts that were within an acceptable margin of error compared to other PM's? OK, so supposedly less drift and better spin down..............

Well. Some people (myself included) do use KICKR in Simulation mode rather then ERG. While it is very easy for 3rd party apps (at least conceptually) to correct KICKR so it matches external power meter in ERG mode it is very difficult/impractical to implement for Sim mode. In firmware however the latter is just as simple as for ERG mode.
Quote Reply
Re: Controlling the Kickr via external power meter now possible [WahooMurray] [ In reply to ]
Quote | Reply
Murray,

First of all, thanks for addressing the issues. I'm looking forward to see what effect the new firmware has.

Can I ask you to clarify a few things in the above ? It's not quite specific enough for me to understand fully what you mean.

WahooMurray wrote:
Linking 3rd party power meters is not to correct the error in the KICKR, its to match the power readings. All power meters (including KICKR) have a margin of error and read power from different points in the system so we are simply offsetting the readings so they match.

I had previously understood 'linking 3rd party power meters' to mean that the KICKR used that external PM as the basis of its decision to increase or decrease resistance. For example, if the workout software (e.g. Wahoo Fitness) asks for 200W, then the KICKR adjusts the resistance until the external PM reports 200W load, and *completely ignores* the internal PM. Here there is only one feedback loop, involving the brake and external PM. That's a perfectly valid approach, especially for those who want to use the same PM on the road.

Or does it mean that the internal firmware conducts power measurements using both the internal and external PM, and determines an offset between the two, but still uses the internal PM with the offset applied to get an alternative estimate of applied resistance which better matches the external PM. Here there are two feedback loops - a high bandwidth loop involving the brake and the internal PM, and a low bandwidth loop involving the internal PM and the external PM. This could also work well, and possibly even better if the bandwidth of the internal brake/PM feedback loop is higher than the ANT+ based external one, and the offset is measured sufficiently often to compensate for any temperature related drift (or drift for any other reason).

There's a third explanation, which I suspect (sincerely hope) you're not doing: measuring the offset between the internal PM and the external PM, and adjusted the reported power from the internal PM by the offset amount. That would mean you report a certain power value, even though the resistance isn't being adjusted accordingly - that would be misleading. Anyone with two power meters would uncover that very quickly indeed...

There may be other explanations, but I can't see what they are right now.

WahooMurray wrote:
Unfortunately this isn't available for hub based power meters.

Presumably because said hub-based power meters need to be removed from the bike to fit it on the KICKR! Or am I missing something?

Finally, I'd also like to hear the estimated accuracy with these fixes to the internal PM. I'd love to be able to believe my KICKR's reported power without having to use an external PM.
Last edited by: marting: Apr 8, 15 1:33
Quote Reply
Re: Controlling the Kickr via external power meter now possible [marting] [ In reply to ]
Quote | Reply
Hi,

Your 2nd explanation is about right, we still use the internal power meter to control the KICKR and this needs to be fast and responsive. We simply use the external power meter to adjust this feedback loop. For this reason users should see slightly better performance than software based solutions.


marting wrote:
Or does it mean that the internal firmware conducts power measurements using both the internal and external PM, and determines an offset between the two, but still uses the internal PM with the offset applied to get an alternative estimate of applied resistance which better matches the external PM. Here there are two feedback loops - a high bandwidth loop involving the brake and the internal PM, and a low bandwidth loop involving the internal PM and the external PM. This could also work well, and possibly even better if the bandwidth of the internal brake/PM feedback loop is higher than the ANT+ based external one, and the offset is measured sufficiently often to compensate for any temperature related drift (or drift for any other reason).

marting wrote:
WahooMurray wrote:
Unfortunately this isn't available for hub based power meters.

Presumably because said hub-based power meters need to be removed from the bike to fit it on the KICKR! Or am I missing something?

Correct.

marting wrote:
Finally, I'd also like to hear the estimated accuracy with these fixes to the internal PM. I'd love to be able to believe my KICKR's reported power without having to use an external PM.

Its a really tuff question, yes I can get you the accuracy against our baseline but that doesn't tell you how well does it match your own power meter. For example, if 2 power meters have a +/- 3% accuracy thats unto 6% assuming they used the same baseline and thats not even taking into account the different locations power is collected. A pedal based power meter should always report a higher power compared to a hub based because of the loss in the chain etc. Just ask Ray how hard it is to compare 2 different power meters.
Quote Reply
Re: Controlling the Kickr via external power meter now possible [WahooMurray] [ In reply to ]
Quote | Reply
WahooMurray wrote:
marting wrote:
Finally, I'd also like to hear the estimated accuracy with these fixes to the internal PM. I'd love to be able to believe my KICKR's reported power without having to use an external PM.


Its a really tuff question, yes I can get you the accuracy against our baseline but that doesn't tell you how well does it match your own power meter. For example, if 2 power meters have a +/- 3% accuracy thats unto 6% assuming they used the same baseline and thats not even taking into account the different locations power is collected. A pedal based power meter should always report a higher power compared to a hub based because of the loss in the chain etc. Just ask Ray how hard it is to compare 2 different power meters.

Thanks Murray, understood. There's no way you can know either (a) where in the spectrum my Vectors sit, or (b) what my drivetrain losses are. So an estimate of the absolute accuracy of the KICKR's internal measurement as delivered to the cassette is about as good as you'll be able to do. If you're able to repeatedly deliver +/-2% across a specified range of operating conditions (e.g. ambient temperatures appropriate for commercial, indoor electronics), that's comparable with others' claims, and would be "trustworthy" enough for me, and I suspect for many others too.
Quote Reply
Re: Controlling the Kickr via external power meter now possible [WahooMurray] [ In reply to ]
Quote | Reply
WahooMurray wrote:

You wouldn't need to do a spin down on the KICKR but you should follow the manufactures instructions for calibration on your power meter, you may need to do a manual zero depending on the brand you have.

Thanks
Certainly something I need to get into a new habit with using the training applications. I've been so accustomed to doing it with the Garmin Edge for so many years that I forgot to do it last night in the software. Your post reminded me that I forgot because I am now comfortable not using the Garmin to record the power meter and just use the application data. :-)

I appreciate you revealing that your team spent a good number of hours testing and tweaking. I hope you all (including TR and PP) will refine as things come up.
For me, at this moment and for the past weeks, my training is going in the right direction.
Quote Reply
Re: Controlling the Kickr via external power meter now possible [WahooMurray] [ In reply to ]
Quote | Reply
WahooMurray wrote:
For example, if 2 power meters have a +/- 3% accuracy thats unto 6% assuming they used the same baseline and thats not even taking into account the different locations power is collected. .

Murray

Since you mention +/- 3% here and didn't get a chance to answer my more specific earlier question, am I to take it that Wahoo is now aiming at the +/- 3% value? It's interesting that the vast majority of reports of Kickrs that are "off" are showing power a good bit higher than the external power meters on the bikes of the reportees. I'd think that a pedal or crank based power meters would read a bit higher than the Kickr due to drive train losses.

Thanks again for participating in this thread.

Hugh

Genetics load the gun, lifestyle pulls the trigger.
Quote Reply
Re: Controlling the Kickr via external power meter now possible [sciguy] [ In reply to ]
Quote | Reply
My KICKR starts out 10-15 watts higher than my P2M and then as it warms up settles in to 20-25 watts higher as it warms up. I'd prefer to use a training bike without a PM attached semi-permanently to the KICKR, so I'm really hoping that the new firmware helps with the offset and drift. If not, I'd really like a way just to be able to increase the resistance of the KICKR by 20W and know that it's going to read low for the first 15 minutes of my warmup.
Quote Reply
Re: Controlling the Kickr via external power meter now possible [bloodninja] [ In reply to ]
Quote | Reply
Thanks Murray - Did few hrs on trailen tonight, it's much improved or more accurate compared to my PM. I rode in erg and SIM mode (2%). Was within 2% or better for Z 5/6 intervals and also on longer Z2/3 efforts.........not so close on recover power.

will play some more this week - didn't bother with PM control option

thanks
Last edited by: gormleyflyer: Apr 8, 15 20:39
Quote Reply
Re: Controlling the Kickr via external power meter now possible [marting] [ In reply to ]
Quote | Reply
Just tested my Kickr with the publicbeta firmware. The kickr took about 4-5 mins to start accurately applying the target resistance that TrainerRoad program was asking for. After that it was extremely accurate until an acceleration in cadence (rpm) which it took approx 7-10 seconds to stabilize. Where the real problem arose was 30sec on 30sec off intervals. These intervals had rpm accelerations from 100rpm to 130 rpm.The kickr would initially apply enough resistance to see the power overshoot the target power and then after peaking within 5 or seconds the resistance would gradually decrease to about 20 below the target power all while the cadence remained between 130-135rpm.
Tomorrow I will try a 3x10min interval session and see how that works.
Quote Reply
Re: Controlling the Kickr via external power meter now possible [Bonedog] [ In reply to ]
Quote | Reply
Power is derived from a combination of cadence speed and pedal force.
The erg unit will always be guessing everytime there is a change in rpm
If to increase or decrease the resistance.

Making this guessing as immediate as possible is tricky
Whilst being able to provide smooth resistance at the same time.
Quote Reply
Re: Controlling the Kickr via external power meter now possible [lennyk] [ In reply to ]
Quote | Reply
lennyk wrote:
Power is derived from a combination of cadence speed and pedal force.
The erg unit will always be guessing everytime there is a change in rpm
If to increase or decrease the resistance.

Making this guessing as immediate as possible is tricky
Whilst being able to provide smooth resistance at the same time.

I have found success in increasing my cadence to where I think I want to be about 5 seconds before the next jump in power, this way the KICKR seems to have less "guessing" to do as my cadence shouldn't change much if at all.
Quote Reply
Re: Controlling the Kickr via external power meter now possible [WahooMurray] [ In reply to ]
Quote | Reply
Just tested the new firmware update on Zwift Island tonight. Recorded the Kickr using a Mac and my SRM using my Garmin. Huge improvement. This was a 1:26:41 ride. Here’s what I got:

Kickr: Avg power: 226. Max power: 665
SRM: Avg power: 215. Max power: 671

The Kickr still seems a bit generous, but these numbers seem pretty good. In comparison, with the previous firmware, the differences were consistently +30-35 watts for the Kickr.

I haven’t tried using the power meter control feature yet, as I’ve had no luck with that in the past. I might give it another try tomorrow. If so, I’ll let you know my results.

Thank you, Wahoo! This update was much needed and much appreciated!
Quote Reply
Re: Controlling the Kickr via external power meter now possible [jmuise] [ In reply to ]
Quote | Reply
jmuise wrote:
lennyk wrote:
Power is derived from a combination of cadence speed and pedal force.
The erg unit will always be guessing everytime there is a change in rpm
If to increase or decrease the resistance.

Making this guessing as immediate as possible is tricky
Whilst being able to provide smooth resistance at the same time.


I have found success in increasing my cadence to where I think I want to be about 5 seconds before the next jump in power, this way the KICKR seems to have less "guessing" to do as my cadence shouldn't change much if at all.

It's definitely a balance in the control system, we have been looking at making the algorithm more aggressive when changing that target resistance by a large amount but still haven't got something that feel right.
Quote Reply
Re: Controlling the Kickr via external power meter now possible [WahooMurray] [ In reply to ]
Quote | Reply
I've got the new beta Kickr firmware installed and I'm using latest production release of TrainerRoad -- NOT the most recent previews shared by Nate. I've got the Kickr set to "Control w/ ANT+ Power Meter" with my Power2Max connected.

Things are NOT great. My TR workout was pyramid intervals at 338W and 84W for the rest intervals. See here...

https://www.trainerroad.com/.../rides/1941485-A3-60

At the beginning of each work interval, the resistance would shoot up like crazy for about 20 seconds and then usually (but not always) stabilize. In 5 of the 8 intervals, the first 20 seconds was over 400W with the worst being in the mid-500's! Intervals at 338W are really tough for me, and having to do the first 20 seconds at 400-500W is super frustrating.

Btw, I was also getting huge oscillations in power during my rest intervals.

In the TrainerRoad devices panel I can see that the Kickr thinks its power is right on, but my Power2Max and my legs beg to differ.

WahooMurray, what's your assessment of what I've described?
Quote Reply
Re: Controlling the Kickr via external power meter now possible [WahooMurray] [ In reply to ]
Quote | Reply
WahooMurray wrote:

We would love for you to try out the new firmware and give us feedback. Initially it would be great if you could just update the firmware, do a spin down and compare the results to your power meter, ie don't link you power meter to the KICKR, we would love to hear how the raw numbers look.

Over the last 3 days I have done 3 rides on my KICKR with TrainerRoad. In each ride I have been recording the data from the KICKR through TrainerRoad as well as from my Garmin Vectors on a Garmin 810 head unit. As requested, I evaluated 3 different scenarios:

Ride A: Old firmware, no ANT+ control of KICKR
Ride B: New firmware, no ANT+ control of KICKR
Ride C: New firmware, ANT+ control of KICKR

Ride A was a 60 minute ride. I first did a static calibration of the Vectors, then did a 15 minute warm-up, then did a spindown of the KICKR (I did not write down the spindown information as this workout was before the new firmware was released). The data presented below is all taken after the spindown (so the last 45 minutes of the workout).



The graphs show the power-duration curve for the KICKR and Vector (top) and the difference in the power-duration curve (bottom). The dashed lines in the top graph show the 3% error bars. The fact that the error bars for the KICKR and Vector do not overlap shows that the difference in their values is more than 6%. The average difference (indicated by the dashed red line in the bottom graph) is around 30 watts (with the KICKR reading high).

I installed the new firmware and repeated the experiment from Ride A, although I did a different workout. Once again I did a static calibration of the Vectors, did a ~15 minute warm-up, then did the KICKR spindown. The spindown offset was 416, spindown time was 29.25 seconds, and the reported temperature was 30.13 C. The data in the below chart is once again taken after the spindown.



Once again, the error bars do not overlap, indicating that the error between the KICKR and Vector is more than 6%. However, the average error decreases from ~30 watts to ~22 watts, an ~8 watt improvement.

In Ride C, I configured the KICKR to be controlled by the Vectors using the newly introduced functionality in the KICKR firmware. For this ride, I show the watts over time to illustrate how it went.



The reported power for both the KICKR and Vector were pretty similar in the beginning. However, after about 13 minutes (vertical purple dashed line), the Vector power jumped up a few watts (this increase was small enough that I didn't notice it during the workout, I only noticed it later when I analyzed the data). Then, the power really jumped at the 23 minute mark (vertical brown dashed line). I definitely noticed this increase and for the rest of the workout I had to use the up/down arrows in TrainerRoad to adjust the target watts to achieve the desired resistance as measured by the Vectors.

I did a fourth workout tonight where I was only recording the Vector data so I do not have any comparative data to show. However, my experience was similar to Ride C where the data did not quite match up.

I am happy to provide the raw data files if anyone is interested.
Quote Reply

Prev Next