KS-18L/XL range difference between 1.x and 2.x

[Chriull]

9 hours ago, shwinston said:

Just found this thread. I recently started using the app Strava to track my rides. It regularly records speeds 25-33 mph. When I saw the 33 mph I had my doubts as the 18XL is limited to 31 mph. However, I looked at my trip as tracked in Darknessbot and it also recorded 33 mph. This data doesn't jive with 18% speed inflation. Explanations? 

 

8 hours ago, shwinston said:

@Joker10 on Strava and Darknessbot I got reading up to 33 mph. Oddly it was on the same ride. I can’t account for this, but today’s speed test would indicate I was really going about 27 mph. 

"High speed" glitches can happen with GPS - one would have to look at the data in detail. Or easier - just ride some distance constantly at the beginning of tiltback on a track with free vision to the sky and take the average value of the GPS reported speed. This should be quite accurate.

Or one makes an Speed vs. GPS Speed plot as described here

This should average out the "glitches" nicely, too.

[WI_Hedgehog]

On 1/31/2020 at 2:55 AM, Seba said:

KS-18L/XL wheels with 1.x firmware was known to elevate both speed and distance by about 18 %. With 2.0 firmware this changed and the things are more weird now - speed is still elevated by 18%, but distance "only" by about 6 %. So as speed is still inflated by 18% (what makes KS-18L/XL real top speed at 41 kph / 25 mph), distance is much more realistic. This change may create impression that update caused 10 % decrease in range, which is simply not true. It's just change in the way distance is calculated by the firmware.

Users of EUC World can apply -18 % distance correction to KS-18L/XL with 1.x firmware using "General settings" -> "Distance correction" to get real distance. Correction of -6% should be applied to KS-18L/XL updated to 2.0 firmware. Regardless of firmware version -18 % speed correction should be applied for these EUCs regardless of firmware version, but personally I recommend to keep speed uncorrected.

Would it make sense to add a CALIBRATE function to EUC World where it uses GPS readings to figure out real-world speed/distance and apply correction factors to what the EUC is reporting? In theory, this should work across all EUCs equally.

 

[.....]

@Seba Has the newest firmware changed these settings of 18% speed and 6% distance on the 18xl/l models. It may be too soon to tell, I was just curious, as Im finally getting around to setting up alarms and other things I barely comprehend. I only use EUCW and i do prefer REAL data tho it hurts my ego. Once the calibration is set, is that what is reported to the csv logs? Meaning my heads up displays in video will be more correct and I'll seem WAY slower than everyone else? How is this calculated in the records of eucw website? Thanks, apologies if i should have asked this in the eucw thread instead.

[svenomous]

I don't have EUCW (I'm in iOS ecosystem), but @Seba, have you considered an auto-calibration feature that uses GPS (as @WI_Hedgehog suggested above)?  My Garmin bike computer does this on my road bike: it uses a rotation sensor on the bike wheel to count wheel revolutions for speed/distance calculations that are more stable and more accurate (cumulatively) than GPS-derived values, and you can set the wheel diameter in mm to "calibrate" the sensor, but since it also has GPS capability, it offers a feature to auto-calibrate the sensor input, where it compares GPS speed against wheel revolutions and updates the wheel diameter value, on some kind of recurring and automated basis.

As with Garmin, this feature could be an optional on/off setting (vs. the user providing their own correction factors) and when it's on the app could do e.g. a 1-minute calibration occasionally (or when a new wheel is connected and ridden).  Given a wheel that's reporting continuous movement during the calibration run, and given a GPS signal that's reporting acceptable position solution and position updates during the same time period, the app could get a few dozen parallel readings from both, come up with speed and distance values from each, average them to reduce noise, and compute pretty good speed and distance correction factors.  Even non-KS wheels that are more correct must have some error, simply due to small variations in wheel diameter due to tire pressure, rider weight, temperature, and whatnot, so this feature would seem useful for any wheel and any rider who likes to have "perfect" values.  Of course, all the speed alerts, tiltbacks, etc., would continue to have to be entered in "wheel speed" vs. "real speed," since that's what the wheel expects and understands.

[null]

I get about 8% in speed difference on my 18XL FW 2.0, measured on a straight line, over several hundred meters with a GPS app (open terrain). It would be nice indeed if the process was automated.

Edited April 30, 2020 by null

[svenomous]

If @null's speed test can be reproduced and confirmed, and if 18% was the previous difference for 18XL (before 2.0 firmware), that would mean the wheel is now objectively faster.  If the max tiltback ("warp") speed has always been 50kmh as measured by the wheel, and previously 50kmh was really about 42kmh due to an 18% inflation, but now 50kmh is really about 46kmh, it means the wheel can now be ridden 4kmh faster (as measured by GPS or other "real" measurement) than before.  Two 18XLs running flat-out side by side, one with 1.13 and one with 2.02, should show the 2.02 wheel able to go a little faster before tiltback.

[travsformation]

On 4/2/2020 at 9:52 PM, WI_Hedgehog said:

Would it make sense to add a CALIBRATE function to EUC World where it uses GPS readings to figure out real-world speed/distance and apply correction factors to what the EUC is reporting? In theory, this should work across all EUCs equally.

I'm not sure it's worth the work it would involve to implement. GPS accuracy varies too much from phone to phone, whereas KS's overblown speed stats are homogeneously inflated among all wheels on the same FW  

Also, GPS doesn't take carving into account...in any case, for now the wheel-reported speed vs. GPS speed are both visible in EUC World tour graphs as a reference. 

[Chriull]

9 hours ago, travsformation said:

I'm not sure it's worth the work it would involve to implement. GPS accuracy varies too much from phone to phone,

This suould average out while riding?

9 hours ago, travsformation said:

whereas KS's overblown speed stats are homogeneously inflated among all wheels on the same FW  

Not to forget different tires and tire pressures - each petcent in diameter is a % in speed..

9 hours ago, travsformation said:

Also, GPS doesn't take carving into account...

 it's the carvers destroying the speed statistics, not KS! 

[travsformation]

21 minutes ago, Chriull said:

it's the carvers destroying the speed statistics, not KS!

Not as much as the face-planters! 

[Waulnut]

So clarification, maybe it has been stated already, the 18% "speed inflation" means the KS18xl doesn't actually do 31 mph, right? The top speed is 25mph? 

[Chriull]

On 5/1/2020 at 1:21 PM, Chriull said:

This suould average out while riding?

Not to forget different tires and tire pressures - each petcent in diameter is a % in speed..

 it's the carvers destroying the speed statistics, not KS! 

 

On 5/1/2020 at 1:44 PM, travsformation said:

Not as much as the face-planters! 

@Seba - Don't know if and how many truth could be in this jokes? Carving should give quite some deviation between wheel and gps speed? Cornering in cities with not too good gps reception, too? Or some of such "systematic faults" just somehow even out?

You have by accident such numbers too for other wheels (from other manufacturers?).

Seems i'll try once my speed over gps speed plot for logs of driving just a straight track compared to random logs...

[Seba]

6 minutes ago, Chriull said:

@Seba - Don't know if and how many truth could be in this jokes? Carving should give quite some deviation between wheel and gps speed? Cornering in cities with not too good gps reception, too? Or some of such "systematic faults" just somehow even out?

Actually factors like tire pressure or carving doesn't introduce any significant differences to GPS-vs-EUC speed. Tire pressure difference accounts for max. 2-3 % of difference. In theory, carving could add several percents but only if ski-type carving would be involved. Carving within relatively narrow, 1-2 meter wide path won't have any significant impact too.

But we both like to experiment  Let's make one and record two tours with similar speed on the same path (at least two km so GPS errors will cancel out). One while riding just straight, second while carving. I'm curious about the result

[travsformation]

5 hours ago, Seba said:

Actually factors like tire pressure or carving doesn't introduce any significant differences to GPS-vs-EUC speed. Tire pressure difference accounts for max. 2-3 % of difference. In theory, carving could add several percents but only if ski-type carving would be involved. Carving within relatively narrow, 1-2 meter wide path won't have any significant impact too.

But we both like to experiment  Let's make one and record two tours with similar speed on the same path (at least two km so GPS errors will cancel out). One while riding just straight, second while carving. I'm curious about the result

Count me in the experiment!

BTW, on the 16X (with FW 2.02, but was the same with 1.0.7) I'm getting approximately 300m differences between GPS distance and wheel distance on 15 km/tours; that's a difference of 20m/km, which isn't bad at all. Haven't tested on the 18XL but will give it a shot sometime soon

Edited May 6, 2020 by travsformation

[Chriull]

13 hours ago, Seba said:

Actually factors like tire pressure or carving doesn't introduce any significant differences to GPS-vs-EUC speed. Tire pressure difference accounts for max. 2-3 % of difference. In theory, carving could add several percents but only if ski-type carving would be involved. Carving within relatively narrow, 1-2 meter wide path won't have any significant impact too.

But we both like to experiment  Let's make one and record two tours with similar speed on the same path (at least two km so GPS errors will cancel out). One while riding just straight, second while carving. I'm curious about the result

 

7 hours ago, travsformation said:

Count me in the experiment!

BTW, on the 16X (with FW 2.02, but was the same with 1.0.7) I'm getting approximately 300m differences between GPS distance and wheel distance on 15 km/tours; that's a difference of 20m/km, which isn't bad at all. Haven't tested on the 18XL but will give it a shot sometime soon

Did a first small ride (1km on a straight track - was to windy and not so real nice ). Post in 

since this was ridden on a KS16S .

Could be that gps speed is not "the reference" to rely on?

[mozart]

You must have also calibrated the gauges of your cars?

[Waulnut]

Does the speed inflation apply to the ks16x? -18% speed and -6% distance?

[EUC Custom Power-Pads]

2 hours ago, Waulnut said:

Does the speed inflation apply to the ks16x? -18% speed and -6% distance?

My 16X has -5% speed inflation