Kingsong S20/S22 (Confirmed)

[supercurio]

59 minutes ago, Tawpie said:

my high school french says: wheel 100 km, gps 96 km

(high school days were pre-disco and it was Québécois, not "French" so I'm likely way off!) 

My native french confirms yours 

Initial charge: 96% (voltage unspecified)

Wheel distance until final titlback: 100.47 km
GPS Distance: 96.43 km
I guess he had the tiltback at 25 km/h - 15.5 mph
Average wheel speed 20.6 km/h - 12.8 mph
Max wheel speed 24.6 km/h

 

To put in perspective, Monsieur Flex also did a similar range test with a 25km/h speed limit (which is the law in France BTW) with the Kingsong S18.

So the S18, with literally half the cells and capacity completed 76 km.

Although the wheels are charged to 124-125V instead of 126V, It gives an idea of the typical energy efficiency of the S20 with its unspecified MOSFET reference & configuration, wide knobby tire compared to the previous Kingsong models.

It does seem unexpectedly low, and seems comparable to previous Kingsong 1500 Wh wheels indeed.

Still hoping to be proven wrong thanks to a 126V charger and better tuned firmware tho.

Edited March 4, 2022 by supercurio

[Seba]

I will make a range test tomorrow. Rider weight (including gear) will be 90 kg, I plan to ride with constant speed between 25 and 30 km/h. Weather forecast says that temperature will be around 0 °C, light wind and no precipitation. Flat terrain. Battery will be charged to 100 % (126 V). Tyre inflated to 1.7 - 1.8 bar. Entire test ride will be public and live on EUC World website, however I don't know exact time I'll start. Just look for the red rider icon in northern Poland  

[Rollin-on-1]

2 hours ago, Tawpie said:

That we know of! V12 didn't have any issues in pre-production either. 

Not entirely correct.  See the video from Hsiang at 13:15.

 

[fbhb]

38 minutes ago, Seba said:

I will make a range test tomorrow. Rider weight (including gear) will be 90 kg, I plan to ride with constant speed between 25 and 30 km/h.

Thanks @Seba for doing this test, as this will definitely put a lot of peoples minds at rest by adding a controlled range test, to those already out there, that can be clearly seen via your own EUC World stats.

I would also like to ask you if you will be able to add BMS monitoring to EUC World for the S20,  along the lines as the fantastic implementation you carried out for the Z10?

 

Edited March 4, 2022 by fbhb

[supercurio]

29 minutes ago, Seba said:

I will make a range test tomorrow. Rider weight (including gear) will be 90 kg, I plan to ride with constant speed between 25 and 30 km/h. Weather forecast says that temperature will be around 0 °C, light wind and no precipitation. Flat terrain. Battery will be charged to 100 % (126 V). Tyre inflated to 1.7 - 1.8 bar. Entire test ride will be public and live on EUC World website, however I don't know exact time I'll start. Just look for the red rider icon in northern Poland  

Haha awesome! Did you get access to the @EUniCycles.eu S20 wheel?

So that's why you were confident about setting the current chargers to 126V: you did it on yours already.
Crazy that they shipped a whole bunch of of uncalibrated chargers, for everyone to run range tests on.

Will there be a reference ride for comparison, like with your 18XL or Inmotion V11?
Same route - same speed.
If you want reasonable repeatable results BTW, setting a titlback is very effective (although makes for a kind of boring ride, unless you're listening for a few podcasts on the way)

[Tawpie]

9 minutes ago, Rollin-on-1 said:

Not entirely correct.  See the video from Hsiang at 13:15.

Thx... I totally missed that. In always 20-20 hindsight, that looks like what happened to some folks with the production wheels (low speed, some torque event, roll over and die, wait for IM to advise).

Edited March 4, 2022 by Tawpie

[supercurio]

11 minutes ago, Tawpie said:

Thx... I totally missed that. In always 20-20 hindsight, that looks like what happened to some folks with the production wheels (low speed, some torque event, roll over and die, wait for IM to advise).

Likewise thanks @Rollin-on-1, it looks like that was the first V12 cut-out (possibly burned MOSFET?) documented on camera.

So yeah most reviewers or shops don't want to throw shade on a product from the initial evaluation batch, therefore we usually don't hear much about these:
That's why the french S20 cut-out mentioned by Adam/Wrong Way had no info, luckily I was still able to extract details (overlean due to confusion between beeps) form people who saw it happen I already been talking to

But hopefully the process can be more transparent, because issues identified by early testers tend to make it towards production models.

Edited March 4, 2022 by supercurio

[Seba]

7 minutes ago, supercurio said:

Haha awesome! Did you get access to the @EUniCycles.eu S20 wheel?

Yup!  

8 minutes ago, supercurio said:

So that's why you were confident about setting the current chargers to 126V: you did it on yours already.
Crazy that they shipped a whole bunch of of uncalibrated chargers, for everyone to run range tests on.

They are getting chargers from external supplier and I suppose there's no additional QC done by King Song. Actually, there's no sticker specifying the manufacturer, although I suspect that it's YZ-Power.

10 minutes ago, supercurio said:

Will there be a reference ride for comparison, like with your 18XL or Inmotion V11?
Same route - same speed.

No, I plan to take a different route, but I'll try to maintain constant speed for most of the ride.

11 minutes ago, supercurio said:

If you want reasonable repeatable results BTW, setting a titlback is very effective (although makes for a kind of boring ride, unless you're listening for a few podcasts on the way)

I'm using EUC World's 1st speed alarm set to "pre-alarm" mode, followed by 2nd speed alarm set 4 km/h above the 1st one. This way I can easily keep the speed in relatively tight range of 26-30 km/h. As prealarm is just a quiet beep without a voice message and it doesn't mute music, it's best config to unobtrusively help me maintain constant speed while listening to my favourite Spotify playlist or audiobook

[Tawpie]

8 minutes ago, Seba said:

I'm using EUC World's 1st speed alarm set to "pre-alarm" mode, followed by 2nd speed alarm set 4 km/h above the 1st one.

What is this "EUC World"? That sounds like a really nice program! 

 

(this is a lame attempt at humor on a Friday, I'm a HUGE fan of EUC World)

Edited March 4, 2022 by Tawpie

[Seba]

25 minutes ago, fbhb said:

I would also like to ask you if you will be able to add BMS monitoring to EUC World for the S20,  along the lines as the fantastic implementation you carried out for the Z10?

I just got S20 for testing today. I'll have few days to test it and also to improve EUC World support for S20. For now I don't know if BMS built into S20 batteries does provide detailed data that can be read over Bluetooth data link.

[Seba]

3 minutes ago, Tawpie said:

What is this "EUC World"? That sounds like a really nice program! 

EUC World? Never heard of her 

4 minutes ago, Tawpie said:

(this is a lame attempt at humor on a Friday, I'm a HUGE fan of EUC World)

Don't worry, I know it well 

[UniBlab]

On 3/3/2022 at 6:21 AM, Paul A said:

OFR, glad it worked, but still think KS using such a strong grade of Loctite is daft.

A lower strength grade of Loctite would probably suffice and make things a lot easier for everyone.

 

Bearing assemblies that 'slot' into position would also be an improvement.

Instead of having to use a blowtorch to heat and expand the surrounding metal to extract the assembly.

Maybe have 'filler' material and make the shape of the outer bearing assembly casing a square, hexagon, some straight sided shape, whatever.

Where heat is not the best option, I have used in the past Brak-Kleen to soften red thread locker - just spray it on and let it soak for a while.  Fastener comes right out.

[supercurio]

46 minutes ago, Seba said:

I just got S20 for testing today. I'll have few days to test it and also to improve EUC World support for S20. For now I don't know if BMS built into S20 batteries does provide detailed data that can be read over Bluetooth data link.

Yes KingSong SoftTuner app can display all 30 voltages for each cell groups, as well as BMS status, alarms, a bunch of temperature sensors - for both packs.
I guess there'll be a lot of BLE packet capture ongoing in the next few days 😉

[Lex Smith]

I'm looking forward to some reviews in a years time when 2nd or 3rd batch wheels are being reviewed.   Until then I will be sticking with my trusty V11.

[Seba]

Just now, supercurio said:

I guess there'll be a lot of BLE packet capture ongoing in the next few days 😉

You're reading my mind!  Thanks you for reminding me about this feature of SoftTuner. I completely forgot about it; I always though that it was supported in KS e-scooters only.

[Miko.cz]

Tip: as S20 was agile in higher speed, dont forget its limited in low/walk speed (180° turn). Enjoy the trip!

[Seba]

Another observation. While charger was adjusted and verified to output exactly 126.0 V (checked with two multimeters that are verified to be precise), motherboard (and then app) reads 125.6 V at the end of charging. While this is inside allowable tolerance, some users may think that there's a problem with the charger or BMS. Please keep in mind that this difference is normal and additionally may vary between wheels. Difference of 0.4 Volt equals to 0.013 V per cell in case of S20. So it translates to cell voltage of ~4.19 V while 4.20 V is considered a fully charged cell.

Edited March 4, 2022 by Seba

[Seba]

3 minutes ago, Miko.cz said:

Tip: as S20 was agile in higher speed, dont forget its limited in low/walk speed (180° turn). Enjoy the trip!

Noted. I'll try to pee in motion 

[supercurio]

32 minutes ago, Seba said:

Another observation. While charger was adjusted and verified to output exactly 126.0 V (checked with two multimeters that are verified to be precise), motherboard (and then app) reads 125.6 V at the end of charging. While this is inside allowable tolerance, some users may think that there's a problem with the charger or BMS. Please keep in mind that this difference is normal and additionally may vary between wheels. Difference of 0.4 Volt equals to 0.013 V per cell in case of S20. So it translates to cell voltage of ~4.19 V while 4.20 V is considered a fully charged cell.

We also observed that the board voltage could be different than the BMS voltages, and on SpeedyFeet unit both BMS measure also differ significantly.
Voltage calibration is 😬 on everything (nothing new for wheels, but this one has many more meters than usual)

edit: Also: the battery % from the wheel is wayyyy off. It's better from the BMS itself tho

Edited March 4, 2022 by supercurio

[Seba]

1 minute ago, supercurio said:

We also observed that the board voltage could be different than the BMS voltages, and on SpeedyFeet unit both BMS measure also significant difference.
Voltage calibration is 😬 on everything (nothing new for wheels, but this one has many meters than usual)

Tomorrow is devoted for range testing, so I won't dig into techical aspects of this S20 unit. Sunday is a demo day, but I plan to at least dig into some S20-specific communication details like logging frames etc. But starting from Monday I will seriously dig deep into this wheel. From my previous experience King Song mainboards were pretty reliable in terms of measurement precision. Battery voltage measurement error was usually well within 0.1 V.

[Chriull]

8 minutes ago, supercurio said:

We also observed that the board voltage could be different than the BMS voltages, and on SpeedyFeet unit both BMS measure also differ significantly.
Voltage calibration is 😬 on everything (nothing new for wheels, but this one has many meters than usual)

Usual accuracies for BMS is in the percent range - ~50mV at 4.2V. This leads to ~1.5V inaccuracy for 30 cells...

5 minutes ago, Seba said:

Battery voltage measurement error was usually well within 0.1 V.

That's some nice calibration work from KS! Would be interesting if this accuracy has some longer time stability which could be not to easy at such accuracies?!

[Miko.cz]

3 hours ago, Seba said:

As we know, chargers shipped with demo wheels are unable to fully charge their batteries. This is because they are incorrectly adjusted (or not adjusted at all) at the factory. If you have a good multimeter, you can adjust it by yourself:

1. ...

Is it OK to measure charger output with no load? And also some protection around charging input (short circuit) could lower Voltage. (speculations only, dont know these PCBs)

16 minutes ago, Seba said:

Noted. I'll try to pee in motion 

Wind, check the wind!

[Seba]

2 minutes ago, Miko.cz said:

Is it OK to measure charger output with no load? And also some protection around charging input (short circuit) could lower Voltage. (speculations only, dont know these PCBs)

Load is not a problem here, because load (thus curent) drops as the battery is near to full charge. But I verified it - charger keeps its 126.0 V in constant-voltage charging phase.

4 minutes ago, Miko.cz said:

Wind, check the wind!

Don't worry, I'm an experienced sailor. I understand perfectly well the difference between windward side and leeward side 

[Paradox]

Hopefully this hasn't been posted before.  Just posted by @Ginger On Wheels:

 

Edited March 4, 2022 by Paradox

[Forwardnbak]

I’m learning a lot about the tech stuff, I just wondered if the bad brake result might come from a power limit on the board similar to the first 3mtr thing but reverse? 
would the power delivery be limited in a similar way when hard braking?