Kingsong S20/S22 (Confirmed)

[Bizra6ot]

25 minutes ago, Rawnei said:

So are you saying MSP, RS and EXN are weak wheels too since they "only" have 12 mosfets?

No they are 100V that's why I specified for 126v

12 minutes ago, Tawpie said:

I have to chuckle at "more MOSFETS" as having meaning when it comes to range or performance or durability. Having more CPU cores in your computer sounds like it would make it faster, but if they're the wrong kind of CPU or if they're not connected or managed properly by the operating system, they're just a higher parts count.

The thing that matters is which MOSFETs are you using, and how are you using them in the system as a whole. When you attempt to boil down the behavior of a very complex system to MOSFET count, you're demonstrating a lack of understanding of how these systems actually work. It's akin to asserting that 4 push pads will allow you to accelerate faster than 2. It might actually help to have 4, but only if they're designed and installed appropriately.

I am not asserting anything I'm just asking a simple question my friend, if Begode increase the number there is a reason behind it, just like there is a reason for KS to put a limit on the first meters

[Tawpie]

17 minutes ago, Bizra6ot said:

I am not asserting anything I'm just asking a simple question my friend, if Begode increase the number there is a reason behind it, just like there is a reason for KS to put a limit on the first meters

Fair enough. 🕊

I'm just saying that simply counting MOSFETs doesn't allow us to infer higher acceleration capability or improved durability because we don't know how the system is designed. As you point out, we don't know why they do what they do.

Edited March 10, 2022 by Tawpie

[Paul A]

Are there actually any serious problems with the S20?

Reviews thus far seem to be quite positive.

[RagingGrandpa]

43 minutes ago, Bizra6ot said:

if Begode increase the number there is a reason behind it, just like there is a reason for KS to put a limit

It was a fair question, no harm in asking.

All EUC's need at least 6 power transistors, to drive 3 motor phases, with each phase (wire) being able to connect to + or - at different motor positions. So: 6 is the minimum.

If we put two FETs in parallel, in each of those same 6 positions, we double the maximum current that a controller can supply. It just takes up more room and costs more money. 
To achieve the same goal, perhaps we could have just used 6 larger (and more-expensive) individual FETs, rather than using 12 cheaper ones. A tradeoff every company must make, based on cost and availability.
There are very few individual FETs on the market that could withstand the currents we want, so "two in parallel" is the norm today.

Monster Pro was a very-high-power EUC (high speed = high wind drag), and while it didn't really need additional motor current, its controller was very stressed from a thermal standpoint. So, they added another parallel FET, making it 6x3 = 18 chips total. This reduces the resistance and heat dissipation, and spreads the heat across a larger area.
(Pity, the heatsinking still wasn't very good, and Monster Pro will overheat riding on flat pavement at 45mph for more than 5 minutes- a ride that's accomplished by Sherman rather easily, using just 12 of the same chips. Only exceptional daredevil riders sustain such speeds for longer than a few seconds, so the MonPro is still "good enough.")

We don't have details for 134V Master yet- only that it has 24 FETs.
Perhaps it's 2-parallel, 2-series, times six, of the same-old Hooyi 125V FET everyone's been using. The FETs in series effectively doubles their voltage rating, enough that 134V can be reliable.
Or, it could be 4-parallel of some other chip which can tolerate 150V. To be seen.

(Hopefully this helps explain the 'reasons behind it' )

 

Edited March 10, 2022 by RagingGrandpa

[Tawpie]

6 minutes ago, RagingGrandpa said:

All EUC's need at least 6 FETs, to drive 3 motor phases, with each phase (wire) being able to connect to + or - at different motor positions. So: 6 is the minimum.

Aren't these H-bridge motor drivers? So you can 'drive' the motor in either direction? Pleading ignorance here.

Edited March 10, 2022 by Tawpie

[chanman]

Would love to know the actual mosfets used in the production S20s, but until we have some shipped and a tear down video I doubt we will get that. What I've seen so far doesn't look like a wheel with twice the power capabilities of an S18, which is what I wanted to see, and I'd like to know if this is a firmware (likely to be made less conservative over time) or a hardware issue which will not without an upgraded board. 

Not sure if there will be a good fix for the weight distribution and turning feel, maybe a mediumish riding mode will help with that and an aftermarket more street oriented tire.

[RagingGrandpa]

On 3/10/2022 at 2:20 PM, Tawpie said:

Aren't these H-bridge motor drivers? So you can 'drive' the motor in either direction? Pleading ignorance here.

Close.

An H-bridge drives two phases, and provides a way to reverse the direction of current flow, as you said. 
Two phases are typical for a brush-type DC motor. 
An H-bridge is made up of two halves, which I like to call "half-bridges." Each half-bridge being two transistors. 
If you draw the resulting circuit, it looks like the letter H.

 

Instead, we need to drive three phases for our fancy brushless motor in the EUC. 
And we still need to be able to reverse the current. 
So, we use 3 "half-bridges." Resulting in a 3-phase motor driver (but not an "H" shape anymore).

 

@mrelwood

Edited March 25, 2022 by RagingGrandpa

[Marc Roberts]

37 minutes ago, Paul A said:

Are there actually any serious problems with the S20?

Reviews thus far seem to be quite positive.

Mainly it seems that power delivery is being throttled at low speed. I’ve not seen a definitive conclusion as to why yet, or if it will change with subsequent firmware or the production hardware. Lots of videos seem to show that it’s quite easy to over torque the wheel at low speeds.

This guy seems to think it’s by design so you don’t spin out when off-road with instant power delivery. I don’t buy that.

 

Edited March 10, 2022 by Marc Roberts

[Paul A]

Ok, thanks.

Not sure that the throttled acceleration over the first 3 meters or so is a major problem though.

 

[Rawnei]

1 hour ago, Bizra6ot said:

No they are 100V that's why I specified for 126v

And that is different how?

[Marc Roberts]

19 minutes ago, Paul A said:

Ok, thanks.

Not sure that the throttled acceleration over the first 3 meters or so is a major problem though.

 

I’m not sure if the “first 3 meters” thing isn’t a slight red herring TBH, just that’s roughly how much distance it takes to get up to speed to bring up the power delivery. If you stay slow it sounds like you’ll stay in the lower end of the power zone? Or I may have missed something and be totally wrong 😳

For me the worry is cutting out trying to do some slow uphill rooty stuff or something like that.

*edit*
That said, I’ll be doing similar stuff to Ian at Speedyfeet, and in 650km he doesn’t seem to have struggled. Jury is still out for me.

Edited March 10, 2022 by Marc Roberts

[rolis]

51 minutes ago, RagingGrandpa said:

Perhaps it's 2-parallel, 2-series, times six, of the same-old Hooyi 125V FET everyone's been using. The FETs in series effectively doubles their voltage rating, enough that 134V can be reliable.

If anybody is bored and wants to know how MOSFETs are being put in Series:
https://resources.altium.com/p/should-you-use-power-mosfets-series

[Freeforester]

Adam’s observations:

 

 

[chanman]

Looks like similar issues with the suspension in the teardown that we saw with the batch 1 S18s. Really hoping the first batch will ship out with improvements and we won't have to wait for batch 3 or 4.

Nice close up of the main board in the video, wasn't able to discern anything about the mosfets from it though, maybe someone else can.

Edited March 10, 2022 by chanman

[Tawpie]

Nice teardown! Doesn't look like early batches will be "low maintenance" though. They're serious when they say to service the suspension monthly… my off-road riding tends to be pretty dusty so I'll be in there fairly frequently (unless you can flush the sliders from the top).

It's a bit sad that this will probably want some disassembly, check and correct upon receipt (mostly to ensure there's blue locktite on the pedal hanger screws that Kevin couldn't get undone, and perhaps to add a dab or two of silicon here and there). It might be a decade or three before a new EUC will be like my car... jump in and drive it for 130k miles without doing anything except replace oil and tires.

Darn. I wanted to see the MOSFET part numbers and really wanted to get the dimensions of the spring so I can start sourcing a lighter one. I do understand the reluctance to remove the control board to see the MOSFET numbers though, until you know exactly how it's put together you don't want to risk undoing something that is difficult to redo (like thermal goo etc.)

One great thing about the over-hype on this wheel is that it's getting a close look from a lot of people. Thank you to all of the reviewers for your efforts, and I hope the YT revenue at least buys you one large bucket of beer!

Edited March 10, 2022 by Tawpie

[techyiam]

Well, the amount of stiction in the slider probably is inherent in the design. They are using polymer bushings. And it doesn't appear there is anything special about those bushings, or the tracks in the battery box. A cheap kid suspension bike has less stiction and play. But it may be good enough for some. It is as low cost as you can get. I was hoping that it would work better, but Wrong Way's video showed its true nature. Unfortunately, it is not the magic bullet for suspension I was hoping for.

Uncharacteristic low end performance was the first strike. Cheap suspension design is the second strike for me. However, the S20's suspension still eats potholes.

I definitely have to wait and see how the Master will compare. There is certainty room for other wheels to edge it out.

Edited March 11, 2022 by techyiam

[Rehab1]

1 hour ago, Freeforester said:

Adam’s observations:

Excellent teardown!

The amount of play I see between the suspension arms and bushings is very disappointing! It’s like KS is intermixing  metric with SAE components.

I ripped apart and rebuilt the suspension on my S18 because I had no choice. If KS and their engineers can’t tighten up the dimensioning and tolerances of the S20 I do have a choice. 

[techyiam]

Good news is that Wrong Way did slide out the wheel out from the bottom. That is a big win, IMO. However, he said that he only use that method, and "not the right way", is because he was having difficulties removing the screws, and that it would be much too difficult if he used the conventional method. Not sure why the quick method is not proper.

[Paradox]

Interesting WW said he mounted the foot plates backwards and it improved the ride.  Pushes the COG forward a bit maybe to counter the heavier rear?  He did not actually say backwards he said they mounted them the wrong way.  No explanation.  He says a follow up video will follow.

Video pasted at relevant time.

 

Edited March 11, 2022 by Paradox
clarification

[Paradox]

In the first picture the pedal is offset toward the rear of the wheel and the second shows the offset toward the front.  They must have just switched right and left pedals?

Edited March 11, 2022 by Paradox
orientation of photos.

[Whalesmash]

Yikes, those plastic sliders... That's going to get real gritty after some dirt riding. I don't even know if that design is salvageable with something like PTFE lube the way the S18's is. Between that and the mosfet problem, looks like KS has some work to iterate on...

[techyiam]

I am wondering whether Kingsong tried too hard to keep the MSRP relatively low? On the one hand, you have a smart BMS, and an undocumented quick wheel removal feature, but on the other hand, you have throttled-back starts from rest, cheap plastic front bumper and kickstand pad, and junkie suspension sliders and links' joints. And this is 2nd gen suspension design. The promotional material says the the wheel is rated at 3300 W, with a peak of 7500 W peak. The battery specs has a rated voltage of 111 V, and a peak charging voltage of 126 V. But so far, no one seemed to be wowed by the wheel's performance. What a shame.

Edited March 11, 2022 by techyiam

[Wolverine]

8 hours ago, chanman said:

Really hoping the first batch will ship out with improvements and we won't have to wait for batch 3 or 4.

Let's not forget that only pre-production models are currently distributed to certain people. The first batch comes with improvements, but not all improvement are included as it would delay the launch of the S20 even more. Batch 1 wheels will arrive in early summer, and no one wants to wait anymore. It is recommended to wait for batch 3 or 4. However, if you are ready to get batch 1, be prepared to make fixes on your own.

[techyiam]

30 minutes ago, Wolverine said:

It is recommended to wait for batch 3 or 4. However, if you are ready to get batch 1, be prepared to make fixes on your own.

Or, for those who don't have to have a S20, and have the luxury to wait a bit more, it may be prudent to wait and see how the competitors will stack up. There are just too many big items that are worrisome, and have to struggle with during ownership. I guess a bad experience can do that to you.

[Greg X]

ks s20 is definitelly NOT worth to buy in this season, especially for such high price. Low quality, and too many serious issues with basic elements. 

At least if someone doesnt like to be "beta tester" and pay full price for this privilege  

It may be nice wheel in 2023, when KS will refine the structure, but not yet.

 

Suspension is very nice idea, but it is so underdeveloped and made "cheap way", that euc without suspension still stay much safer choice. 

Edited March 11, 2022 by Greg X