New GotWay Nikola 17''

[EcoDrift]

https://ecodrift.ru/2018/12/29/gotway-nikola-part-2-close-look-first-try-english-version/

[meepmeepmayer]

38 minutes ago, EcoDrift said:

https://ecodrift.ru/2018/12/29/gotway-nikola-part-2-close-look-first-try-english-version/

Fantastico! We don't even have to use Google Translate now.

(edit: Non Google Translate English link https://ecodrift.ru/2018/12/29/gotway-nikola-disassembling-english-version/)

Nikola disassembly article by EcoDrift!! https://translate.google.com/translate?sl=ru&tl=en&u=https%3A%2F%2Fecodrift.ru%2F2018%2F12%2F29%2Fgotway-nikola-razbiraem-chast-3%2F

Edited December 29, 2018 by meepmeepmayer

[meepmeepmayer]

Reading... well, it's a Gotway. Overall nice and clean (and so much wasted space), but it's hard not to laugh or cry sometimes

Quote: "To prevent the controller from dangling in the bay, Gotway came up with this solution:"

lol

[meepmeepmayer]

Eww the boards now have smaller mosfets!!! NOOOOOOOOOOOO.

The one thing they got so right... they got wrong again.

[Nic]

I don't like how GW use silicone on the connectors ... just fit some better quality connectors that can be easily removed!

Edited December 29, 2018 by Nic

[EcoDrift]

https://ecodrift.ru/2018/12/29/gotway-nikola-disassembling-english-version/

[meepmeepmayer]

Fantastic articles, thank you!

-

Overall, I'm disappointed by the interior. New design, old questionable quality. Smaller mosfets. No water-proofing to be seen, though it may not be needed to too much. Board better get attached properly in the production wheel. No cable channels. Huge shell with badly used and usable space. OK engineering but nothing exciting. Mix of screws. Could need another iteration (or 2 or 5) before production.

Summary: it's a Gotway

[RockyTop]

8 minutes ago, Nic said:

I don't like how GW use silicone on the connectors ... just fit some better quality connectors that can be easily removed!

I agree. I hate working with it!

   They seem to use it a lot, but it is a common practice on encoder sensor connections. (calculates absolute position "Halifax sensors" ) Because of the low voltage signal any vibration that causes an AM signal  (friction/movement in the connector)  can cause havoc. 

Given the importance of the connections I prefer a great connector and the silicone. 

[Marty Backe]

1 hour ago, EcoDrift said:

https://ecodrift.ru/2018/12/29/gotway-nikola-part-2-close-look-first-try-english-version/

Fantastic articles and pictures.

I love your periodic use of "no comments on the quality"   I guess we all know what to expect from Gotway.

As @meepmeepmayer says, there's so much wasted space in the shell. I bet they could make a 2400wh version of this wheel. Can you imagine? 

[Nic]

3 hours ago, RockyTop said:

I agree. I hate working with it!

   They seem to use it a lot, but it is a common practice on encoder sensor connections. (calculates absolute position "Halifax sensors" ) Because of the low voltage signal any vibration that causes an AM signal  (friction/movement in the connector)  can cause havoc. 

Given the importance of the connections I prefer a great connector and the silicone. 

My V8 doesn't have silicone on any connectors and its a reliable wheel.

[Jon Stern]

3 hours ago, RockyTop said:

I agree. I hate working with it!

   They seem to use it a lot, but it is a common practice on encoder sensor connections. (calculates absolute position "Halifax sensors" ) Because of the low voltage signal any vibration that causes an AM signal  (friction/movement in the connector)  can cause havoc. 

Given the importance of the connections I prefer a great connector and the silicone. 

Do you mean Hall Effect sensors?

Hall Effect sensors are used to determine an electromagnet actuator/motor position.

[RockyTop]

2 minutes ago, Jon Stern said:

Do you mean Hall Effect sensors?

Hall Effect sensors are used to determine an electromagnet actuator/motor position.

Yes, The hall effect sensor. thanks Siri

Hall effect sensors are sensors that tell you if two thing are close to each other. They are used in position encoders to determine an electromagnet actuator/motor position. In the larger world calling a position encoder a hall effect is like calling a key board a switch. In my trade (Large motor controllers ) we use tachs (for general speed on older units) and position encoders and absolute position encoders for motor position. Some use optics instead of hall effects.

[meepmeepmayer]

4 hours ago, Marty Backe said:

As @meepmeepmayer says, there's so much wasted space in the shell. I bet they could make a 2400wh version of this wheel. Can you imagine? 

2400Wh, pff. You could essentially do 2600Wh/3200Wh version (2 packs each side) with that space. And the board space, the only apt description is "board cave". First I was wondering if they planned a newer, physically bigger board, but that doesn't make too much sense. Must just be a noob designing the thing. I like me some overengineering, but not like this. If they at least had used all that space for cable channels instead of the usual tangled mess of wires.

Oh well, it's shiny on the outside, works well enough like all the Gotways, is a dream for wheel modders who want to build whatever into their wheel, and of course will ride flawlessly like a Gotway does. Looking forward to your impressions.

Still, with this, I feel like GW is falling behind KS, their advantage got smaller (I'm really salty about those new, smaller mosfets). On the other hand, the wheels are hardly comparable.

Edited December 29, 2018 by meepmeepmayer

[Marty Backe]

2 hours ago, meepmeepmayer said:

2400Wh, pff. You could essentially do 2600Wh/3200Wh version (2 packs each side) with that space. And the board space, the only apt description is "board cave". First I was wondering if they planned a newer, physically bigger board, but that doesn't make too much sense. Must just be a noob designing the thing. I like me some overengineering, but not like this. If they at least had used all that space for cable channels instead of the usual tangled mess of wires.

Oh well, it's shiny on the outside, works well enough like all the Gotways, is a dream for wheel modders who want to build whatever into their wheel, and of course will ride flawlessly like a Gotway does. Looking forward to your impressions.

Still, with this, I feel like GW is falling behind KS, their advantage got smaller (I'm really salty about those new, smaller mosfets). On the other hand, the wheels are hardly comparable.

We (I do at any rate) often complain about lack of engineering by Gotway, so maybe we shouldn't dis their fall-back to smaller MOSFETs. Maybe the improvement in the firmware really is allowing them to reduce the requirements of the power transistors? Maybe we are actually seeing the result of better engineering.

[EUCGUY]

On 12/19/2018 at 4:54 AM, houseofjob said:

 

https://ecodrift.ru/2018/12/18/gotway_nicola_pervaya_vstrecha/

 

Gotway Nicola first meeting in the cold snowy Moscow

Today, the new Gotway Nicola arrived at our warehouse. From a technical point of view - this is the same Tesla, but under a large battery and with a wide tire 3 ″ wide. Tomorrow we will show a more detailed photo report, but for now just roll Nicola in the evening Moscow snow.

A bit of a strange color with white inserts and translucent overlays through which you can see the inner contents of the wheel.

... ..

Instead of the rear side light - GW logo. 

During movement the tape is poured in various colors.

Headlamp - standard, like all Gotway last year and a half.

The road is visible, but not far away.

Snowing.

On the case, Nicola shows the current voltage.

Inside a lot of empty space.

In our case, version 1600Wh. Although in fact it is 1554Wh if you count 3.7V for the rated voltage.

Pay attention to the angle of the pedals:

The blue LEDs are lit on the controller:

A pen:

The controversial decision and lowish in height:

The rubber covers the access to the screw that holds the cover:

According to the plan tomorrow, we will shoot the wheel in more detail and disassemble Nicola.

did they implement a voltage meter at the front?

Clear plastic like this can be very nice if done correctly, here i think it looks cheap.

[houseofjob]

1 hour ago, EUC GUY said:

did they implement a voltage meter at the front?

Yes, in response to all the Korean enthusiasts that do voltage meter mods on their wheels.

[meepmeepmayer]

10 hours ago, Marty Backe said:

We (I do at any rate) often complain about lack of engineering by Gotway, so maybe we shouldn't dis their fall-back to smaller MOSFETs. Maybe the improvement in the firmware really is allowing them to reduce the requirements of the power transistors? Maybe we are actually seeing the result of better engineering.

It's either this, or they are trying to save money/copy KingSong and the other manufacturers. My money is on the latter. What was wrong with the big mosfets? It's not like the required extreme stress currents get less because of some new motor control algorithm. Gotway always did overengineering vs. being "smart" like the other manufacturers, and I strongly prefer that overengineering. This seems to be their (first?) attempt at being "smart" and I don't like it (and don't trust them to do it competently). They went from being ahead of the curve (big mosfet boards) to "with this, I'd rather trust KingSong" for me.

I'm not going to like those new boards until @EUC GUY does another car push test and it works at least as well. With the MSX style board, they literally eliminated the mosfets as weak point, the cabling had to melt for the wheel to fail! How amazing is that? Doesn't get better than this! (Well, except thicker cabling, but something has to give.) Will the new mosfets be as good? Can they? I'm worried this might be a regression or another "We didn't expect our wheels to be used like this" situation.

If they want to do better engineering, they should do away with the electrically nonsensical 12 parallel mosfets and should do it however it's supposed to be done. But not try and skimp on parts.

Their kind of design and overengineering was their big advantage, now they're budget KS technology with disabled speed limit in my eyes. In other words, with this new board, what are they doing better than KingSong, technology-wise? I have no answer.

Don't get me wrong, I'm happy GW are experimenting like this instead of being slow as mollasses like everyone else. Just not sure I like this particular direction.

Edited December 30, 2018 by meepmeepmayer

[Nic]

1 hour ago, meepmeepmayer said:

It's either this, or they are trying to save money/copy KingSong and the other manufacturers. My money is on the latter. What was wrong with the big mosfets?

Looks fine to me ... from what I can see the MSuper v3 had 6 Mosfets, the MSX upgraded to 12 Mosfets, and the Nikola also has 12 Mosfets.

 

MSuper v3 with 6 Mosfets vs MSX with 12 Mosfets...

 

Nikola has 12 Mosfets (looks similar to MSX board) ...

Quote

The motherboard is exactly the same to the one in MSuperX. The only difference is that the MOSFETs now have a small body. I’ve asked the Gotway team why was this unit changed. They told me that they had significantly improved the motor control program. The mainboard performance has become more efficient and that made enhanced MOSFET heatsink unnecessary. They will no longer heat up that much even on the MSuperX. So it’s quite possible that the new mainboards for the MSX will also get that type of MOSFETs.

 

Edited December 30, 2018 by Nic

[Scatcat]

11 hours ago, Marty Backe said:

We (I do at any rate) often complain about lack of engineering by Gotway, so maybe we shouldn't dis their fall-back to smaller MOSFETs. Maybe the improvement in the firmware really is allowing them to reduce the requirements of the power transistors? Maybe we are actually seeing the result of better engineering.

Or maybe they just try to save money and board footprint? Good firmware and all that, it's not as we're even close to the point where EUCs have a reasonable reliability factor. One that means we can push them to their "safe" limit in any and all situations for years on end, without mechanical or electrical failures.

Using 247's would mean a little more physical headroom before any glitches in firmware or extreme riding conditions f*cks things up royally. Unless the manufacturers manage to create truly redundant systems, where failure in one set of circuits means switching over to secondary circuits rather than a cutoff, we cannot have "too much" headroom in the design.

Go for 247s or better, go for redundant gyros, go for the fattest wiring that fits into the design. That is the way forward for the next generation.

[meepmeepmayer]

17 minutes ago, Nic said:

Looks fine to me ... from what I can see the MSuper v3 had 6 Mosfets, the MSX upgraded to 12 Mosfets, and the Nikola also has 12 Mosfets and a much larger Heatsink that they attach to.

ms3 has 12, too, they're directly opposite on the other side of the board. Only the very early 67V GWs had 6 before they figured out it doesn't work with the weak ones they were using.

History of GW mosfets is like this: shitty ones (ACM, ms3), decent ones (Tesla, Tesla-ized ACM2 and ms3 1900W) which are about the same quality as we have on any other brands, and then the big fat godlike TO-247s on the MSX/MCM5/100VMonster style boards.

Nikola seems to be a regress to Tesla level mosfets, just with better cooling/positioning.

18 minutes ago, Scatcat said:

One that means we can push them to their "safe" limit in any and all situations for years on end, without mechanical or electrical failures.

Exactly! The TO-247s were a step in this direction. Beef up the wiring, and monitor the mosfet temperature as (incredibly strong) weak point. 99% done.

[meepmeepmayer]

22 minutes ago, Scatcat said:

Or maybe they just try to save money and board footprint? Good firmware and all that, it's not as we're even close to the point where EUCs have a reasonable reliability factor. One that means we can push them to their "safe" limit in any and all situations for years on end, without mechanical or electrical failures.

Using 247's would mean a little more physical headroom before any glitches in firmware or extreme riding conditions f*cks things up royally. Unless the manufacturers manage to create truly redundant systems, where failure in one set of circuits means switching over to secondary circuits rather than a cutoff, we cannot have "too much" headroom in the design.

Go for 247s or better, go for redundant gyros, go for the fattest wiring that fits into the design. That is the way forward for the next generation.

This sums up my view so well!

EUCs are not something where you see with how little you can get away with using "smart" engineering. Give me ridiculously oversized electronic parts for ridiculous reliability in any conceivable edge case.

[Scatcat]

6 hours ago, meepmeepmayer said:

This sums up my view so well!

EUCs are not something where you see with how little you can get away with using "smart" engineering. Give me ridiculously oversized electronic parts for ridiculous reliability in any conceivable edge case.

Yeah, that was the reason I was so happy when the MSX came out. For once it seemed Gotway had taken the important step needed for the future. Both TO-247s and separation of MOSFETs rather than stacking them two and two on top of each other. And look at the failure rates of this generation of EUCs, MSX is at the very bottom with almost no failures whatsoever.

That's the one thing that always impressed me with the GT16, that they had gone with 247s from the start.

Go with smart firmware, that is all well and good, but not to "mask" the fact that you skimp on the hardware. The hardware must be as good as possible anyway, or there will be failures.

[Marty Backe]

43 minutes ago, Scatcat said:

Yeah, that was the reason I was so happy when the MSX came out. For once it seemed Gotway had taken the important step needed for the future. Both TO-247s and separation of MOSFETs rather than stacking them two and two on top of each other. And look at the failure rates of this generation of EUCs, MSX is at the very bottom with almost no failures whatsoever.

That's the one thing that always impressed me with the GT16, that they had gone with 247s from the start.

Go with smart firmware, that is all well and good, but not to "mask" the fact that you skimp on the hardware. The hardware must be as good as possible anyway, or there will be failures.

I have no engineering insight to Gotway's solution, but to say go with the bigger MOSFETs "just because", feels a bit like a religious argument. Unless we are provided with the schematics of their circuit design we don't have much basis for an argument, unless the new wheel starts burning up control boards.

We just need to find some Gotway early adopters 

[ED209]

3 minutes ago, Marty Backe said:

We just need to find some Gotway early adopters 

@meepmeepmayer? 😊

[meepmeepmayer]

2 hours ago, ED209 said:

@meepmeepmayer? 😊

The only reason I'm not doing kill tests left and right is financial. It would be my personal hobby otherwise. I'm serious. Force a wheel up a crazy mountain, see what happens. Repeat until I'm happy or it dies. Any wheel.

Because that's literally what I want from a EUC - ride up any mountain and either it works with no problems, or it doesn't work in a good manner (alarms). By proxy, this is how EUC electronics in general should always work.

Not interested in "works in 99% of situations" wheels. Especially if you can't know beforehand that something does not work..

3 hours ago, Scatcat said:

That's the one thing that always impressed me with the GT16, that they had gone with 247s from the start.

Really? Never realized/heard of that. That's amazing.

Here it is. Cabling is quite thin, though. But big mosfets.