Dear Gotway, Please Consider These Upgrades

[Goodman]

55 minutes ago, Marty Backe said:

That's good to know. I might use mine once a week, maybe.

On Facebook or here, just recently someone had their MSuper opened and posted some pictures which included what looked like padding that was inside the handle area. It's easy to open the MSuper side panel. Maybe you should open yours and see what has changed inside. It's not very complex. You could probably re-secure the handle mechanism fairly easily.

Good idea since I have a long weekend I will give it a go.

[US69]

11 hours ago, Marty Backe said:

 

12 hours ago, KingSong69 said:

its just a buzzer...i have to replace it 2 times now on my MSuper V3 

Wow, what kind of abuse do you give your V3?

Abuse? Which abuse? The first buzzer died a slow one....getting quieter and quieter every day powering....saying goodbye with a "squeeeeeeek"

The second decided to say goodbye from one power off to the next power on...

 

Now my third is one i received from 1radwerkstatt....seams to hold on and is a bit louder

[Marty Backe]

5 hours ago, KingSong69 said:

Abuse? Which abuse? The first buzzer died a slow one....getting quieter and quieter every day powering....saying goodbye with a "squeeeeeeek"

The second decided to say goodbye from one power off to the next power on...

 

Now my third is one i received from 1radwerkstatt....seams to hold on and is a bit louder

I was just having fun with you 

[KalSeth]

Marty and others asking for it:

Why do you want bluetooth speakers on the wheel?  What do they accomplish?

I Would like a volume control on my wheel beeps, mostly down not up as they are too loud as it is.  Bluetooth speakers just add expense and weight and use battery that could be range so why do you want them?  For playing music? 

Many things would be nice to have but think of how much the cost is. If they did have speakers and other whizzbang add ons I would prefer to see those as optional equipment I could choose to skip. 

[dmethvin]

4 hours ago, KalSeth said:

Why do you want bluetooth speakers on the wheel?  What do they accomplish?

Many things would be nice to have but think of how much the cost is. If they did have speakers and other whizzbang add ons I would prefer to see those as optional equipment I could choose to skip. 

Since the speakers already need to be there to give the warning beeps, there's no cost savings there. The bluetooth needs to be there for the app, so no costs to shave there either. I suspect this is one of those "we might as well" features. I have it on my KS14C but haven't used it because I don't want to bother people as I ride by. I can see how it might be nice though. In Maryland you're not allowed to ride/drive with earbuds so if you want to listen to music the speakers would be handy.

[Marty Backe]

4 hours ago, KalSeth said:

Marty and others asking for it:

Why do you want bluetooth speakers on the wheel?  What do they accomplish?

I Would like a volume control on my wheel beeps, mostly down not up as they are too loud as it is.  Bluetooth speakers just add expense and weight and use battery that could be range so why do you want them?  For playing music? 

Many things would be nice to have but think of how much the cost is. If they did have speakers and other whizzbang add ons I would prefer to see those as optional equipment I could choose to skip. 

If you really ride fast you'll know that it can be hard to hear the beeping over the wind noise in your ears.

KingSong has included speakers in their wheels for, like, forever. Doesn't seem to drive up the cost of their wheels, and when you consider these new wheels have such large capacity wheels, running the speakers is going to be a blip in the total energy usage of the wheel.

Sometimes it can be very enjoyable to ride with music streaming from your wheel. It seems like a gimmick until you've experienced it yourself.

[Marty Backe]

2 minutes ago, dmethvin said:

Since the speakers already need to be there to give the warning beeps, there's no cost savings there. The bluetooth needs to be there for the app, so no costs to shave there either. I suspect this is one of those "we might as well" features. I have it on my KS14C but haven't used it because I don't want to bother people as I ride by. I can see how it might be nice though. In Maryland you're not allowed to ride/drive with earbuds so if you want to listen to music the speakers would be handy.

Just to clarify, the warning beeps come from a simple piezoelectric speaker that's incapable of playing music. All they do is generate a single tone sound.

[Jason McNeil]

@Marty Backe great write-up! This forum is the most important organ for vendor change. Look at the response (after two years !! ) when 'Hello King Song' was complained about recently on the forum: there was acknowledgement, & finally capitulation on this pointless most unpopular 'feature'

There's typically a variety of reasons why manufacturers are reluctant to change; it might be the cost to implement a certain feature, standardized parts need to be completely rethought, not enough manpower (GW's weakness), concerns that the volumes bought will not offset the development or tooling costs, insufficient insight or expertise to make a judgement call, & my particular bug-bear, failure to admit that mistakes were made—a particular problem at one company. 

It drives me crazy on those occasions where you've compiled what should be a complete waterproof case: you spend the time to gather the evidence, present the data so that there could be no possible disagreement about the facts, & the response is some lame "we can't do that, because, basically, we don't want to". The Wheeling World needs competition, at least three brands, this is the Prime motivator towards any sort of positive change. 

Gotway have introduced a lot of innovations, especially recently, it's easy to forget who did what first & when. With the help of my Gotway contacts, I put together an infographic timeline of some of the GW firsts. Going to try to do the same with KS & Inmotion. It would be wonderful if these companies could get more analytical on incorporating Customer feedback—especially considered & experienced ones like Marty—features & capabilities during the design phase, rather than presenting releases as a fait accompli 

[meepmeepmayer]

Very nice graphic. You do like your infographics, and we certainly don't mind

I also like the overall insight into your frustration. Essentially, "BUT WHY?" at so many incomprehensible and unnecessary (or even unprofessional?) behaviors.

32 minutes ago, Jason McNeil said:

not enough manpower (GW's weakness)

Interesting...

32 minutes ago, Jason McNeil said:

my particular bug-bear, failure to admit that mistakes were made—a particular problem at one company. 

You can't tell but I wish you could. Not even guessing in order to prevent a public guessing game. But very interesting to hear. Might be a Chinese saving face thing.

[Marty Backe]

56 minutes ago, Jason McNeil said:

@Marty Backe great write-up! This forum is the most important organ for vendor change. Look at the response (after two years !! ) when 'Hello King Song' was complained about recently on the forum: there was acknowledgement, & finally capitulation on this pointless most unpopular 'feature'

There's typically a variety of reasons why manufacturers are reluctant to change; it might be the cost to implement a certain feature, standardized parts need to be completely rethought, not enough manpower (GW's weakness), concerns that the volumes bought will not offset the development or tooling costs, insufficient insight or expertise to make a judgement call, & my particular bug-bear, failure to admit that mistakes were made—a particular problem at one company. 

It drives me crazy on those occasions where you've compiled what should be a complete waterproof case: you spend the time to gather the evidence, present the data so that there could be no possible disagreement about the facts, & the response is some lame "we can't do that, because, basically, we don't want to". The Wheeling World needs competition, at least three brands, this is the Prime motivator towards any sort of positive change. 

Gotway have introduced a lot of innovations, especially recently, it's easy to forget who did what first & when. With the help of my Gotway contacts, I put together an infographic timeline of some of the GW firsts. Going to try to do the same with KS & Inmotion. It would be wonderful if these companies could get more analytical on incorporating Customer feedback—especially considered & experienced ones like Marty—features & capabilities during the design phase, rather than presenting releases as a fait accompli 

Fantastic. As someone who didn't know what an EUC was prior to August 2016 it's interesting to see the innovations attributable to Gotway. This industrial is so incredibly young. Just look at the changes in 3 years.

I didn't realize that they invented the phone app for the wheel. I guess I should give them slack with regards to its quality considering that we might not have the apps without Gotway.

I know that you're agitator for positive changes. I have to hope that with time and increased sales (not just with you) that we will see continued improvements.

I recognize that I was a little over-the-top in praise of Gotway, but I actually do hope that someday there's some serious competition within the performance wheel category. That may be the only thing that drives some of the changes we want.

Look forward to your education on the KingSong front 

[Marty Backe]

1 hour ago, Jason McNeil said:

@Marty Backe great write-up! This forum is the most important organ for vendor change. Look at the response (after two years !! ) when 'Hello King Song' was complained about recently on the forum: there was acknowledgement, & finally capitulation on this pointless most unpopular 'feature'

There's typically a variety of reasons why manufacturers are reluctant to change; it might be the cost to implement a certain feature, standardized parts need to be completely rethought, not enough manpower (GW's weakness), concerns that the volumes bought will not offset the development or tooling costs, insufficient insight or expertise to make a judgement call, & my particular bug-bear, failure to admit that mistakes were made—a particular problem at one company. 

It drives me crazy on those occasions where you've compiled what should be a complete waterproof case: you spend the time to gather the evidence, present the data so that there could be no possible disagreement about the facts, & the response is some lame "we can't do that, because, basically, we don't want to". The Wheeling World needs competition, at least three brands, this is the Prime motivator towards any sort of positive change. 

Gotway have introduced a lot of innovations, especially recently, it's easy to forget who did what first & when. With the help of my Gotway contacts, I put together an infographic timeline of some of the GW firsts. Going to try to do the same with KS & Inmotion. It would be wonderful if these companies could get more analytical on incorporating Customer feedback—especially considered & experienced ones like Marty—features & capabilities during the design phase, rather than presenting releases as a fait accompli 

P.s. I'm not sure about your dates for the ACM and MSuper. I thought the ACM+ (12 mosfets, etc.) came out in mid 2016, and the original ACM was 2015. And the original MSuper came out in September 2016. The s+ models didn't arrive until 2017.

But perhaps all of my dates are based on when they became available in the western world.

[Carlos E Rodriguez]

Dear Gotway. 

 Consider this improvement. hire a real Electrical Engineer before you kill one of us. 

[US69]

very Nice Graphics, but with some more or less questionmarks in the timeline....

wasnt the Firewheel f779 the first high capacity wheel, going 28+???

Yes ACM June 2015.... released in a hurried rush after the popular KS16 arrived April/May, then first with 6 Mosfets, which then failed on their Beta Customers, 2 months later 12 Mosfets.

to call 12 Mosfets an improvement is also questionable. Pure from electrical standpoints it is Nonsens. The right way would have been to implement stronger Mosfets. But Marketing won this battle and 12 Mosfets is seen as an "advantage" by most customers nowadays.

September 2016 was V3....84 Volt Versions about 2-3 months later....v3s+....just know this exactly because i was on the angry side as my 2 monthsV3 was on the "old" side from one day to the other....

@Jason McNeil

nice graphics, Jason, don't get me wrong on this "little" criticism above....but partly it looks like written from an GW employee, nicing the things a little bit to much up ;-)

you will also do this for other companies?

[esaj]

9 hours ago, Jason McNeil said:

@Marty Backe great write-up! This forum is the most important organ for vendor change. Look at the response (after two years !! ) when 'Hello King Song' was complained about recently on the forum: there was acknowledgement, & finally capitulation on this pointless most unpopular 'feature'

There's typically a variety of reasons why manufacturers are reluctant to change; it might be the cost to implement a certain feature, standardized parts need to be completely rethought, not enough manpower (GW's weakness), concerns that the volumes bought will not offset the development or tooling costs, insufficient insight or expertise to make a judgement call, & my particular bug-bear, failure to admit that mistakes were made—a particular problem at one company. 

It drives me crazy on those occasions where you've compiled what should be a complete waterproof case: you spend the time to gather the evidence, present the data so that there could be no possible disagreement about the facts, & the response is some lame "we can't do that, because, basically, we don't want to". The Wheeling World needs competition, at least three brands, this is the Prime motivator towards any sort of positive change. 

Gotway have introduced a lot of innovations, especially recently, it's easy to forget who did what first & when. With the help of my Gotway contacts, I put together an infographic timeline of some of the GW firsts. Going to try to do the same with KS & Inmotion. It would be wonderful if these companies could get more analytical on incorporating Customer feedback—especially considered & experienced ones like Marty—features & capabilities during the design phase, rather than presenting releases as a fait accompli 

Not to rain on your parade, but first Ninebots (C or E?) were released in 2014, and the Android app is already mentioned in posts dating back to November 2014 in the Russian forums & elsewhere (AppAnnie could probably tell when the app was added to Google Play, but required registration). 

On the 64-cell config I'm not sure, Firewheel F779 was also released sometime in 2014 (autumn?) having 64-cell configuration, but I don't know if even that was the first (or if it actually came before or after MCM2).

 

[esaj]

4 hours ago, Carlos E Rodriguez said:

Dear Gotway. 

 Consider this improvement. hire a real Electrical Engineer before you kill one of us. 

I'd think they have a real EE or two in the payroll, shooting off remarks like this isn't exactly considerate   Whether the quality/possible design problems are due to to having too few (having multiple engineers gives more alternative views on things and they can peer-review each others work), or trying to make too many products at the same time, I don't know. Probably keeping the costs down also matters, so the hardware designers might have their "hands tied" when it comes to making more expensive boards?

2 hours ago, KingSong69 said:

to call 12 Mosfets an improvement is also questionable. Pure from electrical standpoints it is Nonsens. The right way would have been to implement stronger Mosfets. But Marketing won this battle and 12 Mosfets is seen as an "advantage" by most customers nowadays.

I think Chris from 1Radwerkstatt or someone other "more authoritative", ie. real electronics engineer, once mentioned here or in the private messages that "paralleled mosfets belong to hobbyist devices" or something along those lines. Even then, paralleled mosfets have been used in ("even" western designed) products, but slapping them directly in parallel isn't exactly a good way, the app notes I've gone through recommend at least separate resistors between the driver and the gates, as the gates may have different threshold-voltages and charges, and one or the other might turn on/off sooner than the other, or the other one might not open "fully", and the load may then be spread unevenly (they do even out a bit due to effects happening when the mosfet heats up). There are also recommendations about placement in certain patterns, so that they stay at roughly the same temperature, and use of ferrite beads to prevent gate ringing, which can cause higher switching losses if the gate voltage momentarily drops below full conduction, or even kill the fets if the voltage shoots to too high/low value. Art of Electronics 3rd edition (Horowitz & Hill, ISBN 978-0521809269, considered as the bible of basic electronics by many, https://artofelectronics.net/ ) mentions something along the lines that care should be taken with mosfets circuits, as the manufacturing spreads (tolerances in the component properties, like gate threshold voltages and such) can be relatively high, especially if the components are not from the same batch, and either need to be compensated for through external components or the fets need to be closely matched with testing them by hand. But apparently it hasn't been that much of an issue, as there aren't boards blowing left and right all the time. Still, it seems like it's done to save money on the board, as it's cheaper to put two cheaper mosfets in parallel than one better. The dead ACM -board Rehab1 sent me has the parallel mosfets soldered directly to same pin-holes, so the board looks like it's been designed for 6 mosfets at first, and then they've just added the parallel ones at some point.

[Chriull]

1 hour ago, esaj said:

...

I think Chris from 1Radwerkstatt or someone other "more authoritative", ie. real electronics engineer, once mentioned here or in the private messages that "paralleled mosfets belong to hobbyist devices" or something along those lines. Even then, paralleled mosfets have been used in ("even" western designed) products, but slapping them directly in parallel isn't exactly a good way, the app notes I've gone through recommend ...

I remember the same statement from chris - so it should have been somewhere here on this board...

1 hour ago, esaj said:

... But apparently it hasn't been that much of an issue, as there aren't boards blowing left and right all the time. Still, it seems like it's done to save money on the board, as it's cheaper to put two cheaper mosfets in parallel than one better. The dead ACM -board Rehab1 sent me has the parallel mosfets soldered directly to same pin-holes, so the board looks like it's been designed for 6 mosfets at first, and then they've just added the parallel ones at some point.

If the electronic design is done properly, two (identical - not cheaper) mosfets in parallel give one huge advandage: the dissipated power of the mosfets is proportional to the square of the current to the motor. If one manages to half this current by paralleling two mosfets the dissipated power per mosfet is 1/4 and in total it is halved...

So as heatsinks are not really overdesigned this could lead to a big relief in heat managment...

... but on the other side, choosing a mosfet with 1/2 the Rdson one can reach the same effect without all the possible design probs...

[US69]

1 hour ago, esaj said:

I think Chris from 1Radwerkstatt or someone other "more authoritative", ie. real electronics engineer, once mentioned here or in the private messages that "paralleled mosfets belong to hobbyist devices" or something along those lines. Even then, paralleled mosfets have been used in ("even" western designed) products, but slapping them directly in parallel isn't exactly a good way, the app notes I've gone through recommend at least separate resistors between the driver and the gates, as the gates may have different threshold-voltages and charges, and one or the other might turn on/off sooner than the other, or the other one might not open "fully", and the load may then be spread unevenly (they do even out a bit due to effects happening when the mosfet heats up). There are also recommendations about placement in certain patterns, so that they stay at roughly the same temperature, and use of ferrite beads to prevent gate ringing, which can cause higher switching losses if the gate voltage momentarily drops below full conduction, or even kill the fets if the voltage shoots to too high/low value. Art of Electronics 3rd edition (Horowitz & Hill, ISBN 978-0521809269, considered as the bible of basic electronics by many, https://artofelectronics.net/ ) mentions something along the lines that care should be taken with mosfets circuits, as the manufacturing spreads (tolerances in the component properties, like gate threshold voltages and such) can be relatively high, especially if the components are not from the same batch, and either need to be compensated for through external components or the fets need to be closely matched with testing them by hand. But apparently it hasn't been that much of an issue, as there aren't boards blowing left and right all the time. Still, it seems like it's done to save money on the board, as it's cheaper to put two cheaper mosfets in parallel than one better. The dead ACM -board Rehab1 sent me has the parallel mosfets soldered directly to same pin-holes, so the board looks like it's been designed for 6 mosfets at first, and then they've just added the parallel ones at some point.

 

8 minutes ago, Chriull said:

... but on the other side, choosing a mosfet with 1/2 the Rdson one can reach the same effect without all the possible design probs...

That's what i meant with "Nonsens" and i also got it from a Quotation from Chris, .....

Als Allways, @esaj and @Chriull got my harsh words "electrical Nonsens"   into  a lot better Explanation :-)

[esaj]

Found the quote:

"But i can say (this is real) 12 Mosfet´s is electrical trash!(you can make this on hobby table when you have not the right parts on hand) You can stack stones for a wall but not MosFet´s for HighPower.

When a car need´s more Engine Power you can put a Engine on top to the other Engine and can not see to the Street  in fact the right way is one larger Engine with more Power. When you double MosFet´s on one side is split current but Energy is fast and very exact one MosFet switch faster on later one is internal a little bit different and more stress to the gate driver...many electrical things...other side you double the failure in MosFet self....wrong way is stacking...

You can buy one MosFet with can handle hundreds of Ampere or a module with 6 matching MosFet´s in one case...this and only this is the right way.

Problem in all Wheels is mounting a MosFet correct...first: stupid 1mm thick heat conductive rubber (stupid stuff) second: mounting in the corner from the heat sink third: mechnical stress...screw mosfet with soft rubber on heat sink.....all not OK.... one industrial case (IGBT) with correct internal mounting and.....work

"

I've wondered why 1RadWerkstatt stopped selling all other brands than King Song, maybe it's the only one he sees as "good enough" design from the electronics standpoint?

[US69]

1 hour ago, esaj said:

I've wondered why 1RadWerkstatt stopped selling all other brands than King Song, maybe it's the only one he sees as "good enough" design from the electronics standpoint?

As far as i know he did not stop selling them in general.....he is only not delivering them to the end customer anymore, but providing them to the Re-Sellers. (but i can be wrong on this!)

Also: As a "exclusive" KS importer for Germany he has to buy a lot of KS each order....he is providing all other German EUC seller's with these.

In the end....if you buy a KS in Germany it is Coming from his stock.

He is doing a lot of his Business  in repairs and Batterie upgrades also....and what he thinks about others brand Quality....he might tell you in a personal discussion, but not written on a Forum :-) But Yes, i would guess from an electronical/electrical standpoint he sees KS much in front.

 

As i have heard in Asia ´from KS Thailand and also through the wheelies Thomas Hoon, he has a very good Reputation at KS, and even helped them to develop their boards and BMS's further. Good example might be the newest 350Amp Mosfets, which where introduced on his recommendation.

 

[Marty Backe]

2 hours ago, esaj said:

Found the quote:

"But i can say (this is real) 12 Mosfet´s is electrical trash!(you can make this on hobby table when you have not the right parts on hand) You can stack stones for a wall but not MosFet´s for HighPower.

When a car need´s more Engine Power you can put a Engine on top to the other Engine and can not see to the Street  in fact the right way is one larger Engine with more Power. When you double MosFet´s on one side is split current but Energy is fast and very exact one MosFet switch faster on later one is internal a little bit different and more stress to the gate driver...many electrical things...other side you double the failure in MosFet self....wrong way is stacking...

You can buy one MosFet with can handle hundreds of Ampere or a module with 6 matching MosFet´s in one case...this and only this is the right way.

Problem in all Wheels is mounting a MosFet correct...first: stupid 1mm thick heat conductive rubber (stupid stuff) second: mounting in the corner from the heat sink third: mechnical stress...screw mosfet with soft rubber on heat sink.....all not OK.... one industrial case (IGBT) with correct internal mounting and.....work

"

I've wondered why 1RadWerkstatt stopped selling all other brands than King Song, maybe it's the only one he sees as "good enough" design from the electronics standpoint?

I can't argue with this design discussion but at the end of the day, isn't what works what actually matters. Unless I'm wrong Gotway boards are reliable and perform their function. Apparently thousands of 12 Mosfet based wheels have been sold and they work with no overheating issues. Nearly all of the failure modes discussed here and elsewhere are due to poor quality assembly and design choices associated with cabling. Regarding the control boards all I hear is how well they function in terms of running cool.

If all this discussion here and elsewhere regarding their questionable control board design choices was actually relevant to the real world I would expect to be reading about control board failures. I never do.

[esaj]

1 hour ago, Marty Backe said:

I can't argue with this design discussion but at the end of the day, isn't what works what actually matters. Unless I'm wrong Gotway boards are reliable and perform their function. Apparently thousands of 12 Mosfet based wheels have been sold and they work with no overheating issues. Nearly all of the failure modes discussed here and elsewhere are due to poor quality assembly and design choices associated with cabling. Regarding the control boards all I hear is how well they function in terms of running cool.

I can't give professional opinions on the board designs either, being just a hobbyist/amateur in this stuff, it just seems that paralleling mosfets is frowned upon, when there is the option to use better (but more costly) single mosfets and/or better cooling to keep the design simpler and lower the amount of possible points of failure. There are real applications which use paralleled mosfets outside of wheels too, although only example that comes to mind right now was some sort of massive artificial load used to test power stations, where the paralleled mosfets were used to control current sinking of hundreds of kilowatts of power at kiloampere (1000's of amperes) current levels. I doubt there's a way to do that without a very large amount of paralleled mosfets, but it also had complicated gate driver-circuitry with each mosfet having their own driver (and it's not a switching application like the wheels). But why else would there be so many application notes that concentrate on pitfalls of paralleling mosfets and how to avoid them, if paralleling really wasn't an option? Vishay's note actually says in the end that:

The conclusions presented in the above mentioned paper indicate, however, that the problem is not severe, as long as attention is paid to the following points, in order to ensure satisfactory sharing of current between parallel Power MOSFETs at turn-on:

• Threshold voltages should be within determined limits.

• Stray inductances throughout the circuit should be equalized by careful layout.

• Gates should be decoupled with individual resistors, but not more than strictly required, as it will be explained later.

(regarding Dynamic Sharing of Current at Turn-on)

While a turn-off unbalance is potentially a more serious problem, the analysis in ref. [1] shows that this is not so in practice as long as the devices are turned off with a “hard” (very low impedance) gate drive. This by itself will almost guarantee limited dynamic unbalance at turn-off. In summary, to achieve good sharing at turn-off the same precautions should be used as for turn-on, with the addition of a “hard” drive. 

(regarding Dynamic Sharing at Turn-off)

So, even (some) fet-manufacturers say it's possible to parallel them without more or less complex balancing schemes. Still, in general it seems that it's not recommended to parallel mosfets without balancing components (decoupling resistors, possibly inductors) between the gates, but if it works just fine in real life, does it really matter?

All the manufacturers have had (and probably will continue to have) failed boards here and there, I doubt in the long run any manufacturer can have 0% failure rate. It happens with "better" products too, like cars, where the design and component selection is much stricter with standards and testing. AFAIK, the 12-mosfet boards have worked well, or at least the problems seem to be nowadays (mainly) elsewhere than in the motor drive-part. And even if the board of wheel X was designed "perfectly" (if such was even possible, because probably if you ask the opinion from many enough EE's, you'll have many different opinions about what is "perfect"), you still have at least the firmware, mechanical parts etc that can cause problems   No wheel model is 100% safe and fault-free.

 

1 hour ago, Marty Backe said:

If all this discussion here and elsewhere regarding their questionable control board design choices was actually relevant to the real world I would expect to be reading about control board failures. I never do.

Luckily, nowadays broken boards seem to be much more rare occurrence generally in all wheels than, say, 2015, so there's definitely improvement. If memory serves, on Gotways MSuper2's had some board faults, also maybe the early ACMs before they put the parallel mosfets in? Vee's (EUC Extreme's) mods were also a story of their own, but I think he overvolted the boards anyway and used LiPo-packs that could give >100A continuous current, with at least one occasion where the component legs melted/exploded away completely    Also most wheels no longer simply cutout at high speeds, which seemed far too common back then too  

[Marty Backe]

52 minutes ago, esaj said:

I can't give professional opinions on the board designs either, being just a hobbyist/amateur in this stuff, it just seems that paralleling mosfets is frowned upon, when there is the option to use better (but more costly) single mosfets and/or better cooling to keep the design simpler and lower the amount of possible points of failure. There are real applications which use paralleled mosfets outside of wheels too, although only example that comes to mind right now was some sort of massive artificial load used to test power stations, where the paralleled mosfets were used to control current sinking of hundreds of kilowatts of power at kiloampere (1000's of amperes) current levels. I doubt there's a way to do that without a very large amount of paralleled mosfets, but it also had complicated gate driver-circuitry with each mosfet having their own driver (and it's not a switching application like the wheels). But why else would there be so many application notes that concentrate on pitfalls of paralleling mosfets and how to avoid them, if paralleling really wasn't an option? Vishay's note actually says in the end that:

The conclusions presented in the above mentioned paper indicate, however, that the problem is not severe, as long as attention is paid to the following points, in order to ensure satisfactory sharing of current between parallel Power MOSFETs at turn-on:

• Threshold voltages should be within determined limits.

• Stray inductances throughout the circuit should be equalized by careful layout.

• Gates should be decoupled with individual resistors, but not more than strictly required, as it will be explained later.

(regarding Dynamic Sharing of Current at Turn-on)

While a turn-off unbalance is potentially a more serious problem, the analysis in ref. [1] shows that this is not so in practice as long as the devices are turned off with a “hard” (very low impedance) gate drive. This by itself will almost guarantee limited dynamic unbalance at turn-off. In summary, to achieve good sharing at turn-off the same precautions should be used as for turn-on, with the addition of a “hard” drive. 

(regarding Dynamic Sharing at Turn-off)

So, even (some) fet-manufacturers say it's possible to parallel them without more or less complex balancing schemes. Still, in general it seems that it's not recommended to parallel mosfets without balancing components (decoupling resistors, possibly inductors) between the gates, but if it works just fine in real life, does it really matter?

All the manufacturers have had (and probably will continue to have) failed boards here and there, I doubt in the long run any manufacturer can have 0% failure rate. It happens with "better" products too, like cars, where the design and component selection is much stricter with standards and testing. AFAIK, the 12-mosfet boards have worked well, or at least the problems seem to be nowadays (mainly) elsewhere than in the motor drive-part. And even if the board of wheel X was designed "perfectly" (if such was even possible, because probably if you ask the opinion from many enough EE's, you'll have many different opinions about what is "perfect"), you still have at least the firmware, mechanical parts etc that can cause problems   No wheel model is 100% safe and fault-free.

 

Luckily, nowadays broken boards seem to be much more rare occurrence generally in all wheels than, say, 2015, so there's definitely improvement. If memory serves, on Gotways MSuper2's had some board faults, also maybe the early ACMs before they put the parallel mosfets in? Vee's (EUC Extreme's) mods were also a story of their own, but I think he overvolted the boards anyway and used LiPo-packs that could give >100A continuous current, with at least one occasion where the component legs melted/exploded away completely    Also most wheels no longer simply cutout at high speeds, which seemed far too common back then too  

Well said. It's certainly interesting to read your guys opinions on this stuff. I just think it's an academic exercise at this point.

[Spinner]

TESLA, please make an electric unicycle...

[Carlos E Rodriguez]

9 hours ago, esaj said:

I'd think they have a real EE or two in the payroll, shooting off remarks like this isn't exactly considerate   Whether the quality/possible design problems are due to to having too few (having multiple engineers gives more alternative views on things and they can peer-review each others work), or trying to make too many products at the same time, I don't know. Probably keeping the costs down also matters, so the hardware designers might have their "hands tied" when it comes to making more expensive boards?

I think Chris from 1Radwerkstatt or someone other "more authoritative", ie. real electronics engineer, once mentioned here or in the private messages that "paralleled mosfets belong to hobbyist devices" or something along those lines. Even then, paralleled mosfets have been used in ("even" western designed) products, but slapping them directly in parallel isn't exactly a good way, the app notes I've gone through recommend at least separate resistors between the driver and the gates, as the gates may have different threshold-voltages and charges, and one or the other might turn on/off sooner than the other, or the other one might not open "fully", and the load may then be spread unevenly (they do even out a bit due to effects happening when the mosfet heats up). There are also recommendations about placement in certain patterns, so that they stay at roughly the same temperature, and use of ferrite beads to prevent gate ringing, which can cause higher switching losses if the gate voltage momentarily drops below full conduction, or even kill the fets if the voltage shoots to too high/low value. Art of Electronics 3rd edition (Horowitz & Hill, ISBN 978-0521809269, considered as the bible of basic electronics by many, https://artofelectronics.net/ ) mentions something along the lines that care should be taken with mosfets circuits, as the manufacturing spreads (tolerances in the component properties, like gate threshold voltages and such) can be relatively high, especially if the components are not from the same batch, and either need to be compensated for through external components or the fets need to be closely matched with testing them by hand. But apparently it hasn't been that much of an issue, as there aren't boards blowing left and right all the time. Still, it seems like it's done to save money on the board, as it's cheaper to put two cheaper mosfets in parallel than one better. The dead ACM -board Rehab1 sent me has the parallel mosfets soldered directly to same pin-holes, so the board looks like it's been designed for 6 mosfets at first, and then they've just added the parallel ones at some point.

THAT IS HILLARIUS!   MOSFET's installed in the same hole!   LOL that is just SPECTACULAR!  

This is very entertaining. I think we shold use all this information and write a book. Real world examples of how not to build electronics!

[Duf]

2 hours ago, Marty Backe said:

Well said. It's certainly interesting to read your guys opinions on this stuff. I just think it's an academic exercise at this point.

Yes I would agree, based on public response from the company (none) regarding the string of dangerous safety issues and lack of consistent quality control. 

Much of what you mention is already implemented on other platforms.  The issues affecting rider safety should get utmost priority of course but damn, they should really put some effort into hardening their product line.  My Ninebot One E+ was built like a tank and from what I see the King Song cases are also incredibly durable.  The Gotway line seems to utilize very cheap plastic/fasteners that is not designed for surviving impacts, even small ones.