My new MSuper V3s+ and Gotway Screwed Us

[US69]

1 minute ago, meepmeepmayer said:

You want to but a >1000Wh 18 inch wheel? Gotway, or the comparably overpriced KS18.

with all other you said i am with you....but this "overpriced" argument on the 18" i dont get..

compare chinese or american prices, they are the same(or KS18 is even cheaper), compare european prices...gotway is even more expensive on the 84volt wheels

But sorry, this is offtopic here...

Gotway can be as "better, faster, more range" as whoever, when this stories of burned wires and totally nonsens firmware go on or repeat, no one who even gets "a bit informed before buying" would choose this faceplanting garantie....

[Tishawn Fahie]

1 minute ago, Maximus said:

I wanted that...but not at the risk of a face plant.

At this point, I don't think it is overpriced, I think it's the price of safety, right?

You are right, currently they are the only option, but I guess at this point I'll stick with the limitations of the manufacturers I trust, and when they come out with the upgrades, then that's when I'll purchase them.  I'm not interested in the cutting edge at the risk of safety, particularly when these wheels will be used by not just me, but members of my family or friends, it's just not worth it.

couldn't have said it any better - I guess other manufacturers put these limitations in for a reason...smh..

[Marty Backe]

22 minutes ago, meepmeepmayer said:

Hmm, will watch again. But it was so quick, seemed just like a cut out. But I understand you don't want to relive the crash

Wrist guards actually feel quite nice, make your wrists warm and comfortable and give you a safer feeling.

It's not a cut out. Hard to describe, but the wheel was still balancing at some level. I've experience a cutout (a generic wheel) and this is not it.

Who knows, may only a random number of boards are affected. Since Gotmay may not have any kind of configuration management, there could be good wheels among the bad wheels.

Wouldn't it be nice to call up the app and see the firmware version???

[houseofjob]

9 minutes ago, meepmeepmayer said:

Thanks for the info. Cut out, cut off, is there a difference? I use "overlean" or "overpower" for the non-hardware-error crash reason. With cut out, I mean the motor simply switching off, and (due to tire resistance from the magnets) the wheel tilting forward and basically stopping instantly.

This gets thrown around with different meanings and understandings on here.

My understanding is cutout, cutoff, overlean, the wheel stays on. But there are no wobbles felt with this, only the slight hanging, limbo sensation before faceplanting

[Marty Backe]

23 minutes ago, Maximus said:

I guess it's just speculation on my part that forums like this one have a major impact on the sales of wheels.  If anyone is willing to get on a Gotway product at this point, then either they don't mind face planting, are unaware of the problem, or are a little crazy.  Once you are emotionally connected or disconnected to a "brand" then it's kind of with you for life, in my opinion.  Apple products, for example, are my preference.  Somewhere in my past they became my product of choice and no matter what functionality a competitor has, it still isn't an Apple product.  Gotway had this opportunity with me and my new ACM, but they just lost me.  I'm more that willing to use a product now like Inmotion or KingSong, even though they may have "inferior" products in terms of speed or range, but once you've lost your confidence in a brand, it's lost.  Maybe this is not true for everyone, and maybe in time I'll try again, but for the foreseeable future my money is not going to a Gotway product.

I'm on the fence right now. Truly depends on Gotway's next move.

[meepmeepmayer]

KS18S 1680Wh looks might fine, Marty

[Marty Backe]

21 minutes ago, houseofjob said:

Ummmm.... super hard to cutout on an 84V 1500W nominal wheel.

Also, wobbles are super noticeable.

I was getting a little side-to-side wobble early in the ride. This was my very first fast ride with the wheel so I was getting my "sea legs" so to speak.

[OliverH]

@Marty Backe which is the production date of your new MSuper?

Glad you had no serious injury.

[Hunka Hunka Burning Love]

Marty, I'm so glad that you are not seriously injured.  You were going at a pretty good clip when you hit the sidewalk ramp edge.  Maybe it is best to go to a medical centre to have things checked out.  I'm thankful that there wasn't like a fire hydrant or sign pole in your path.  You were going straight, but imagine if you were turning from the street onto the sidewalk, your trajectory might have been directed to the side instead where hazardous items are located.  I usually wear a motorcycle type jacket when riding.  It has shoulder pads, elbow L shaped guards, and an optional back pad insert.  I have a mesh summer one which is remarkably cool as it lets air pass through and a warmer one that is also rain resistant.  You can also get matching padded pants.

I watched the accident footage about 10 times, and you can see the wheel hit the slight surface bump and start to oscillate back and forth quickly while beeping to buck you off.  I wonder if the bump triggered a current spike or gyro signal irregularity which the firmware was not able to deal with properly.  It's almost as if it has lost it's orientation of where level is and the oscillations are the controller trying to stabilize itself and recover.  I wonder what the beep was triggered by - over-current, controller loss of balance reference, or other?  In testing firmware, they need a simulation rig that can introduce bumps to simulate real world situations while a 250 pound weight is applied to the wheel.

This is quite a setback for the sport/hobby, dealers who have to manage the issue, and customers who expect a wheel to perform safely.  An user applied firmware update would likely be a simple fix especially in situations like this where a large number of wheels are affected, but unfortunately it sounds like this is not going to be forthcoming from Gotway.  To require customers/dealers to swap out the control board just to fix a firmware issue is a bit ridiculous.  It would be like Dell refusing to issue firmware updates to motherboards and requiring everyone to swap motherboards out due to firmware bugs.  It just doesn't make good business sense.  If someone wants to hack the firmware, I'm sure they could send a controller to one of the hacking services to get the STM32F103 MPU dumped without a huge hassle.  I think they may be shooting themselves in the foot and hurting customers/dealers by trying to protect their IP.

[Michael Vu]

Wow I'm glad you're ok @Marty Backe! Good that you were able to capture it on video. Interesting for you that it was almost instant while for me I was able to ride the wobble for about a second before I fell off. What was your battery level at by the way? For me it was around 50% and since I've charged it back up, I haven't been able to recreate the issue even when dropping off curbs although I do feel like is almost on the verge of doing so. But maybe I need to get the battery level down to 50% again for this behavior to happen.

[meepmeepmayer]

24 minutes ago, KingSong69 said:

with all other you said i am with you....but this "overpriced" argument on the 18" i dont get..

Ok you're right, maybe not overpriced, but expensive. >2000€ makes the wallet cry My 1600€ 1300Wh ACM from the UK is my point of comparison, so >2000€ for >1000Wh seems expensive. But maybe I should have compared to msuper prices or even the higher € prices.

21 minutes ago, houseofjob said:

This gets thrown around with different meanings and understandings on here.

My understanding is cutout, cutoff, overlean, the wheel stays on. But there are no wobbles felt with this, only the slight hanging, limbo sensation before faceplanting

cut out/off = motor cuts out = it deactivates. overlean/overpower = motor stays on at maximum performance, but you slowly (compared to a cut out)) fall forward because it's still not enough. Maybe we can start another thread to argue about the exact meaning of each term

@Hunka Hunka Burning Love I suspect it was just a small change in the balance algorithm that's the reason. And suddenly, an algorithm that was supposed to dampen the oscillation (aka balance the wheel horizontally) did in fact amplify it, until it was too much and, boom,  emergency beeps and shut down. Might just be a little parameter changed, or some numerical error from rewriting existing functionality, who knows.

[flass]

1 hour ago, Marty Backe said:

I still want Gotway to succeed because I love the MSuper. 

You know you are a fanboy when you love a wheel that wants to kill you.... 

I felt bad watching that video. I wish you a quick recovery and I sincerely hope you will get a functional Gotway wheel soon!

[Rehab1]

1 hour ago, Jason McNeil said:

Let's get this sorted out!  

Jason, I am so sorry this is happening to you as well! You were so looking forward to this new shipment of GW wheels along with your customers! You were even going to put the wheels though hell to make sure they were safe for your customers! What a huge blow! Prayers buddy!

[Hunka Hunka Burning Love]

25 minutes ago, meepmeepmayer said:

 I suspect it was just a small change in the balance algorithm that's the reason. And suddenly, an algorithm that was supposed to dampen the oscillation (aka balance the wheel horizontally) did in fact amplify it, until it was too much and, boom,  emergency beeps and shut down. Might just be a little parameter changed, or some numerical error from rewriting existing functionality, who knows.

I wonder whether one other possibility might be a short in one of the hall sensor legs.  It appears that Gotway doesn't insulate the legs like KingSong does plus they solder directly to the frail legs instead of using a PCB junction.  If vibration caused a wire to move and short out a sensor temporarily, I wonder if that would throw the balance algorithm off.  Or if there is a short in the new motor wiring due to sharp edges during harness installation it may not present itself until the wheel is ridden at higher speeds where harsher vibration occurs.

I wish there was some sort of testing lab to experiment on these wheels.  That way you could test a wheel while it is under load and running, introduce bumps of differing amplitude,  monitor temperature rises under increased load (to simulate steep hill climbing), and to see what occurs when you disable hall sensors.

@Marty Backe are you shipping the wheel back to Jason to do a tear down and testing analysis?  My concern if I were a dealer of these wheels is that even though you swap out all affected control boards that the problem continues.  Without definitively identifying the source of the problem, picking easy solution A may not solve it and result in further injuries.  I wonder if Gotway even has one of the affected wheels in their possession to do an analysis on or whether they are guessing at the solution.

[Questor]

Hey Marty.

I'm sorry to hear you had a less than satisfactory experience with your new wheel. I know you were very excited to receive it.  I'll keep up to date on your progress with the analysis and repair.

Q~

[OliverH]

7 minutes ago, Hunka Hunka Burning Love said:

I wonder whether one other possibility might be a short in one of the hall sensor legs.  It appears that Gotway doesn't insulate the legs like KingSong does plus they solder directly to the frail legs instead of using a PCB junction.  If vibration caused a wire to move and short out a sensor temporarily, I wonder if that would throw the balance algorithm off.  Or if there is a short in the new motor wiring due to sharp edges during harness installation it may not present itself until the wheel is ridden at higher speeds where harsher vibration occurs.

I wish there was some sort of testing lab to experiment on these wheels.  That way you could test a wheel while it is under load and running, introduce bumps of differing amplitude,  monitor temperature rises under increased load (to simulate steep hill climbing), and to see what occurs when you disable hall sensors.

@Marty Backe are you shipping the wheel back to Jason to do a tear down and testing analysis?

Isn't the motor running in FOC at that speeds? So a HALL sensor should be responsible only at low revs.

[Hunka Hunka Burning Love]

On 2017-06-22 at 2:57 PM, OliverH said:

Isn't the motor running in FOC at that speeds? So a HALL sensor should be responsible only at low revs.

Yes, I read about that, but I thought someone refuted it later on?  Maybe @electric_vehicle_lover or @esaj can comment.  If that is the case, then at speed, would the wheel continue to function if you unplugged the hall sensor cable or shorted some of the wires out?

Edit:

 

[alfu]

I would never put 4 ounces of any substance, Slime or otherwise, inside a small high-speed tire, for fear of unbalancing it (and causing it to oscillate).

[Marty Backe]

53 minutes ago, Hunka Hunka Burning Love said:

Marty, I'm so glad that you are not seriously injured.  You were going at a pretty good clip when you hit the sidewalk ramp edge.  Maybe it is best to go to a medical centre to have things checked out.  I'm thankful that there wasn't like a fire hydrant or sign pole in your path.  You were going straight, but imagine if you were turning from the street onto the sidewalk, your trajectory might have been directed to the side instead where hazardous items are located.  I usually wear a motorcycle type jacket when riding.  It has shoulder pads, elbow L shaped guards, and an optional back pad insert.  I have a mesh summer one which is remarkably cool as it lets air pass through and a warmer one that is also rain resistant.  You can also get matching padded pants.

I watched the accident footage about 10 times, and you can see the wheel hit the slight surface bump and start to oscillate back and forth quickly while beeping to buck you off.  I wonder if the bump triggered a current spike or gyro signal irregularity which the firmware was not able to deal with properly.  It's almost as if it has lost it's orientation of where level is and the oscillations are the controller trying to stabilize itself and recover.  I wonder what the beep was triggered by - over-current, controller loss of balance reference, or other?  In testing firmware, they need a simulation rig that can introduce bumps to simulate real world situations while a 250 pound weight is applied to the wheel.

This is quite a setback for the sport/hobby, dealers who have to manage the issue, and customers who expect a wheel to perform safely.  An user applied firmware update would likely be a simple fix especially in situations like this where a large number of wheels are affected, but unfortunately it sounds like this is not going to be forthcoming from Gotway.  To require customers/dealers to swap out the control board just to fix a firmware issue is a bit ridiculous.  It would be like Dell refusing to issue firmware updates to motherboards and requiring everyone to swap motherboards out due to firmware bugs.  It just doesn't make good business sense.  If someone wants to hack the firmware, I'm sure they could send a controller to one of the hacking services to get the STM32F103 MPU dumped without a huge hassle.  I think they may be shooting themselves in the foot and hurting customers/dealers by trying to protect their IP.

 

23 minutes ago, Hunka Hunka Burning Love said:

I wonder whether one other possibility might be a short in one of the hall sensor legs.  It appears that Gotway doesn't insulate the legs like KingSong does plus they solder directly to the frail legs instead of using a PCB junction.  If vibration caused a wire to move and short out a sensor temporarily, I wonder if that would throw the balance algorithm off.  Or if there is a short in the new motor wiring due to sharp edges during harness installation it may not present itself until the wheel is ridden at higher speeds where harsher vibration occurs.

I wish there was some sort of testing lab to experiment on these wheels.  That way you could test a wheel while it is under load and running, introduce bumps of differing amplitude,  monitor temperature rises under increased load (to simulate steep hill climbing), and to see what occurs when you disable hall sensors.

@Marty Backe are you shipping the wheel back to Jason to do a tear down and testing analysis?  My concern if I were a dealer of these wheels is that even though you swap out all affected control boards that the problem continues.  Without definitively identifying the source of the problem, picking easy solution A may not solve it and result in further injuries.  I wonder if Gotway even has one of the affected wheels in their possession to do an analysis on or whether they are guessing at the solution.

It's very difficult to analyze this - happens so fast. I'd have my doubts if it weren't for all the other occurrences.

I'm not sending it back to Jason. What's to tear down and analyze? It's just a bunch of bits in the firmware.

I so wish that an English speaking person at Gotway would explain (details) what they changed to make their reliable (from a control board perspective)  wheels unreliable. Why in Gods name did they screw with firmware that has served them so well.

If they send me a new board how am I to have faith that anything is different. How this plays out is going to be very interesting

[Marty Backe]

Z

52 minutes ago, Michael Vu said:

Wow I'm glad you're ok @Marty Backe! Good that you were able to capture it on video. Interesting for you that it was almost instant while for me I was able to ride the wobble for about a second before I fell off. What was your battery level at by the way? For me it was around 50% and since I've charged it back up, I haven't been able to recreate the issue even when dropping off curbs although I do feel like is almost on the verge of doing so. But maybe I need to get the battery level down to 50% again for this behavior to happen.

The battery was fully charged. I have a feeling the wheel has to be moving fairly fast for this to happen. Who knows...

[Marty Backe]

14 minutes ago, alfu said:

I would never put 4 ounces of any substance, Slime or otherwise, inside a small high-speed tire, for fear of unbalancing it (and causing it to oscillate).

It wasn't the Slime if that's what you're thinking. Many of have been using Slime for a long.

[Marty Backe]

44 minutes ago, flass said:

You know you are a fanboy when you love a wheel that wants to kill you.... 

I felt bad watching that video. I wish you a quick recovery and I sincerely hope you will get a functional Gotway wheel soon!

I love the MSuper and ACM. The current rendition has problems, yes. But the fundamental design is great.

I am a fanboy, although slightly shaken in my beliefs right now 

[Tilmann]

@Marty Backe Oh boy, that hurts! I'm really sorry for how the GotWay Gods are treating you lately! Hope your wrist will heal soon! (...diligently digging through my drawers in search of my wrist guards...).

Building upon other's assumptions regarding the (non existing) use of version control systems for the firmware development and my personal knowledge and trust in @Linnea Lin Gotway, I got my own theory, of how this could happen: when the trouble reports startet pouring in after the May 1 FW update, the GW boss called the FW developer to "fix it". I am ready to bet a case of fine Bavarian beer, the FW sources "live" on the C: drive of that developers laptop and the Release Notes (if any) say little more than "changed something". So, the poor guy has no idea, which of the files on his drive actually contains the last stable release. Consequently, he starts fooling around with the troubled release, tries to revert his changes or fix what he assumes might be broken. At the end of his work (or his nerves - whatever came first) he presents the result to GotWay with a desperate "it's fixed!". 

The now following "test" for some reason did not uncover, that the problem still exists (for example, because the FW changes to correct it just moved it up to a higher speed range).

GW flashes the boards with the "repaired" FW and resumes shipping...

While this explanation sounds somewhat "harmless" and like an excuse for GotWay, It's not meant to be - I am just not willing to believe, that Linnea outright and knowingly lied to poor @Jason McNeil. I am convinced, that issues like this will persist as long as quality management is left to goodwill and luck instead of a systematic methodology. The software industry has developed release management processes ages ago and even some back room distributors have startet setting up test beds for EUCs - just none of that ever occurred to GotWay. And I don't even know, whether the FW is developed by a GotWay employee or some freelancer with a laptop.

Even though none of my GW wheels has harmed me so far, I know now, that this was a good portion of plain luck. While I am not ready to disown GotWay just yet, I will not buy or recommend an EUC from them until they demonstrate an honest effort to improve QM.

[Marty Backe]

3 minutes ago, Tilmann said:

@Marty Backe Oh boy, that hurts! I'm really sorry for how the GotWay Gods are treating you lately! Hope your wrist will heal soon! (...diligently digging through my drawers in search of my wrist guards...).

Building upon other's assumptions regarding the (non existing) use of version control systems for the firmware development and my personal knowledge and trust in @Linnea Lin Gotway, I got my own theory, of how this could happen: when the trouble reports startet pouring in after the May 1 FW update, the GW boss called the FW developer to "fix it". I am ready to bet a case of fine Bavarian beer, the FW sources "live" on the C: drive of that developers laptop and the Release Notes (if any) say little more than "changed something". So, the poor guy has no idea, which of the files on his drive actually contains the last stable release. Consequently, he starts fooling around with the troubled release, tries to revert his changes or fix what he assumes might be broken. At the end of his work (or his nerves - whatever came first) he presents the result to GotWay with a desperate "it's fixed!". 

The now following "test" for some reason did not uncover, that the problem still exists (for example, because the FW changes to correct it just moved it up to a higher speed range).

GW flashes the boards with the "repaired" FW and resumes shipping...

While this explanation sounds somewhat "harmless" and like an excuse for GotWay, It's not meant to be - I am just not willing to believe, that Linnea outright and knowingly lied to poor @Jason McNeil. I am convinced, that issues like this will persist as long as quality management is left to goodwill and luck instead of a systematic methodology. The software industry has developed release management processes ages ago and even some back room distributors have startet setting up test beds for EUCs - just none of that ever occurred to GotWay. And I don't even know, whether the FW is developed by a GotWay employee or some freelancer with a laptop.

Even though none of my GW wheels has harmed me so far, I know now, that this was a good portion of plain luck. While I am not ready to disown GotWay just yet, I will not buy or recommend an EUC from them until they demonstrate an honest effort to improve QM.

Well said, and I agree. The sad thing is this is all speculation. Gotway continues to be black hole. I have my doubts that we will learn anything more. I don't know how they are going to prove to Jason or others that these news boards are "fixed".

It's 2017. If they really aren't using source control for their software all hope is lost.

[Tilmann]

27 minutes ago, Marty Backe said:

Well said, and I agree. The sad thing is this is all speculation. Gotway continues to be black hole. I have my doubts that we will learn anything more. I don't know how they are going to prove to Jason or others that these news boards are "fixed".

How can GW demonstrate QM improvements? Just as an example, I would love to see a video showing a test rig, allowing the wheel to spin under load at any selected speed and then induce a "bump" to simulate riding over a curb, tree root, etc. I am not pretending, that's the ultimate proof of a successful fix, but I would feel much better seeing such an effort.

27 minutes ago, Marty Backe said:

It's 2017. If they really aren't using source control for their software all hope is lost.

I run large automation projects in the industry for a living. With a recent multi-million project, it was part of my job to convince a smaller supplier to trade his "it's on my laptop, would you have a USB stick for me to deliver it to you?" to adherence to the most fundamental disciplines of release management and delivery discipline...