Rehab1 Posted June 23, 2017 Share Posted June 23, 2017 @Marty BackeAs much as I try I cannot freeze frame the exact moment but your tire could have lost contact with the surface a millisecond after this shot. That probably screwed with the algorithms in the firmware. Can you possibly capture the exact moment as I do not have the 1st generation footage? Link to comment Share on other sites More sharing options...
electric_vehicle_lover Posted June 23, 2017 Share Posted June 23, 2017 1 hour ago, Chriull said: @electric_vehicle_loveror some others from "our firmware members" can help them out to implement a sensorless control (or some mixed mode model to use the sensors just for "calibration"/startup or whatever...), if the interference with the motor cables are not be handled... I can't read everything that everyone already wrote but seeing the video and the explanations of the author, I would say the problem can be the balance algorithm that fails on that situation... I think that balance algorithm have inputs of ECU angle and motor speed, also depends on capacity for motor acceleration/power, that depends on the hardware capacity and rider weight, etc. Balance algorithm should have quite a big math involved (at least for me ;-) ) and I am afraid this companies got the original firmware from someone that did the math for a specific EUC but this companies after started to scale the EUCs for more speed, heavier EUCs, boards placed on different position on the EUC, etc, without redoing the math and the EUC just oscillate because it gets on instable point for the control system/balance algorithm. This is why I am taking time to make the OpenSource firmware, I am learning more math to understand the control algorithm, I do not want to copy from someone else without understand same ground base!! I think EUC is dangerous, at least the potential problems of falling at high speed are really scary!! That's why I am being looking more at 2 wheels... to travel at low speed to minimize the risk, looks like a toy for me as I am interested on commuting. I stopped using hoverboard after 5 days because I understand was not safe to ride at more than 10km/h. Now the EUC, I would say 15km/h is a safe velocity but very slow compared to the 25km/h of an EBike or other 2 wheel devices. Link to comment Share on other sites More sharing options...
lizardmech Posted June 23, 2017 Share Posted June 23, 2017 They need to open source their firmware, EUC users can easily help solve their firmware and control board issues, they should focus on what they're good at, designing, building and selling unicycles. It's not like their competitors currently have any trouble cloning their controllers to begin with. Link to comment Share on other sites More sharing options...
Rehab1 Posted June 23, 2017 Share Posted June 23, 2017 I just received this email from a Chinese company that sold me a replacement ACM board that arrived 2 weeks ago. Now I have 2 bad boards! New message from: xiaoq-cn (136) Hi Daniel, sorry to disturb you! We got to know that the mainboard you bought last time may have problem in firmware. So please stop riding it recently, we will arrange ship out new mainboard to you once we got the.new mainboard from factory. Please reply our message if you can see our message. Thanks! Link to comment Share on other sites More sharing options...
OliverH Posted June 23, 2017 Share Posted June 23, 2017 23 minutes ago, lizardmech said: They need to open source their firmware, EUC users can easily help solve their firmware and control board issues, they should focus on what they're good at, designing, building and selling unicycles. It's not like their competitors currently have any trouble cloning their controllers to begin with. No way. This is protecting your business case and this is the core asset. Link to comment Share on other sites More sharing options...
MaxLinux Posted June 23, 2017 Share Posted June 23, 2017 17 hours ago, Marty Backe said: that transition was enough to cause it to instantly oscillate and throw me off the wheel. What theories have come out regarding the cause of this? Did Gotway change something? Why suddenly did this problem appear in their wheels? Link to comment Share on other sites More sharing options...
MaxLinux Posted June 23, 2017 Share Posted June 23, 2017 16 hours ago, meepmeepmayer said: Their real danger is, as soon as KS or someone else makes a better designed wheel that competes with their high powered ones, they're toast. Why buy Gotway if you can get the same, but better. This is a big opportunity for King Song! Link to comment Share on other sites More sharing options...
Chriull Posted June 23, 2017 Share Posted June 23, 2017 1 hour ago, electric_vehicle_lover said: I think that balance algorithm have inputs of ECU angle and motor speed, also depends on capacity for motor acceleration/power, that depends on the hardware capacity and rider weight, etc. Balance algorithm should have quite a big math involved (at least for me ;-) ) It's not trivial - but well explored and settled "basic knowledge" for people working in this field. 1 hour ago, electric_vehicle_lover said: and I am afraid this companies got the original firmware from someone that did the math for a specific EUC but this companies after started to scale the EUCs for more speed, heavier EUCs, boards placed on different position on the EUC, etc, without redoing the math and the EUC just oscillate because it gets on instable point for the control system/balance algorithm. Could easily be that the greatly increased motor power is the cause - the rest of the system stayed more or less the same. So the compensation by the motor happens now to quick which leads to overshoots and oscillations which just requires a small adaption of the previous controller design (which they presumably never did... ;( ) Link to comment Share on other sites More sharing options...
lizardmech Posted June 23, 2017 Share Posted June 23, 2017 1 hour ago, OliverH said: No way. This is protecting your business case and this is the core asset. The firmware and control boards seem more like a liability to the smaller brands than an asset. They don't have the resources to have lots of people continuously working on firmware and control boards which puts them at a disadvantage to bigger companies like ninebot or inmotion. Opening up their software would be a disruptive business strategy that would allow them to leverage thousands of hours of coding and testing as hobbyists play around with their EUCs. Link to comment Share on other sites More sharing options...
vladmarks Posted June 23, 2017 Share Posted June 23, 2017 What else is new... Sorry to hear about the fall and yet another unsafe board @Marty Backe. As far as I remember, Gotway was attempting to reassure Jason several times regarding this shipment. Turns out they are really unaware of what is happening with the software/pcb dynamics. I do want to give them the benefit of the doubt, since who in their sane mind would not revert the shipment knowing it's /still/ unsafe. Link to comment Share on other sites More sharing options...
steve454 Posted June 23, 2017 Share Posted June 23, 2017 2 hours ago, Rehab1 said: I just received this email from a Chinese company that sold me a replacement ACM board that arrived 2 weeks ago. Now I have 2 bad boards! New message from: xiaoq-cn (136) Hi Daniel, sorry to disturb you! We got to know that the mainboard you bought last time may have problem in firmware. So please stop riding it recently, we will arrange ship out new mainboard to you once we got the.new mainboard from factory. Please reply our message if you can see our message. Thanks! Well, at least they are aware and are shipping you a new one. Hopefully the new board will be a good one. Link to comment Share on other sites More sharing options...
meepmeepmayer Posted June 23, 2017 Share Posted June 23, 2017 6 hours ago, Chriull said: A first step should be a mathematical model of the whole system (motor, physics, control loop) to design a stable controller. That's a well understood topic easy to accomplish by experts. A PID controller is nothing magic and can be perfectly designed... Yes! Manufacturers must start to actually know what they are doing. This, rest of the software, electronics design, everything. Testing is nice, but you can't just throw parts together and test the result and that's it, there will be an unending stream of problems... Link to comment Share on other sites More sharing options...
Marty Backe Posted June 23, 2017 Author Share Posted June 23, 2017 3 hours ago, Rehab1 said: I just received this email from a Chinese company that sold me a replacement ACM board that arrived 2 weeks ago. Now I have 2 bad boards! New message from: xiaoq-cn (136) Hi Daniel, sorry to disturb you! We got to know that the mainboard you bought last time may have problem in firmware. So please stop riding it recently, we will arrange ship out new mainboard to you once we got the.new mainboard from factory. Please reply our message if you can see our message. Thanks! This is very interesting. I would not have expected such a proactive response from a 2nd/3rd party Chinese vendor. Gotway must be taking this seriously and putting the word out to everyone that they've supplied boards to. Guess this confirms that the ACM is affected too. Link to comment Share on other sites More sharing options...
US69 Posted June 23, 2017 Share Posted June 23, 2017 25 minutes ago, Marty Backe said: I would not have expected such a proactive response from a 2nd/3rd party Chinese vendor. Gotway must be taking this seriously and putting the word out to everyone that they've supplied boards to. Guess this confirms that the ACM is affected too. I really found this interesting, too! - Because that small chinese resellers take their sells serious and proactive like you said, Respect for that to this small company! - And because that confirms my theory of GW's saying "mainly sold to chinese " is worth nothing as "xiaoq-cn" is an Ebay worldwide seller - That GW sees its a real problem as this is the first worldwide "recall" in EUC business (while it can alsobe that this reseller did the "real" recall on his own and was only informed by GW) - yip, and that perhaps all kind of GW wheels are in doubt - this gives me again the question which tests have been done on the changed firmware, if any at all.... Link to comment Share on other sites More sharing options...
Marty Backe Posted June 23, 2017 Author Share Posted June 23, 2017 10 hours ago, meepmeepmayer said: In 0.25x slomo you can actually see the oscillation quite well. So the cause of the crash is not the wheel cutting out (as I thought) but the wheel throwing you off like a rodeo bull? It's so hard to know for sure. I remember a bouncing sensation, but maybe that was just the normal bounce from riding into the slope at speed. Link to comment Share on other sites More sharing options...
Marty Backe Posted June 23, 2017 Author Share Posted June 23, 2017 11 minutes ago, KingSong69 said: I really found this interesting, too! - Because that small chinese resellers take their sells serious and proactive like you said, Respect for that to this small company! - And because that confirms my theory of GW's saying "mainly sold to chinese " is worth nothing as "xiaoq-cn" is an Ebay worldwide seller - That GW sees its a real problem as this is the first worldwide "recall" in EUC business (while it can alsobe that this reseller did the "real" recall on his own and was only informed by GW) - yip, and that perhaps all kind of GW wheels are in doubt - this gives me again the question which tests have been done on the changed firmware, if any at all.... I see that Speedyfeet has been affected also, and is recalling units. So it's everyone. Gotway has to be taking this who issue extremely seriously. I really hope they are simply reverting to the firmware that was used previously and not just changing the new firmware. They can't afford ($$$, reputation, etc.) to send out replacement boards that have the same or slightly different problems. Jason told me that they are for the first time allowing him to flash new firmware directly (without taking the boards out). That's a big first. He's coming out to California next week to fix all of his wheels. He also told me that my video provide irrefutable evidence that the wheels have a problem - they can't ignore what they can see with their own eyes. I'm feeling really optimistic that I'll be able to enjoy my MSuper again within a week or so Link to comment Share on other sites More sharing options...
US69 Posted June 23, 2017 Share Posted June 23, 2017 3 minutes ago, Marty Backe said: Jason told me that they are for the first time allowing him to flash new firmware directly (without taking the boards out). That's a big first. He's coming out to California next week to fix all of his wheels. when i have read that they first refused that, i was very irritated...i know for sure that french resellers had the option to reflash boards at the time of the ACM 67 volt release(mai/june 2016?). competition must get very hard if they stopped that option in the time between now. But very good that they reject their position and @Jason McNeil will be able to flash himself! Link to comment Share on other sites More sharing options...
Jason McNeil Posted June 23, 2017 Share Posted June 23, 2017 Late last night Linnea revealed more details into the nature of the fault: the Engineer(s) thought it would be a good idea to implement current limiting (exact code snippet please?!) in an attempt to mitigate the type MOSFET of burnouts that Marty recorded on his ACM during that steep hill climb only a couple days earlier—obviously not from that specific example, but from a handful of other Users. 14 hours ago, Hunka Hunka Burning Love said: If I were a gambling man, I'd trust totally in Gotway and just mail out replacement controller boards to affected users. If I were wanting to be 100% sure, I'd see if I could get Marty's wheel that demonstrated the accident couriered to me on my dime, give him a new replacement wheel from stock I don't think it's fair for Marty to be the unofficial test pilot for GW Wheels Heading out to LA on Thursday to get started on this project. 6 hours ago, Dingfelder said: It will be interesting seeing how this all turns out. And I hope nobody loses any money, including Jason. When dealing product safety, monetary considerations are subordinated to reliability. Link to comment Share on other sites More sharing options...
Scatcat Posted June 23, 2017 Share Posted June 23, 2017 4 minutes ago, Jason McNeil said: Late last night Linnea revealed more details into the nature of the fault: the Engineer(s) thought it would be a good idea to implement current limiting (exact code snippet please?!) in an attempt to mitigate the type MOSFET of burnouts that Marty recorded on his ACM during that steep hill climb only a couple days earlier—obviously not from that specific example, but from a handful of other Users. I don't think it's fair for Marty to be the unofficial test pilot for GW Wheels Heading out to LA on Thursday to get started on this project. When dealing product safety, monetary considerations are subordinated to reliability. How I wish the same attitude was the truth for the manufacturers too! Link to comment Share on other sites More sharing options...
meepmeepmayer Posted June 23, 2017 Share Posted June 23, 2017 @Jason McNeil Those weren't mosfet burnouts, those were cable insulation melting and shorting the motor cables. The mosfets survived. But good to see GW don't even know the exact nature of their problems (unless there's an entire series of blown mosfets we didn't hear about) Also, how does current limiting end in such an oscillation fuck up? Anyways, you're doing the good work hope it does not cost you too much money. Link to comment Share on other sites More sharing options...
Jason McNeil Posted June 23, 2017 Share Posted June 23, 2017 Just now, meepmeepmayer said: Those weren't mosfet burnouts, those were cable insulation melting and shorting the motor cables. The mosfets survived This one is... Link to comment Share on other sites More sharing options...
eddiemoy Posted June 23, 2017 Share Posted June 23, 2017 2 minutes ago, Jason McNeil said: This one is... Jason, I think Marty concluded that the melting cables created the short that blew out the mosfet. Link to comment Share on other sites More sharing options...
Jason McNeil Posted June 23, 2017 Share Posted June 23, 2017 Thanks, that makes sense. Link to comment Share on other sites More sharing options...
meepmeepmayer Posted June 23, 2017 Share Posted June 23, 2017 11 minutes ago, eddiemoy said: Jason, I think Marty concluded that the melting cables created the short that blew out the mosfet. Yep, if the mosfet had blown first, the cabling would be fine (or at least not shorted). It seems to be like this with the 84V ACM: the old (replaced) motor connectors are more or less the same strength as the old (16AWG) wiring. Usually the connectors failed first (the known and solved connector issue), but at least in my case the cabling insulation melted first (there was another case where the old connectors failed right before the cabling would have, so it's close). Mosfets seem more robust than both old connectors and old cabling (for this kind of failure from sustained high currents on hills, not crazy current spikes for other reasons). Now (with better connectors) the (old 16AWG) cabling is most likely the weak point when heat from sustained high current occurs. That's what happened to Marty (I believe he re-did his connectors? Or maybe he had the old ones and the cabling failed first - just like mine). Now that the newer motors have thicker cabling (14AWG), no idea what is the weak point and would fail first from high currents. Anybody want to try? Link to comment Share on other sites More sharing options...
Marty Backe Posted June 23, 2017 Author Share Posted June 23, 2017 7 hours ago, Rehab1 said: @Marty BackeAs much as I try I cannot freeze frame the exact moment but your tire could have lost contact with the surface a millisecond after this shot. That probably screwed with the algorithms in the firmware. Can you possibly capture the exact moment as I do not have the 1st generation footage? Just for you @Rehab1 I took the raw footage and made a nice slow-motion film of the crash. The back-and-forth oscillation can clearly be seen right before I'm ejected. You can also clearly hear the single beep before it cuts out. For grins, watch my feet fly thru the air at the end Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.