My 84-volt Nikola Triumphs, Tribulations, and Failures

[meepmeepmayer]

5 hours ago, Unventor said:

Ohh I forgot you had other wheels to ride, almost 

Those could have been good time too to test the KS16X 

[Dzlchef]

7 hours ago, Marty Backe said:

It's no fair that your odometer is spinning and mine is stuck at 100-miles 

If you come to the ride Sunday I’d let you take it for a spin as long as you promise to not burn up my MOSFETs. 😁

[Jim Martin]

So is Jason going to accept returns of the $2000 time bomb?  I feel sorry for him, but surely people can't feel confident in this wheel.

[esaj]

15 hours ago, Lukas83 said:

question for the electronic gurus here : 

in theory: if you can get the to247 mosfets and get rid of all that hot glue residue, do you guys think it's possible to change the mosfets myself to the 247 ones?? I know a guy who is an expert in soldering, would not be a problem for him to change the mosfets... but I'm not sure if everything will work if you only change them? of course I will go for a lot better heatsink as well... 

what do you guys think about? 

 

Still reading the earlier pages, so this might be answered already, but in my opinion, IF the mosfet has enough "contact" with the heatsink, larger size mosfet should just shed the heat even better. Still, it's not that simple, a different mosfet will have different electrical properties, and then the question becomes can the gate driver switch it on and off fast enough, or if gate ringing occurs or such... Still, the switching frequency of these devices is pretty low (<20kHz), and probably they have separate gate drivers, so even a higher gate charge probably wouldn't be a problem, but nothing can be said for certain without testing. I'd expect it to work (done properly, ie. "correct" mosfet chosen, good contact with the heatsink, electrically isolated etc), just fine.

Edited June 21, 2019 by esaj

[esaj]

11 hours ago, Marty Backe said:

 

BTW, hot-glue melts at ~120-degrees Celsius. That's probably hotter than these MOSFETs are supposed to get, so the hot-glue likely stays a solid and acts as a nice thermal break.

This is my theory, and I'm sticking to it 

I miss my Nikola  

I hope you're wrong, really   The critical temperature for most mosfets is 175 Celsius, at the semiconductor junction inside the mosfet. If you check the datasheets you find numbers for thermal resistance between the junction and the case (and junction-to-ambient, among others, but they aren't of interest with a heatsink), so you can't really measure the junction temperature directly, but rather you measure the case temperature and calculate the internal temperature according to the thermal resistances and (estimated) power loss of the mosfet, there are ways to estimate the switching losses, but in this case adding a few celcius to the resistance would probably cover that just fine (the switching is relatively infrequent, if you compare to a something like a high-frequency SMPS doing the switching at several hundred kHZ or MHz)? Of course it's down to how high current the gate driver uses to charge the mosfet gate and the gate charge, but the usual numbers are something like <200 nano coulombs and gate driver pushing 1A or more...

Still, the 175C is more of an absolute maximum, a value you should stay away from under normal usage... if there is indeed a hot-glue -barrier between the mosfets and the heatsink, this is going to be a serious problem over time, as the hot-glue will have high thermal resistance (not to mention when it melts and moves away, plain air has much higher thermal resistance, basically you're looking at closer to the junction-to-ambient -values when there's not "thorough" thermal contact).

Here's an old post of mine looking into the basics of thermals:

 

Edited June 21, 2019 by esaj

[Han Vo]

1 hour ago, Jim Martin said:

So is Jason going to accept returns of the $2000 time bomb?  I feel sorry for him, but surely people can't feel confident in this wheel.

+1 for no confident here

[Harold Farrenkopf]

21 minutes ago, Han Vo said:

+1 for no confident here

It might be a Chinese retaliation to you Americans and only your wheels are tainted.

[Mike Sacristan]

Today I checked out Ali and can now see that the Nikola is available in black or white according to :

https://www.aliexpress.com/store/product/GotWay-new-unicycle-84V-Nikola-17-inch-electric-unicycle-2000W-motor-1600wh-big-battery/3896031_32966472876.html?spm=2114.12010612.8148356.13.3c861baeH6OPHi

With the white being the white we know and the black looking like the plus.

I don't even know who I am anymore at this point.

[Kens]

9 minutes ago, Mike Sacristan said:

Today I checked out Ali and can now see that the Nikola is available in black or white according to :

https://www.aliexpress.com/store/product/GotWay-new-unicycle-84V-Nikola-17-inch-electric-unicycle-2000W-motor-1600wh-big-battery/3896031_32966472876.html?spm=2114.12010612.8148356.13.3c861baeH6OPHi

With the white being the white we know and the black looking like the plus.

I don't even know who I am anymore at this point.

ooooh~ they sell Nikola without battery 

@Marty Backe can I use ACM2 battery in 84v Nikola? Are the battery and connector the same??

[Jim Martin]

1 hour ago, Harold Farrenkopf said:

It might be a Chinese retaliation to you Americans and only your wheels are tainted.

Haha, probably so, they are trying to extinguish us Americans one face plant at a time.... We're doomed, time to open my own american made EUC company

[Marty Backe]

12 minutes ago, Kens said:

ooooh~ they sell Nikola without battery 

@Marty Backe can I use ACM2 battery in 84v Nikola? Are the battery and connector the same??

I really don't know. I haven't looked closely at the battery configuration on my Nikola.

[Flying W]

I really hope this is isolated and not effecting all of them. 

I think I assumed the boards came from a supplier assembled ready to be hooked up and that supplier knows what they are doing. 

This seems like non electrical minded folks are just slapping them together and if it powers on QC is passed. 

The addiction level of this hobby makes us want the new gen wheels but it would seem smarter to wait for a long time before pulling the trigger. 

Buying an MSX now makes perfect sense as its proven to be one of the most reliable wheels ever made. Problem is "new shinny itis" has me wanting a nicola or 16x. 

I'm sure the "plus" with the 247s will be bullet proof but that price jump is a little steep in my opinion. 

I'm crossing my fingers and toes that Marty's will conquer over heat hill once its its fixed and the other wheels in the wild are fine.....but in light of the hot glue discovery and the way assembling works, makes it unlikely that only one board had hot glue used. 

Huge thank you to Marty for putting himself, and his wallet on the line to stress these wheels to the max!

[Marty Backe]

32 minutes ago, Flying W said:

I really hope this is isolated and not effecting all of them. 

I think I assumed the boards came from a supplier assembled ready to be hooked up and that supplier knows what they are doing. 

This seems like non electrical minded folks are just slapping them together and if it powers on QC is passed. 

The addiction level of this hobby makes us want the new gen wheels but it would seem smarter to wait for a long time before pulling the trigger. 

Buying an MSX now makes perfect sense as its proven to be one of the most reliable wheels ever made. Problem is "new shinny itis" has me wanting a nicola or 16x. 

I'm sure the "plus" with the 247s will be bullet proof but that price jump is a little steep in my opinion. 

I'm crossing my fingers and toes that Marty's will conquer over heat hill once its its fixed and the other wheels in the wild are fine.....but in light of the hot glue discovery and the way assembling works, makes it unlikely that only one board had hot glue used. 

Huge thank you to Marty for putting himself, and his wallet on the line to stress these wheels to the max!

Thanks.

I have to be honest that I'm not sure when I'll try overheat hill again with the Nikola. I'd like to enjoy riding this wheel during the summer. It could be mid July or later before I get mine back. If it fails again there goes my summer. Maybe I'm done with the Nikola. I'm in a holding pattern with no info about what's going on. I just saw another video by Lin (owner of Gotway) where he's on a long journey to Tibet. So who knows if the lights are on at Gotway right now 

[Harold Farrenkopf]

7 minutes ago, Marty Backe said:

Thanks.

I have to be honest that I'm not sure when I'll try overheat hill again with the Nikola. I'd like to enjoy riding this wheel during the summer. It could be mid July or later before I get mine back. If it fails again there goes my summer. Maybe I'm done with the Nikola. I'm in a holding pattern with no info about what's going on. I just saw another video by Lin (owner of Gotway) where he's on a long journey to Tibet. So who knows if the lights are on at Gotway right now 

Is that the rider going through water with it?

[Rywokast]

2 hours ago, Kens said:

ooooh~ they sell Nikola without battery 

@Marty Backe can I use ACM2 battery in 84v Nikola? Are the battery and connector the same??

yes, same as v3s+ and monster as well

[Rywokast]

5 hours ago, Jim Martin said:

So is Jason going to accept returns of the $2000 time bomb?  I feel sorry for him, but surely people can't feel confident in this wheel.

one isolated incident proves a QC oversight, nothing more.. if there were reports left and right of this happening, or if say this was confirmed to be an issue by gotway, or you cracked yours open and saw the same issue then I'm sure yes but as of now this is an isolated incident.. I ordered the nikola 2100 a week ago when I get it I'm going to be opening it up to have a look

[stephen]

5 minutes ago, Rywokast said:

one isolated incident proves a QC oversight, nothing more.. if there were reports left and right of this happening, or if say this was confirmed to be an issue by gotway, or you cracked yours open and saw the same issue then I'm sure yes but as of now this is an isolated incident.. I ordered the nikola 2100 a week ago when I get it I'm going to be opening it up to have a look

I don't think I'd feel confident riding up the hills where i live knowing in the back of my mind i might be hitting the floor because the mosfets burnt out , if marty didn't find this out it could of been okay worse depending where you ride

[Nils]

15 hours ago, Marty Backe said:

But if my theory holds, seems to me that every Gotway wheel that has an inner row of MOSFETs (not just the Nikola) is a time-bomb waiting to blow. Now I think the other wheels (MSX, Monster, etc) are using the bigger MOSFETs and the hot-glue thermal barrier might not be an issue for those. But when they switched to the smaller MOSFETs, the assembly process became the weak link. 

 

5 hours ago, Jim Martin said:

So is Jason going to accept returns of the $2000 time bomb?  I feel sorry for him, but surely people can't feel confident in this wheel.

 

3 hours ago, Han Vo said:

+1 for no confident here

Plausible theory regarding the bigger MOSFETs saving the other wheels from assembly issues manifesting Marty, and justified reactions from you others - Gotway screwed up here, such a shame following the MSX slam dunk last year (in competition with the other manufacturers with regards to issues reported for the flagships). Remains to be seen how widespread an issue this is for other Nikola owners though. Since I've personally put my Nikola through the paces with my own climbing load test I feel that I'm as confident as I can be for any unicycle with regards to my own wheel and that it will handle the load I'll put on it. Comparing my V10F and Nikola experiences with Marty's suggests my V10F is either worse than the V10F Marty tested, or my Nikola is better than Marty's was.  I personally believe it's the latter.

Regardless, that helps me, but does not help anyone else. It could mean I was one of the few lucky ones with the Nikola, or instead that Marty was one of the few unlucky ones, or that it's 50/50 or whatever. We haven't heard of any other reports as of yet with burned MOSFETs however, but then explicitly pushing the wheels under these conditions are not common either, and perhaps the same error would cause issues for riders with more normal riding patterns in the long run.

Only time will tell, and of course this causes a lot of uncertainty and doubt.

In an ideal world I'd like to see Gotway rush a new motherboard/wheel to Marty for immediate retesting (and it'd make sense in term of damage control you'd think), alas Gotway doesn't seem to really care much about these things.

 

[Marty Backe]

14 minutes ago, Rywokast said:

one isolated incident proves a QC oversight, nothing more.. if there were reports left and right of this happening, or if say this was confirmed to be an issue by gotway, or you cracked yours open and saw the same issue then I'm sure yes but as of now this is an isolated incident.. I ordered the nikola 2100 a week ago when I get it I'm going to be opening it up to have a look

Unfortunately you will discover that it's impossible to inspect the middle row of MOSFETs without de-constructing the control board assembly. And then you may need to buy replacement thermal transfer sheets to reassemble it.

[Rywokast]

10 minutes ago, stephen said:

I don't think I'd feel confident riding up the hills where i live knowing in the back of my mind i might be hitting the floor because the mosfets burnt out , if marty didn't find this out it could of been okay worse depending where you ride

just do it and if you her the incessant beeping heard in martys video STOP lol.. and at that point you should probably inspect your wheel

[Rywokast]

5 minutes ago, Marty Backe said:

Unfortunately you will discover that it's impossible to inspect the middle row of MOSFETs without de-constructing the control board assembly. And then you may need to buy replacement thermal transfer sheets to reassemble it.

I would be fine with that, already have sheets laying around anyways.. I didn't pay 2200 for something that's going to conk out on me :p

Edited June 21, 2019 by Rywokast

[meepmeepmayer]

If it is really glue (looks like it), I wouldn't buy a Nikola now, at least from Ali or any other source without excellent service. Because then all boards might be affected and it is no isolated incident, just the first.

With the uncertainty if they will use a better board in future 84V Nikolas I wouldn't buy a Nikola now, too.

2 good reasons to wait. Frustrating, but the reasonable thing to do. I'm certainly waiting until more info comes up.

Though I certainly appreciate any involuntary beta testers ordering the 2100Wh one and giving us a look (if only at the crash and subsequent service odyssey).

[Jim Martin]

11 minutes ago, meepmeepmayer said:

If it is really glue (looks like it), I wouldn't buy a Nikola now, at least from Ali or any other source without excellent service. Because then all boards might be affected and it is no isolated incident, just the first.

With the uncertainty if they will use a better board in future 84V Nikolas I wouldn't buy a Nikola now, too.

2 good reasons to wait. Frustrating, but the reasonable thing to do. I'm certainly waiting until more info comes up.

Though I certainly appreciate any involuntary beta testers ordering the 2100Wh one and giving us a look (if only at the crash and subsequent service odyssey).

I have the 84v 1600 Nikola from Jason now, and about 3 weeks ago ordered the 100v 1845 from Ali.  I guess you can say I took a major risk, oh well, gotta roll with it.

[Marty Backe]

14 minutes ago, Jim Martin said:

I have the 84v 1600 Nikola from Jason now, and about 3 weeks ago ordered the 100v 1845 from Ali.  I guess you can say I took a major risk, oh well, gotta roll with it.

You're be covered by EWheels if this is a bigger issue. With Ali you're probably screwed  I'm exaggerating a bit of course. Worst case you'll have to get a new control board, but dealing directly with China for this will probably not be fun experience.

One upside to all of this is the Nikola is an amazing riding wheel. It would really suck if we had to deal with these inconveniences for a mediocre wheel.

Edited June 21, 2019 by Marty Backe

[UniVehje]

56 minutes ago, Nils said:

Comparing my V10F and Nikola experiences with Marty's suggests my V10F is either worse than the V10F Marty tested, or my Nikola is better than Marty's was.  I personally believe it's the latter.

In my books even the V10F beats Nikola on this test. Here’s why: 

Number one task of the wheel is to protect the rider. 

Number two is to protect the wheel from damage. 

Only third most important is to take the rider to the most extreme hills. 

The V10F is kicks off the rider too easily but that is still the only correct behavior. 18L/XL does the same but not that easily. They will never cause a faceplant or get destroyed by testing their limits. It’s another issue how large their safety margins are. 

The Nikola didn’t only fail on the third task but failed BIG TIME on the two more important tasks. It caused a faceplant and got broken. It also didn’t report the temperature accurately. Marty was commenting that it’s a cool running wheel while some parts were overheated to the point of destruction.

Same goes for some other Gotways. Their performance could be an illusion because we don’t really know how close to their limit they have been and what the real temperatures were. Only way to know is to have a faceplant and a broken wheel. On some other wheels we can safely test where the safety margins are and they simply let you know that it’s time to stop. Having some protective behavior naturally also has to limit the performance. The manufacturers must set the margins large enough to cover all possible scenarios (same with speed limits). In my books this is the right way to do it. I understand why some people don’t want that, you get more performance but then there is always the risk of not knowing how close to a faceplant you are riding. 

In any case, while I applaud Marty for doing these tests (very entertaining), we should remember that a wheel isn’t bad just because it overheated on the overheat hill. Unless you really need to ride those kinds of hills (say, you live and work in San Francisco and your daily commute goes over the steepest roads) you will almost never encounter anything like that. The 84v Nikola will do fine and seems like a great wheel.