My 84-volt Nikola Triumphs, Tribulations, and Failures

[Jon Stern]

29 minutes ago, eddiemoy said:

Marty, the thermal silicon sheets usually come with some paper adhered to hit so to keep the sticking surface clean.  from what i can tell, that is is the paper and it wasn't removed when applied.  LOL.

There's usually a plastic cover sheet. It's hard to tell what that stuff is around Marty's MOSFETs from the video, but there shouldn't be anything there. I think your hypothesis could be right.

[eddiemoy]

2 minutes ago, Jon Stern said:

There's usually a plastic cover sheet. It's hard to tell what that stuff is around Marty's MOSFETs from the video, but there shouldn't be anything there. I think your hypothesis could be right.

I’ve worked with this stuff a few times with water cooling drives on PC’s. Proper contact is definitely important. 

[Nils]

18 hours ago, Marty Backe said:

Thanks. Please note that the board burned up because of excessive current and small MOSFETs. This was not a heat related problem (my weather vs yours). The Nikola was actually running super cool.

So you are at risk if riding in steep terrain. As long as you stop when the 90-amp alarm starts triggering continuously you'll be OK. But again, this had nothing to do with temperatures.

Yeah, Jason will get me up and running again. No worries there.

Thanks Marty, I'll try to monitor things on my end. I've bought a Bluetooth headset for use with wheellog so that will help, and hopefully sometime in the future wheellog will feature some kind of sustained current alarm that makes this easier to respond to (pretty please @Seba? Could I send a bottle of Whisky perhaps? ). Did you log the ride by the way? Would be interesting to see the current graphs prior to the burn out.

Please let us know how it goes with the new motherboard, and whether the plus version will be an option. Having now watched the video I as well admire and applaud your stoicism! It wasn't even my wheel and I'm still mad as hell. What's the secret? Five hours of meditation per day? Heavy medication? Asking for a friend.

[Afeez Kay]

58 minutes ago, Marty Backe said:

He's looking for a 1600wh wheel

Ah.. get the MSX 100v PLUS then..shipped from Europe. Am getting one ☝🏾 

[EUCGUY]

45 minutes ago, eddiemoy said:

I was looking at that too.  It is almost like GW tech who applied the heat spreader didn't remove the adhesive sheet completely from the thermal silicone sheet.  BAD, this would explain why those 3 blew because it wasn't make direct contact with the thermal silicone.  It was touching paper, then silicone sheet then aluminum heat spreader.

does the paper when remoing often rip and leave some parts behind?  whats the color of the paper?  

[Jon Stern]

6 minutes ago, eddiemoy said:

I’ve worked with this stuff a few times with water cooling drives on PC’s. Proper contact is definitely important. 

I've spent a lot of my career working on difficult electronics cooling problems.

At my current company, we have a very tight thermal budget and gap pads are one of several tools we use. All of this is backed up by extensive thermal simulation work and testing throughout the development cycle.

None of it counts for anything if there's a process error on the assembly line.

[stephen]

7 minutes ago, Afeez Kay said:

Ah.. get the MSX 100v PLUS then..shipped from Europe. Am getting one ☝🏾 

When you getting it have you ordered it yet 

[Dzlchef]

Hmm...I have a Nikola probably from the same shipment as Marty's.   Not sure I want to open up my new wheel, but not sure that shouldn't either.  

[Marty Backe]

12 minutes ago, Nils said:

Thanks Marty, I'll try to monitor things on my end. I've bought a Bluetooth headset for use with wheellog so that will help, and hopefully sometime in the future wheellog will feature some kind of sustained current alarm that makes this easier to respond to (pretty please @Seba? Could I send a bottle of Whisky perhaps? ). Did you log the ride by the way? Would be interesting to see the current graphs prior to the burn out.

Please let us know how it goes with the new motherboard, and whether the plus version will be an option. Having now watched the video I as well admire and applaud your stoicism! It wasn't even my wheel and I'm still mad as hell. What's the secret? Five hours of meditation per day? Heavy medication? Asking for a friend.

Thanks :-)   Unfortunately I did not have WheelLog configured for logging. I'll try to remember to turn on logging on my future wheel tests.

Regarding Stoicism - a good book. Here is one that I've read and offers a good introduction to a calmer way of living life: A Guide to the Good Life: The Ancient Art of Stoic Joy

[Marty Backe]

2 minutes ago, Dzlchef said:

Hmm...I have a Nikola probably from the same shipment as Marty's.   Not sure I want to open up my new wheel, but not sure that shouldn't either.  

I don't think there's much to learn from opening the wheel. To get access to the MOSFETs is pretty much a destructive process. All of the wiring on mine looked great.

[Afeez Kay]

14 minutes ago, stephen said:

When you getting it have you ordered it yet 

Early next month.. the value of the MSX just went up after this Nikola MOSFET issue..

[Nils]

1 hour ago, eddiemoy said:

I was looking at that too.  It is almost like GW tech who applied the heat spreader didn't remove the adhesive sheet completely from the thermal silicone sheet.  BAD, this would explain why those 3 blew because it wasn't make direct contact with the thermal silicone.  It was touching paper, then silicone sheet then aluminum heat spreader.

Hmm, very interesting, thanks! (and nice spotting by @Rehab1). I'm not well versed enough in these things to add anything useful myself, but here's a link to Marty's post mortem at the time he peels off the pads (unexpanded for easier copying). The pad in question does not seem like it melted and ripped from what I can tell from the video:

https://youtu.be/UstREObHBX8?t=1821

 

[Marty Backe]

4 minutes ago, Afeez Kay said:

Early next month.. the value of the MSX just went up after this Nikola MOSFET issue..

LOL

[Afeez Kay]

3 minutes ago, Marty Backe said:

LOL

 it’s true @Marty Backe treasure your MSX 

[US69]

50 minutes ago, Unventor said:

My point here is the GW vs KS doesn't really matter if you push past the limits of the wheel. Doing so will likely get you hurt. 

I just don't see any need to point finger at a fault at someone's new wheel that was in a special situation. 

I do understand things could have been designed better. But most of the EUCs we are riding have some querks. We are still in a young business.

People cry for newer, faster, stronger... But at the expense of testing. 

Now I do think it would have been nice to let some time pass before doing the GW/KS opinion wars. 

I am not special happy with GW, in fact I am happy with my KS18L. I do however see a benefits in competition between brands in the long run. 

Great comment!!!

 

No need to make this a GW/KS question.....

However...Hotway has decided to throw the smaller mosfets on the board....while the steady triggering og Martys Power alarm even showed that the Nicola is consuming amps as crazy...much more than other GW wheels! That makes this decision to go back to smaller MOS even more questionable. But as you said: All/most EUC have starting problems....to be honest: i dont know ONE Euc which did NOT receive any updates/improvements in its „iterations“ „versions“ „batches“....whatever you like to call it

[Seba]

2 hours ago, Rehab1 said:

Yes it appears that GW is using thermal pads instead of paste. Most experts state that paste is more effective than the pads as it fills in all of the microscopic pores between MOSFET and the heat sink. I hate thermal paste as it is messy to work with.

You can't use paste in this case, because metal tab of each MOSFET is electrically connected with center pin (so called "drain"). You can also notice a plastic insulation sleeves used to insulate thermal tab from the screw (that is electrically connected to the radiator).

1 hour ago, eddiemoy said:

Marty, the thermal silicon sheets usually come with some paper adhered to hit so to keep the sticking surface clean.

This type of insulation thermal sheets is not adhesive and there is no additional paper or foil.

[Seba]

20 minutes ago, US69 said:

Hotway has decided to throw the smaller mosfets on the board...

[Unventor]

6 minutes ago, US69 said:

Great comment!!!

No need to make this a GW/KS question.....

Thank you. 

I expect there to be issues with KS16X too, yet I choose to be an early adopter still... I can't help it... I am too stoked about this wheel. And I don't use the wheel to stress it like on those mountain trails. I wish I could but my bad knee don't see it the same way as I do. 😕

I just hope the new trolley is better that KS18L. I use the trolley a lot getting out of the basement at home, and when shopping groceries. Luckily I didn't have too much issues but this rework service and later self paid parts since I didn't trust my dealer. Long story. But V3 stems are fine. 

That is why I also feel so sorry for Marty since he was so happy about the Nikola. 

Now hypotecial GW could maybe fix this easy at first glance. It isn't hard to make new control board with bigger MOSFETs and adding a fuse reposition fan. But it could move a problem down the river sort of speak. Moving the bottleneck somewhere else. 

I hope GW get a future solution so it is less likely to happen. I wounder how tight KS follow this as they could be next in line. Tight cost savings in construction on this type of product is not good. 

[Marty Backe]

OK. Hats off to @eddiemoy. I believe he has identified to root cause of my failure - crappy Gotway workmanship (or whoever builds their boards - do we know?).

I took a high resolution photograph (click if you want to access the full-res version) of the MOSFETs in question and the thermal barrier. You can clearly see a bit of the plastic still stuck on the sheet. I closely inspected both thermal sheets and they look clean except for this section.

So whomever assembled this board, nonchalantly pealed off the backing material and shrugged their shoulders when some of it tore and stuck to the thermal sheet. It's so obvious now. Those three MOSFETs could not dissipate the heat and blew.

This makes me feel better that probably most of the Nilola boards are OK, but unfortunately it's now a bit of a roll-of-the-dice if a Nikola owner gets a crap board. And it's basically impossible to inspect the board without destroying it to gain access   If this was not a second rate Chinese company, they would do a recall out of an abundance of caution and replace the boards with fully inspected ones. But we know that's not going to happen.

I'm tagging @Jason McNeil so that he sees the root cause.

Thanks again @eddiemoy, @Rehab1 and @Jon Stern 

 

Edited June 15, 2019 by Marty Backe

[Rama Douglas]

3 hours ago, Dzlchef said:

The whole build situation is disappointing and I hope for a bigger MOSFET board release eventually, but I still love this wheel and feel confident that it will hold up under my normal riding.   I’ve seen quick spikes over 90A but nothing continuous and the temperature hasn’t varied much.  I only have 100 miles so far and have been riding it like I stole it!  

Sweet! You got a Nikola? Awesome! Don't let all this get you down. Not like Overheat Hill is standard riding conditions. Congrats on the new 1554wh wheel! 👊😎💜

[Marty Backe]

6 minutes ago, Seba said:

You can't use paste in this case, because metal tab of each MOSFET is electrically connected with center pin (so called "drain"). You can also notice a plastic insulation sleeves used to insulate thermal tab from the screw (that is electrically connected to the radiator).

This type of insulation thermal sheets is not adhesive and there is no additional paper or foil.

See the new picture I added. Clearly there was some "backing material" stuck to the thermal sheet. Any ideas?

[Adel]

Going away from the proven and relatively trouble free electronics of the MSX was a dumb move by Gotway. What did they actually save with the change? In general all EUC companies seem to be plagued with various design/quality issues and none of them have produced good quality products on consistent basis. So every new model we have to wonder what they screwed up, as if there are no lessons learned and no processes in place.

[esaj]

14 minutes ago, Seba said:

You can't use paste in this case, because metal tab of each MOSFET is electrically connected with center pin (so called "drain"). You can also notice a plastic insulation sleeves used to insulate thermal tab from the screw (that is electrically connected to the radiator).

This type of insulation thermal sheets is not adhesive and there is no additional paper or foil.

What they could do is separate the heatsinks so that the 3 high-side mosfets with drains on the battery-line are directly connected (with paste) to the heatsink, of course the heatsink will then have the same voltage as the battery, which could be an unpleasant suprise working with a 100V wheel     Then 3 separate heatsinks for the low-side mosfets, one per phase, again with direct contact + thermal paste.  Probably too much of a hassle to build, and not very safe, but they could get lower thermal resistance

[Seba]

Just now, esaj said:

What they could do is separate the heatsinks so that the 3 high-side mosfets with drains on the battery-line are directly connected (with paste) to the heatsink, of course the heatsink will then have the same voltage as the battery, which could be an unpleasant suprise working with a 100V wheel     Then 3 separate heatsinks for the low-side mosfets, one per phase, again with direct contact + thermal paste.  Probably too much of a hassle to build, and not very safe, but they could get lower thermal resistance

In general it's a bad design practice to keep heatsink at any electrical potential. It's also common that low-side MOSFET drains are not connected together and directly to the negative battery terminal, but there are separate shunt resistors to measure phase current. Also I wouldn't blame silicone insulating sheets as they're very goot thermal conductors. Not as good as thermal silicone paste, but difference is not important here.

[meepmeepmayer]

31 minutes ago, Marty Backe said:

This makes me feel better that probably most of the Nilola boards are OK

I hate how this gives me hope but no certainty lol