WARNING: Gotway Nikola production (assembly) issue [Solved]

[Dzlchef]

38 minutes ago, wheelr said:

Why don't you take her to over heat hill to watch the sunset? 

Yeah, thought about it but I don’t want to lug that b*tch down the hill if she gets all cray cray.  The stress part of my recent tests was at least easy access for Uber or the paramedics depending on how hard I hit, and no traffic. 

[Smoother]

2 minutes ago, Dzlchef said:

The stress part of my recent tests was at least easy access for Uber or the paramedics depending on how hard I hit, and no traffic. 

But seriously, you make a good point.  If Marty had damaged a wrist during the fall, even if just sprained, he would have had twice as much trouble lugging that boat anchor back to civilization with only one good wrist.  And he said the trolly was no use at it doesn't lock in the open position, so no chance of trollying it dead.

Edited June 25, 2019 by Smoother
added text

[Nils]

So, I had planned to open my wheel some time later, but went shopping today for some materials to prepare for later. Having some prerequisites at home it turned out that I couldn't live with the suspense any longer though, so I rolled up my sleeves and opened the wheel up. After a rough fight with silicon I freed up the board enough so that I could somewhat inspect the MOSFETs without taking apart the board.  Everything looked clean, and I couldn't see any traces of glue anywhere. My wheel has performed well even when I've purposefully loaded it, so I already knew it wasn't a lemon, and the first inspection seemed to reflect this. However, you can't really see that much from the outside though, and there could still be glue hiding under the thermal pads which would be impossible to determine from the outside. Hence I proceeded with disconnecting the board and after besting the silicon foe again I disassembled it (a pox on anyone putting silicon in screw heads!) . I suck at documenting stuff well apparently but here's a picture of my MOSFETs with the inner row thermal pad removed:

It was completely clean for all practical purposes. Only the MOSFET to the left had some glue/stuff on it's sides, but nothing covering the pad. The two MOSFETs to the right might look a bit weird in the photo, but those are discolorations and not anything do to with any adhesive. While not shown in the photo, the outer row of MOSFETs was completely clean as well.

I fitted new thermal pads and reassembled the board, which was finicky but not super hard (heat sink at the bottom, place thermal pads with ready made holes in their correct positions, fit the board looking through the MOSFET screw holes, keep pressure on the board and insert screws for the middle row. The outer thermal pad can be adjusted if needed easily enough).

In conclusion, not all Nikola boards are glued it would seem. I have a very early version for what it's worth, product code indicates it was assembled on April the 2nd. This would seem to lend some credibility to the theory that only certain wheels have been assembled badly. I believe that my wheel is a very early one, and that for instance  @Marty Backe's is a later one. If so, that would suggest that it's not a case of any early process being improved on shortly after but crappy assembly for certain wheels.

Anyway, after reassembling the wheel I went for a ride, and a great one it was! My Nikola is now open for business again!

 

Edited June 25, 2019 by Nils

[Marty Backe]

9 minutes ago, Nils said:

So, I had planned to open my wheel some time later, but went shopping today for some materials to prepare for later. Having some prerequisites at home it turned out that I couldn't live with the suspense any longer though, so I rolled up my sleeves and opened the wheel up. After a rough fight with silicon I freed up the board enough so that I could somewhat inspect the MOSFETs without taking apart the board.  Everything looked clean, and I couldn't see any traces of glue anywhere. My wheel has performed well even when I've purposefully loaded it, so I already knew it wasn't a lemon, and the first inspection seemed to reflect this. However, you can't really see that much from the outside though, and there could still be glue hiding under the thermal pads which would be impossible to determine from the outside. Hence I proceeded with disconnecting the board and after besting the silicon foe again I disassembled it (a pox on anyone putting silicon in screw heads!) . I suck at documenting stuff well apparently but here's a picture of my MOSFETs with the inner row thermal pad removed:

It was completely clean for all practical purposes. Only the MOSFET to the left had some glue/stuff on it's sides, but nothing covering the pad. The two MOSFETs to the right might look a bit weird in the photo, but those are discolorations and not anything do to with any adhesive. While not shown in the photo, the outer row of MOSFETs was completely clean as well.

I fitted new thermal pads and reassembled the board, which was finicky but not super hard (heat sink at the bottom, place thermal pads with ready made holes in their correct positions, fit the board looking through the MOSFET screw holes, keep pressure on the board and insert screws for the middle row. The outer thermal pad can be adjusted if needed easily enough).

In conclusion, not all Nikola boards are glued it would seem. I have a very early version for what it's worth, product code indicates it was assembled on April the 2nd. This would seem to lend some credibility to the theory that only certain wheels have been assembled badly. I believe that my wheel is a very early one, and that for instance  @Marty Backe's is a later one. If so, that would suggest that it's not a case of any early process being improved on shortly after but crappy assembly for certain wheels.

Anyway, after reassembling the wheel I went for a ride, and a great one it was! My Nikola is now open for business again!

 

Fantastic. This is very good news. If only there was a way to know which wheels have the suspect boards 

Very happy for you, and perhaps this explains why you had no issues on your hill tests.

I'll trade you wheels 

[Unventor]

2 minutes ago, Marty Backe said:

I'll trade you wheels 

No way I be trading off my KS16X for a Nikola with altitude sickness 

[Dzlchef]

Verifying that there are properly built control boards helps but the fact that there was still some hot glue found doesn't inspire total confidence.   It'd be nice if we check a little more easily.

[Nils]

16 minutes ago, Marty Backe said:

Fantastic. This is very good news. If only there was a way to know which wheels have the suspect boards 

Very happy for you, and perhaps this explains why you had no issues on your hill tests.

Yeah, this really is where Gotway should step in after doing an investigation, with some information on what happened, the extents and assurances that they've addressed this since XXXX-XX-XX.  Regarding the tests I'm sure yours is a lot rougher, but I do believe you would have been fine with my wheel.

 

16 minutes ago, Marty Backe said:

I'll trade you wheels 

Sorry, this will be a family heirloom I'm afraid, I imagine it being passed down from father to son for countless generations

[Nils]

19 minutes ago, Dzlchef said:

Verifying that there are properly built control boards helps but the fact that there was still some hot glue found doesn't inspire total confidence.   It'd be nice if we check a little more easily.

Agreed, and there isn't any data indicating they've stopped using glue at all for assembling the boards. There really shouldn't be anything there at all, period. Again, in the ideal world Gotway would acknowledge this, take action and communicate clearly, but we'll see what they actually do.

Edited June 25, 2019 by Nils

[Smoother]

It might be that one individual assembler, not satisfied with how long it took to align each thermal pad, took it upon himself to increase his piece rate (and pay) by whipping out a glue gun.  If I were Gotway I'd look to see if one assembler had a higher piece rate to the average assembler.

[Marty Backe]

4 minutes ago, Smoother said:

It might be that one individual assembler, not satisfied with how long it took to align each thermal pad, took it upon himself to increase his piece rate (and pay) by whipping out a glue gun.  If I were Gotway I'd look to see if one assembler had a higher piece rate to the average assembler.

I can't believe Gotway is producing so much product that they are paying people based on piece-rates. I like the theory though 

We do need another sample so I was hoping that @Dzlchef would volunteer to take his apart. He's been riding it damn too much anyway 

[Nils]

23 minutes ago, Smoother said:

It might be that one individual assembler, not satisfied with how long it took to align each thermal pad, took it upon himself to increase his piece rate (and pay) by whipping out a glue gun.  If I were Gotway I'd look to see if one assembler had a higher piece rate to the average assembler.

Could be! However, compare the small discrete irrelevant dot of of glue on my board to the glue party on Marty's board. Since I believe my board was probably assembled earlier than Marty's it does not strike me as the same person (unless they had a falling down moment.. ) . I guessing that perhaps the use of some glue has been suggested generally as a helper for assembly, but that some perhaps did not understand the consequences of using too much or the importance of absolutely not covering the pads. All speculation though of course.

Edited June 25, 2019 by Nils

[Dzlchef]

14 minutes ago, Marty Backe said:

I can't believe Gotway is producing so much product that they are paying people based on piece-rates. I like the theory though 

We do need another sample so I was hoping that @Dzlchef would volunteer to take his apart. He's been riding it damn too much anyway 

I'm still contemplating checking my board but I do have 350 miles of fairly stressful riding on the wheel and it's been doing great.   

[Unventor]

13 minutes ago, Marty Backe said:

I can't believe Gotway is producing so much product that they are paying people based on piece-rates. I like the theory though 

We do need another sample so I was hoping that @Dzlchef would volunteer to take his apart. He's been riding it damn too much anyway 

Just a thought @Marty Backe... You need to mod your wheel.

Use some of that space in the Nikola to store bread and sausage....

When you reach the summit you have hotdogs ready for a lunch break. 🌭🍽️

[Smoother]

On 6/25/2019 at 9:44 PM, Marty Backe said:

I can't believe Gotway is producing so much product that they are paying people based on piece-rates. I like the theory though 

We do need another sample so I was hoping that @Dzlchef would volunteer to take his apart. He's been riding it damn too much anyway 

Historically paying piece rate for assemblers is the most efficient way of maximizing productivity relative to cost.  I was under the assumption that most assembly work like this was piece work as the key to keeping costs down, and we all know why so much manufacturing takes place in the far east...low labor costs.  It would cost many times more if European or American workers rates were paid, piece OR hourly. 

In twenty years we will probably see manufacturing centers move to another developing area, as has historically been the case.  We once "lost" "all" our manufacturing jobs to Japan, then they lost them and so on and so on.  There are many places on the planet where labor rates are less than China, and as the standard of living is rising there, the people will naturally demand more, as was seen in the west.  The cycle will inevitably continue.

Edited June 30, 2019 by Smoother

[Smoother]

On 6/25/2019 at 9:59 PM, Nils said:

Could be! However, compare the small discrete irrelevant dot of of glue on my board to the glue party on Marty's board. Since I believe my board was probably assembled earlier than Martys it does not strike me as the same person (unless they had a falling down moment.. ) . I guessing that perhaps the use of some glue has been suggested generally as a helper for assembly, but that some perhaps did not understand the consequences of using too much or the importance of absolutely not covering the pads. All speculation though of course.

Yes, I thought that too.  "use glue but only here" says management some douch bag decides he has a better way.  EDIT or wasn't listening, or misunderstood, or missed the meeting, or couldn't see the AV demo, etc, etc.  There are many possible scenarios.

But the question still remains...who was doing the quality control and observing for manufacturing correctness? And does GotWay stress test each wheel like we see with King Song?

Edited June 30, 2019 by Smoother

[mrelwood]

 

4 hours ago, Smoother said:

In twenty years we will probably see manufacturing centers move to another developing area, as has historically been the case.  We once "lost" "all" our manufacturing jobs to Japan, then they lost them and so on and so on.  There are many places on the planet where labor rates are less than China, and as the standard of living is rising there, the people will naturally demand more, as was seen in the west.  The cycle will inevitably continue.

The cycle has been very evident for example in mid-priced electric guitar manufacturing. From USA to Japan in the 70’s, to Korea in the 90’s, and later Indonesia, China, Vietnam, and India.

China has been able to manufacture good quality products for quite a while already. But as we have learned, it’s not a certainty. Perhaps the quality transformed to a new level when out-sourcing western companies learned to watch over the manufacturing process more closely than before. That leaves the solemly Chinese companies to figure things out themselves, resulting in products that don’t fit to a western mindset as quality ones.

I’m just babbling though, trying to make sense of the situation. What do I know.

[Rywokast]

well i have some shred of hope that this was not something that was done across the board and was maybe just one idiotic employee or something done for a certain batch mistakenly.. but on a slightly positive note my seller from aliexpress tells me that there will be a delay with my nikola because the factory is aware of a manufacturing issue.. apparently they are solving this issue and the next batch will be ready at the end of this month to ship out... my fingers are crossed, but no amount of words from anyone will stop me from checking the board personally when i get mine.. i dont know (and i doubt) that there will be any hardware changes but he did say that they are waiting for the improved version... i ordered the 2100 wh model and as far as i am aware these have not yet been distributed,, i noticed the listing from aliexpress has actually been removed (which in my mind says that they dont want to sell any more of them right now while they wait for the changes to be made and get priority to people who placed their order early, as per aliexpress terms it could mean trouble for them if they dont get them out in a "timely" manner).., so i am quite certain they are ensuring that these will not be going out like that again, and my hopes are that they will be using the 247 mos like on the 100V version, which is still listed for sale, from now on... we'll see. for my own peace of mind i will be checking the board myself but it seems hopeful to me that they will be fixing this, at least going forward.

Edited June 26, 2019 by Rywokast

[erk1024]

6 hours ago, Smoother said:

There are many places on the planet where labor rates are less than China, and as the standard of living is rising there, the people will naturally demand more, as was seen in the west.

Here is a video about how Africa is becoming China's China.

 

Edited June 26, 2019 by erk1024

[Phil McLaughlin]

I have been busy over the last few days exploring the MOSFET thermal coupling issue on the Nikola. I put together a two part video that describes the current situation with the Nikolas and what I saw when inspecting my control board. It also documents a number of informal experiments that I did to test various thermal solutions using the Nikola heat sink, thermal pads and other thermal coupling solutions. The first video is available now and I should have the other done tomorrow that describes what seems to be a significant improvement over Gotway's standard assembly technique.

Of course there may not actually be a need for better thermal performance and the bad units so far may have just been overzealous application of glue. I hope that these videos are useful to those of you that are considering your options on this issue.

 

[Marty Backe]

45 minutes ago, Phil McLaughlin said:

I have been busy over the last few days exploring the MOSFET thermal coupling issue on the Nikola. I put together a two part video that describes the current situation with the Nikolas and what I saw when inspecting my control board. It also documents a number of informal experiments that I did to test various thermal solutions using the Nikola heat sink, thermal pads and other thermal coupling solutions. The first video is available now and I should have the other done tomorrow that describes what seems to be a significant improvement over Gotway's standard assembly technique.

Of course there may not actually be a need for better thermal performance and the bad units so far may have just been overzealous application of glue. I hope that these videos are useful to those of you that are considering your options on this issue.

 

Wow Phil. That was a great video. You were a tiny bit hesitant to call a spade a spade, but that scotch tape experiment was brilliant, and I'm going to say that was close enough to the glue on my MOSFETs. And then your commentary about the MOSFETs working in pairs and that the complimentary pair had no damage because there was no glue, wraps this case up with a nice bow on top of it. 

I hope @Jason McNeil watches this too.

Your frog commentary made me laugh out loud 

Looking forward to Part 2.

Hope you don't mind, but I'm going to post this Facebook so that community gets a little education too.

P.s. I love this community that we have here. We can accomplish anything 

Edited June 26, 2019 by Marty Backe

[stephen]

2 hours ago, Phil McLaughlin said:

I have been busy over the last few days exploring the MOSFET thermal coupling issue on the Nikola. I put together a two part video that describes the current situation with the Nikolas and what I saw when inspecting my control board. It also documents a number of informal experiments that I did to test various thermal solutions using the Nikola heat sink, thermal pads and other thermal coupling solutions. The first video is available now and I should have the other done tomorrow that describes what seems to be a significant improvement over Gotway's standard assembly technique.

Of course there may not actually be a need for better thermal performance and the bad units so far may have just been overzealous application of glue. I hope that these videos are useful to those of you that are considering your options on this issue.

 

Well if I'd of done a test it probably would of been the same😊 and also said just as intellectual also,,,NOT😊,,

Thanks for this video i thoroughly enjoyed it and learnt something new , I'm not tech minded but you explained it well enough that someone with so little knowledge (like me) understood it 👍

Looking forward to the next video thanks

[Nils]

2 hours ago, Phil McLaughlin said:

I have been busy over the last few days exploring the MOSFET thermal coupling issue on the Nikola. I put together a two part video that describes the current situation with the Nikolas and what I saw when inspecting my control board. It also documents a number of informal experiments that I did to test various thermal solutions using the Nikola heat sink, thermal pads and other thermal coupling solutions. The first video is available now and I should have the other done tomorrow that describes what seems to be a significant improvement over Gotway's standard assembly technique.

Of course there may not actually be a need for better thermal performance and the bad units so far may have just been overzealous application of glue. I hope that these videos are useful to those of you that are considering your options on this issue.

Very nice Phil, I really enjoyed watching that and am looking forward to part 2 and what you'll be doing there! Like your room setup as well

[meepmeepmayer]

8 hours ago, Rywokast said:

but on a slightly positive note my seller from aliexpress tells me that there will be a delay with my nikola because the factory is aware of a manufacturing issue..

Big news!

That means they either stopped using glue, or they change to TO-247s, or both!

[meepmeepmayer]

@Phil McLaughlin Fantastic! 

[wheelr]

1 hour ago, meepmeepmayer said:

That means they either stopped using glue, or they change to TO-247s, or both!

It's not an either or issue. Glue should NEVER be used anywhere near any MOSFET