Phil McLaughlin

In previous posts I have incorrectly referred to the supply coming from the BT speaker board as being 5V; it is a 12V supply. The 84V power connection to the BT speaker module supplies the power to this module from the battery. The 12V positive lead from the BT speaker module is a supply voltage made on the BT board by regulating down the 84V. This is an output from the BT speaker module that was used to power the fan when using the original Nikola board.

My apologies to all for incorrectly referring to the Nikola fan supply voltage as 5V; it is 12V. My post on fan connection was meant to point out that the new control board can be connected to the original fan wiring without cutting wires or crimping on new connectors. This is something useful to try for anyone with a Nikola fan problem (and two fans) since it eliminates both the potentially defective fan and the voltage supply from the new daughter board. I have corrected the post to clean up the record on this.

Oh good catch on that... I do not remember actually measuring the fan supply voltage but the new fan and old fan are the same. eWheels are now shipping the new boards without fans and provide instructions to connect the old fan up to the old BT speaker connection as I described. My new board which was shipped early came with a fan installed so I just used it as-is although I had to make my own jumper wire.

I am guessing that they determined the control board 12V power budget did not support the fan on the original board but they had additional amperage available on the BT speaker board 12V supply. I agree that the decision seems a little odd. This was one of the first things I noticed when I opened up my Nikola for glue inspection but I decided not to comment on it because it was not really a problem and would draw attention away from the very real and serious MOSFET surface contamination issue. As a designer there are things that I would never personally do that are not necessarily "errors". For example the high voltage capacitors on the Nikola are rated at 100V and must support 84V DC and regenerative braking spikes in typical use. I would have specified 150V caps for this application. This does not necessarily mean Gotway was wrong but clearly there is not a lot of margin in this design. But we must also understand that these wheels are consumer items and component cost matters.

@Nils, so sorry to hear that your new board setup is failing! The old fan can be connected directly to the new board without cutting wires or crimping on new connectors. The old board did not have a 12V supply for the fan on the control board and instead used 12V from the Bluetooth speaker module. The ground wire from the fan connected to the control board and the fan was controlled by the board selectively grounding the fan negative connection when the board was too hot. The wire and connector from the BT module only carried +12V but no ground. The header on the old control board only had a switched ground pin. The new board has a new daughter board that makes a +12V for the fan so that the BT speaker module +12V does not need to be used. On the new board the +12V and Ground wires from the fan are connected to a 2 pin JST connector on the control board. a jumper wire is connected from the daughter board to the header next to the fan connection header to connect the 12V to the fan connection. You can use the original fan on a new board by connecting the +12V to the BT speaker module just as before and connecting the ground cable from the original fan to the fan header on the new board just as before. In this case the jumper wire to the daughter board should be removed. All you need to do is bolt on the old fan and connect the wires with original cables and connectors. Using the old fan on your new board will solve two potential problems. In case your new fan is bad you will be using your original working fan. Also, if the +12V daughter board supply is intermittent you will have removed it from the equation by using +12V from the speaker module as with your original board. If the fan is still intermittent after this change you will know that the problem is with the thermal control of the switched ground on the control board header.

It would be really interesting to see what your original board looks like underneath. If your MOSFETs were slathered with glue and you managed 1000 trouble free miles anyway it would be a very interesting data point. I am guessing that the Nikola boards fail readily only when both MOSFETs in at least one pair are contaminated. Great to hear you are up and running with the new board!

Just a small clarification... The MOSFETs that are potentially contaminated with glue on the first generation 84V Nikola boards are the six that are in the center of the control board (the pull-up MOSFETs of the H-bridge). Only the mounting screws of these MOSFETs are visible from the top of the board. The picture that @Fco Javier Padilla Jimenez posted only shows the outer edge MOSFETs (the pull-down MOSFETs). We have not seen any examples of contamination on these outer row MOSFETs most likely because glue would not aid the assembly process. I think it is important to point out that the absence of visible glue at the edges of these outer MOSFETs does NOT mean that the board is glue-free. The only way to determine if there is glue contamination is to remove the board from the heatsink and examine the underside of the middle row MOSFETs. There have been cases such as @Nils board where there was no apparent glue and the board was probably fine but this was only determined by full disassembly. That being said I agree with the view that all 84V Nikola owners with the first generation board should find a way to upgrade to the new control board.

Great! This is the new board with the larger MOSFETs so you have nothing to worry about.

Congratulations on the new wheel! It is really a great ride in spite of these early manufacturing glitches.The picture attached limits our knowledge on your situation. You did not say if your wheel was a 100V or 84V variant. In either case, If it was my wheel I would not worry about glue after seeing the full size thermal pads. My reasoning is: if it is a 100V Nikola a different board was used and we have no reports of manufacturing problems with this board; if it is an 84V Nikola it either has the new board installed or it has the original board with a full size thermal pad indicating that it was manufactured after the glue problem was recognized by Gotway. Applying glue would not aid the assembly process if the larger pad is used so I would not expect that it was applied. If the wheel is an 84V variant I would recommend taking a better picture so that it can be verified whether it is the new board. If it is the old board with larger thermal pads (we have not seen this) you should probably request an upgraded board. I am still riding my now glue free 84V Nikola with the original board but plan to swap in the new board after some additional study of the new board.

Yes I fully agree with you; the TO-220 self adhesive thermal silicone pads worked very well in testing and are easy to use. I demonstrated the application of thermal paste and mica insulators because otherwise there would have been many comments about how much better thermal paste is. Yes it is better but it is my belief that the silicone pads are probably good enough for any use of the Nikola and A LOT less messy!

Well said! I completely agree with your analysis of the situation.

Congratulations on the new wheel! Personally I love the Nikola and feel that the overall riding experience is probably the best that I have experienced. I would probably not recommend that owners rebuild the current Nikola board at this point until we find out more about the situation with availability of the new board. Although we have a good DIY solution, it seems that Gotway’s response to this problem has been very strong and quick. On a side note, I think that the rapid response to this quality issue is an example of the power of this forum. Knowledgable people working together can really move the dial!

Wow! Just walked off a plane to see this. Looks Great! Aside from larger MOSFET packages, we are seeing a different thermal coupling solution. The long strip of thermal silicone that was shared by all six MOSFETs in a given row has been replaced by some type of thermal air gap solution that is individual to each MOSFET. I think that I see an edge of double sided Scotch Tape peeking out from under one of the MOSFET edges... it’s a joke... IT’S JUST A JOKE! Actually, I am guessing that they have used individual self adhesive thermal silicone pads and it looks like from this picture that the gap between the MOSFETs and the heat sink seem to be smaller. If thermal grease was being used I think that we would be able to see the mica insulators since they are always oversized to account for imperfect alignment. As shown in the tests in my videos the self-adhesive thermal pads can work quite well. Larger surface area packages coupled with a better thermal air gap solution coupled with the absence of adhesive should make a substantial improvement in temperature control. However, what goes between the MOSFET and the heatsink is at least as important as the size of the MOSFET package so I will hold off on cracking open the first celebratory beer until I see the board; there is more to this than just TO-247=GOOD. That being said this is so far looking VERY promising. can’t wait to get a closer look! I am so happy to see Gotway stepping up and taking a sledgehammer to this thermal problem. They make unique products and I really want them to be successful. Marty I am thinking that this new board may have your Nikola laughing at you all the way up Overheat Hill.

Yes, well put! There was some mention of the insulating shoulder washer inserts in the first video and also some cautions about checking for them in the second. I did not discuss sources for these because all of the factory setups would already have them installed since the through-bolts that pass through the MOSFET tab are bolted to the shared electrically conductive heat sink and thereby cannot contact the metal mounting tab of the MOSFET. I just reused the original hardware. Sorry about the length of these videos but I wanted to do a thorough coverage of the subject with close to real-life tests to see if practice actually matched theory and myth.

https://www.digikey.com/product-detail/en/53-77-4ACG/53-77-4ACG-ND/1625215/?itemSeq=295984971 The question with the paste/mica job is whether it is really necessary... It did amaze me with its ability to transfer heat though.