Sandy's sarcasm is wearing a little thin. His videos are becoming more showmanship than actually informative as he hams it up for his non-industry audience.
@@Miata822 If I was chief engineer over the Mach E, I would be showing this to the entire team. Plenty of meat there. You want the industrial grade info, that's probably gonna cost you. Buy the report.
@@rockon7848 I have the i3 report. It's mostly part-by-part costing and source info. The Mach E inverter is made by Denso. Ford will probably bring it in-house if they think they can improve it or cut cost. I was impressed by the use of SiC switching.
@@Miata822 So Ford threw a bunch of off the shelf components to get the Mach E out the door ASAP, they will re-engineering it if they live long enough. Something like that?
@@carholic-sz3qv The bigger thing is that it is being mass produced I suppose by a machine created by themselves. Will be interesting to see if there's anything else new with the motors too though of course. We've seen the 3/Y motors with Munro but never S/X. :)
@@alaricgoldkuhl155 really?! Stfu! The model is quick but not fast. Impressive things are buggati chiron(fastest car in the world) with 300mph, mercedes unimog with tousands configurations possible for any applications, a scania 770s.......
From what I have read Ford is not building these electric motors, BorgWarner is supplying them as complete modules. Love the tear down videos, keep up the great work.
@@aitorbleda8267 If you write the class that handles the parking brake the right way, you do not have to maintain it, other than maybe improving it overtime. I'm petty sure making modifications to the code and much, much cheaper then producing all these parts for every single vehicle, plus - if you want to improve the behavior of the mechanical part that means you have to replace it in most cases.
@AnalogueKid2112. Then you probably know that both Ford and GM farm out the vast majority of their design work to the Detroit design infrastructure and beyond ("beyond" being for example the Chevy Bolt having an LG Drivetrain). You can tell that Tesla has control of every detail and are designing organically/systemically, there is design iteration being co-ordinated from a central locus, and even often manufactured in-house. US legacy manufacturers are full of teams that oversee parts design from a zillion vendors off-site and around the world. Fords look like design-by-committee, GMs look like Sears appliances, where contracts are let based on spec to the lowest vendor bid. Teslas are more like the classic band, Boston...where a really smart guy designs all his equipment, plays all his equipment himself and it works really well and sounds different from everyone else.
@@johncahill3644 Spot on. This is how we ended up with the now "infamous" Six Liter Power Stroke. A really good engine that got butchered up in an effort to get it EPA qualified (not a bad thing in itself) by several committees. In so doing they nearly killed a legendary Ford cash cow.
I think 🤔 that engineer was gonna do the 1 gram everyday reduction plan and he never got around to starting it, lol 😂 He was set for complete development cycle with that housing !
6:48 "You don't really need that second washer ..." The washer is used because the shape of the under head of the bolt is typically slightly conical (not flat), designed for mating with harder steels. You need a flatter surface as found in the washer to make the torque reading more reliable on installation and not cut into softer materials such as low carbon steel lamination or aluminum. 15:40 Most castings, like most injection molded plastics, have draft angle. If they don't get a surface perpendicular to the spring surface, it will want to walk out or shift the coil, possibly breaking the silicon carbide transistors or the coil fins. I suppose they couldn't put the mating draft into the spring (probably poke yoke issues during assembly), thus the "standoffs".
I'm pretty sure that spacer at 15:40 would make sense if they had understood the order of assembly. The gate driver board with mounted FET modules is installed, with the FET modules slotting into the cooling plate assembly. Then the spring and wedges/spaces are installed to apply clamping force to the stack.
Large circuit boards subject to vibration tend to flex in not supported spans as a result of vibratory accelerations. This flexing can crack solder joints. Solutions are to use thicker PCB material (high cost) or provide more supports over the span. We did larger PCBs and vib testing will kill them if you don't have the needed support points. Brackets or support rails are a typical solution. I believe that the vibration environment of the car and associated motor is the reason.
I don't see how someone could challenge the intelligence of a design god that understands everything man-made that moves. He could've saved every OEM billions of dollars by using least amount of components to achieve highest level of design perfection. Thing about how much all these money can do to world peace.
resonant frequencies are a b1tch... a simple interaction between a couple of fans and a hdd can make a pc case vibrate so loud you'd think it will disintegrate. :) also by the look of the board at 25:56 the fasteners appear to be thermal couplers and the plate that mounts on top is a large surface heat sink.
Yes, that's undoubtedly how it's actually done. I imagine the MOSFETs are attached to the PCB, and the entire control circuit side is assembled before being fed through the cooling plate with the TIM applied to the pre-applied MOSFETs. This makes it a very simple process that isn't very error prone. The top side assembly could be largely automated then... one tech installed the first plate, sends it to a machine which handles all the welds, then it goes on to the next station... insulation is applied, next plate, back to the next welding station, etc...
@@AndrejGobec Sure, but that's not much of an issue for this particular component since each step is basic enough you can train replacement workers easily and quickly and run several production lines at once. The numerous fasteners are the greatest issue, as always.
@@looncraz The 3 welding steps they described probably go in different sequence - put both coper contacting sides at one time maybe manually, then machine welds it, puts the middle top part and weld it again.
One of the important things that Tesla and SpaceX is doing which is completely over looked is inspiring people in all sorts of areas. It's getting young people interested in engineering, electronics, even construction with the drone videos of the factory builds and so on. My wife teaches primary school kids and often shows them videos of rockets and other such things and the kids love it. It's important to get those kids to want to learn but also injects some positivity in their lifes and future, which is often lacking.
Looks like your team does not have enough power electronics and motor drives background since you focus a lot on the mechanical and manufacturing part. For a motor with or without oil cooling, the difference is the cooling efficiency which determines the size of motor under the given maximum power. So the Mach-E definitely has higher power density than vw ID. For a inverter, the most important indexes are power capacity and control accuracy. The relevant parts are power module, cooling system, current measurement and position sensor. Looking forward to seeing these information in the later e-powertrain tear down videos.
I love that the outside of the Ford inverter casting has "FoMoCo". I'm sure it's supposed to be "Ford Motor Company", but I immediately thought of "Fear of Missing Out Company" lol😂
Imagine the joy on a dealers face, when someone comes in with a rattle in the inverter, and occasional cutting out. " Oh yeah, we get that a lot, someone's dropped a screw inside the inverter." That'll be $1849.63 please.
Have you ever heard of an EV inverter failure? I haven't, but I'm sure it must have happened somewhere. Parts like this are replaced, not repaired. The three copper strips Sandy complains about allow inverter replacement w/o opening the motor housing.
Thank you so much for showing these insights here for free. There is so much any engineer can learn from that, even in totally different industries. Keep up the good stuff!
@@billy-go9kx It's a Borg Warner IDU. B-W did the overall design, subcontracting out the design and part of the manufacturing for the motor and controller. The case and gearing was done at B-W, and the final integration was to Ford specs.
Exactly what I was yelling at the screen lol. I understand his point. People do maybe oversimplify the making of an electric motor. But it does have massively less complexity and individual moving parts
The multitude of stampings in the body shell are all welded together and do not move individually. they are still in the count of parts as an inseparable assy.
Most people don't know about all of the independently moving parts in common machines. How many moving parts are there in a hydraulic lifter? What about the variable valve timing actuators in your favorite ICE engine? And the throttle actuator?
You also never need to disassemble the laminates on the rotor for service. That's like saying babbitt bearings are multiple parts because they are layers of metal.
Why so many screws on PCBs? Mikeselectricstuff pointed it out : it's for avoiding mechanical resonance issues or prevent any flexing on these boards, BGAs don't like this at all. And this inverter assy is exposed to an extreme vibration environment.
Couldn't they have used snap in standoffs to do the same thing? I've been involved in a few product designs and we only used screw standoffs for prototyping. I can't think of a single case where we used them in production.
The design team likely has a spec for the area and weight between circuit board supports for each environment. Since this is rigidly mounted to the drive unit, it's a high shock and vibration environment. Without the freedom to select a thicker, more rigid circuit board stack-up (which creates its own process change problems), the solution is more clamping supports.
@@6AM_YT Latest project is a smallish remote crawler with manipulator attachments for making repairs inside sewer pipes. I'm mostly providing guidance, experience and refining the requirements. My engineering student intern is doing all the heavy lifting. He's been working part time for me since high school and can now out program me in one language and is definitely faster using Fusion 360 for mechanical design. So he is doing most of the hands on work.. I actually quote Sandy to him and say "look we can leave that screw in for now but remember you'll have to figure out how to remove it from the next iteration." I like Sandy's "we call them un-fasteners" . We try to limit them to where a tech might need to access a part for troubleshooting, repair or maintenance. We try to avoid someone accidentally breaking an assembly because they can't figure out how to disassemble it, so we use "un-fasteners" because that is what they are for, not for assembly purposes.
There are many better ways of preventing flexing and fixing a pair of boards and a shield between them. Especially as you have a ridged cover on top of them. With proper design maybe 10 screws for both boards.
Another area of concern is the high number of soldered interconnections between circuit boards. I have seen solder joint failures in the form of “fatigue cracking” (possibly originating from a poorly prepared solder joint) on automotive circuit boards that could potentially occur here over time from motor/drivetrain vibration subjected to the circuit boards. It would be beneficial to reduce part count by integrating more of circuitry onto one or two boards at most.
@@dangraff8467 You can weld certain things, but when you are attempting to electrically attach delicate components, it's not easy to contain the current or magnetic flux welding produces. It would also be extraordinarily time consuming and expensive. Plus, welding is impossible on ball grid ICs of which many are used these days in dense boards. Properly soldered circuit boards should easily last 30 years in a car environment from a physical standpoint (including shock and thermal stresses).
@@dangraff8467 Thicker, more thermally stable circuit boards can improve long term resistance to solder related failures, but the old leaded solder was less prone to crack. Also, proper ventilation of the board and arranging the connectors so they don't get stresses on them which help accelerate solder failures. It comes down to the manufacturer and how much they want to spend to extend life.
What's wild is, this sort of deep-dive, technical episode is really for me as an EV customer. I would never buy a Mach-E after seeing this. I wouldn't want the "many opportunities for failure." I like the idea of buying a clever and efficiently manufactured (lean design) EV.
Skewing magnets is done to rduce torque ripple because of magnet flux linkage to stator fin poles. Optimal total skew is always 1/2 * 360 / Stator_slots degrees. It also smoothens harmonics from motors back EMF, but reduces your torque output by max 10%.
13:10 i was told (FWIW) that we dont use the disk brakes because in wet weather int he winter or in really cold temps the pads can stick to the rotors and take while to thaw in the morning
the IGBTs used in this inverter are called dsc (double side cooling) which have been extensively adopted by Japanese suppliers such as Denso for Prius for years.
That's my guess, the fets are soldered to the gate drive board and they drop in from the top together. On the other hand, the control board must be soldered in on the line so who knows.
Yep. Honestly, them thinking that board was installed after the FETS were welded in rather than it being populated and then installed into the heat sink really makes me reconsider the credibility of this channel.
@@petergabrielsson2013 The thermal expansion spring is installed only after the SiC MOSFETs are inserted between the heat sink plates, thus the separate bracket that they didn't understand in the beginning.
21:56 Is it possible they’re welding and pre assembling the mosfets and then sliding that entire assembly down or up into the cooler? If they did that process they could weld and pre assemble the electric components and bolt them together after they fact.
That's what i think too. Sandy just tried to assemble it in the wrong sequence. They would solder all the mosfets to the PCB first, probably by dirt cheap wave soldering and then they insert it from the bottom into the uncompressed heatsink. They use a spunger to drop in the blank aluminium spacer and the Spring from the top then. Easy as pie. Those boards get tested for sure with a fixture, pogopins, automated testing and the best would be to have the mosfets already soldered to it when testing so that nothing can get wrong in a later soldering stage. By the way, even this steel plate between those boards and all those nickel plated brass standoffs can be assembled beforehand as do the solder bridges between those two pcbs. It is one preassembled and tested unit before they insert it into the housing and heatsink plates. I also thought about why they solder so much and do not use board to board connectors. One thing could be current capability, as a solder joint will always be better and more compact than any crimped connector. The other thing is deteiorating leaf springs inside the connectors. Also due to the vibration any gold plating in such a connector will shave off quickly and then it starts corroding the contact. Solderjoints can crack too, but there are solutions to alleviate excess stress like the fancy curvy Pins they used in the Mach e front Inverter. A very large single PCB would be prone to flexing and cracking due to vibrations, but i would have preferred Rigid Flex PCBs over all those stacked soldered connectors. Also there are a lot of options of floating spring loaded board to board connectors, which are suitable for a high vibration enviroment (hirose, amphenol, etc.). But i guess this is where they try to save the cost.
24:50 I am certain this pcb is mounted first and after that the mosfets are placed one by one and soldered. Or the mosfets are soldered on this pcb first and then this pcb is mounted sliding thos mosfets between the cooling ribs. Both ways are much simpler and are way faster as it will 'auto align' the parts.
Unless Sandy understands a design instantly, he implies there is an issue. This may not be the case. What if there were reasons things were done a certain way? For example, I would expect those soldered connections to be more reliable than a connector, especially for those high voltages.
He is a perfectionist, he brings up valid points but in the end none of it matters, as long as the customer is happy, and the device functions. Nobody typically cares about how their car is assembled.
@@slayyou Until it starts having problems after the honeymoon is over. Triple the amount of coolant hoses, parts, brackets, clamps, etc. as the Model Y -- can't be easy to repair. As much as I like the Mach E's interior and driving experience, I'm going with an Ioniq 5 or a Model Y. Just as fun to drive and better engineered.
About the inverter assembly, I think the Mosfets are first soldered in place on that PCB, then slide into the cooling thingy and the side on the PCB is assembled first, then flipped over and all the bus bars are put in place and welded.. This prevents to get these 90 or so little pins to line up with the PCB, which would indeed be a big headache for the operator assigned to this job. It even could be that the PCB is already fitted with the mosfets from the place where the PCB is assembled, and this could be automated and makes sure the solder connections to the Mosfets are high quality and consistent.
Regarding the parking brake, in some markets (ie the EU) it is a legal requirement to have an independant mechanically interlocked parking brake on an automatic car, which an EV falls under, despite having no actual gears to change. It also tends to be used to ensure the safety case for "un-intended movement" for which an EV is particularly vunerable as compared to an ICE. With an EV, if the battery contactors are in, then the vehicle can suddenly and un-expectidly move if the inverter applies torque. For an ICE, the faults required for the engine to start, fuel pumps to run, gears to engage and the vehicle to move un-expectidly are just about impossible to occur. Not the case with the EV......
@@Fix_It_Again_Tony Legislation doesn't demand that it's a traditional park pawl/park lock - many OEMs are now just relying on the worm-gear driven e-brake integrated into the foundation brake calipers.
I feel like sandy is too focused on cost reduction. Faulting Ford for having an oil cooler while praising others for not having one is nuts IMO. Give me the oil cooler any day. Or welding a high voltage electrical connection over a snap in. I’ll take the weld. Especially in an application with plenty of vibration. He knows his stuff, but this mentality is why we get cheaply built crap. It’s anti-consumer and pro bottom line for the company.
"There was not enough room to get SEVENTEEN in there..." LOL! Sandy is RUTHLESS! Also, holy crap - lining up the board with 90 pins below = bent pins on every unit 100%. Definitely at least 25 dropped screws rattling around in every one of these... Great work Sandy and Ben!
@@AudiTTQuattro2003 what time is necessary? they have the competition right there next to it. You can SEE the differences. And the performance numbers are already known. Ford's motors are 'the best' at nothing I've seen measured
Nice video.. love watching. One thing is interesting though. You are assuming they don't mount the circuits to the silicon board first.. if that were the case there would be no problem aligning a few welded connections..
Can't the MOSFETs be pre-soldered to the board and the whole board go in *first* into the radiator from the bottom, *then* weld the bus bars? EDIT: Saw the second board...
EMI shielding could be indeed the reason for the countless screws in the inverter box. The high power impulses from the switching mosfets can have very high frequency content. To shield those high frequencies you need to have much surface-area to contact the shield parts together. Too little contact-area between the shielding parts renders the shield useless because of skin effect and high parasitic inductance. I saw that happen many times because of oxid layers on the aluminium which acts as an insulator. So in the first glance you think you have a wonderful farraday cage but with all the oxid on your aluminium parts you only have a partly closed cage which is not obvious. So you have to take care of it as an engineer. Just an educated guess....
Shielding for EMI can be a real headache, no? I don't know if they used used a conductive gasket on the lid, perhaps a molded rubber one with a woven metal strip around the entire perimeter.
D. G-S, how do the other manufacturers get around that without the mess that we saw here? I'm sure there's a way to design this better, especially given that we've seen evidence of this being done.
24:30 Wouldn't the transistors be soldered onto the board before and then the whole stack is being put in between the cooling fins? I reckon this is the first step and then you do the other side.
and regardless.. why is Ford the only one doing it this way? It doesn't seem like, regardless of the order of assembly, that this is a very intelligent and efficient way of making this component and connecting it to the motor. Looks like if they took the time to see how others were already doing it, they might have come up with more efficient methods
I don't think so. Look at 16:22 when Ben inserts the slotted MOSFET holder. That component needs to be inserted from the side slot and if the top had been assembled before, then the MOSFETs would get in the way of sliding that component into the spring.
@@Zedus-rl9hp It's almost certainly put together standing on its side, so you can access the top and bottom without having to manually flipping it over.
Please, please, PLEASE!! A Sandy Munro + AvE teardown would be comedy gold of epic proportions. The best of engineering snark from both sides of the border. Please make it happen!
@@MegaPraystation if AvE would hold firm on never being on camera, we would only see AvE's hands. But I would be so happy just to hear the two discuss. I am sure Sandy and AvE would have an amazing exchange over something like the inverter shown here.
You "never see" brazed plate heat exchangers on EV drive units then walk right over to a Tesla drive unit and lay your hand right on their brazed plate heat exchanger. (???)
@@charlesball6519 In the second brief clip barely one second into the video you can clearly see it on the outside of the drive unit (upper left in portion of the image).
"If I was that cover, I'd feel dejected." - Sandy Munro, the greatest Engineer Ford forgot about. Also, did anyone notice the logo Ford had on the casing? "FoMoCo - Guess their FOMO is shining through in these hasty design decisions . . .
I'am a retired electronics engineer, when ever I bought a new pice of equipment and it was held together by screws, I would always open it up to see what's inside, since if the used screws they must want you to take it apart.
8:15 Perhaps it could have been made more compact, but they need to pass the connection through a watertight seal, and they did it with that component.
24:30 I was thinking they'd install the MOSFETs with the same pick'n'place equipment that populated the rest of the board, then drop the board into the inverter casing, slotting the MOSFETs into the heat exchanger. THEN you turn it over and do all the goofy welding. Note also that the inverter is probably assembled in a separate plant, most likely in a low-grade cleanroom. Still a pretty wacky design, I'll give you that.
3 года назад+7
Doesn't the PCB, on the other side of the inverter, is added first? I can believe that it can be mounted in the order that you guys show!
It seems each sub component is designed by a different person or team. Then when they meet for the 1st time, to assemble, they have another team that tries to get the subcomponents connect with additional sub components (connectors).
I think the comparison between the parts on ICE vs electric are MOVING parts. Those 500+ parts in that rotor are essentially one part, since they don't move independently.
@@vidznstuff1 Mentioned in the video. They were earlier switching semiconductors, and now they are moving to silicon carbide Mosfets, which have advantages, but are more expensive.
@Munro Live: Could you have the assembly sequence wrong? If the MOSFETS fall through the casting, they could be inserted from either side. It seems to me that if you started at 23:06, you could have previously coated the MOSFETs, inserted them individually into the circuit board, used a precision jig with slots on the sides that duplicates the casting slot spacing to keep them in alignment, and flow soldered them to the board in one step? That unit would be easier to insert in the casting from that side than inserting the individual MOSFETs and aligning the circuit board with all the pins and the casting. The only thing that would sort of justify the standoffs and aluminum plate is if they act as a power bus or a barrier shield between the two boards. The welding is still a mess.
The EMI shield may be to isolate between the two PCBs as noise fron the one below is highly inductive and probably induces false logic signals into the one above.
With 7000 miles on my Mach E, the vehicle is running flawlessly. (Phone as a key has had issues). With 9000 miles on my Tesla model 3, I’ve had two visits to Tesla service for minor issues. Rear view mirror fell off and the charging door wouldn’t open. I’ll reserve judgement until both have had 50,000 miles on them.
My friend just sold her Tesla model 3 last Saturday and bought the Mustang Mach E and she said it's so much better to her. She had hers for 9 months and had several issues and she had enough.
That's a reasonable way to assess them from the consumer level. But from an engineering level (and I am NOT an engineer.. so looking at it from a limited perspective).. cost analysis and engineering complexity.. Ford is up and down from one area to another. They got some things done well.. other areas seem to need a lot of refinement. Not to say Tesla is perfect lol.. but a lot of their issues is the fit and finish an quality control that they need to improve on.. as a fairly new company, you can expect these to naturally improve and mature over time. They already have drastically. As well as their under body design, going towards the mega castings. As for Ford. I am not so sure these engineering challenges just improve naturally. My concern is these are things pushed on these engineers based on Fords culture of operations. Silos.. bean counters.. parts bins.. management making decisions when they have no real interest in what's best for the customers. Not a fan of the MBAs making decisions on the guts of the cars.
@@jodrhh yes the phone as a Key issue has been resolved to an extent. But every time there is an OTA update it seems I have to reinstall the phone as a key. My model 3 was purchased in December 2020 but is a 2021 model year. The Mach E was purchased in February 2021. We wanted to replace our 2013 infinity jx35 and the obvious choice was the model y buy we decided on the Mach E after test driving both.
23:30 Wouldn't it make more sense to put the circuit board on first, so you plug the MOSFETs in and they hold the MOSFET in place for the welding and assembly?
Like them both as he needs to poke the Ford team to do better, and the fact that they have turned their backs on many of the things he implemented when he was there, well, I don't think he has much sympathy for them.
This shows a term I call " Engineering Inertia". When you design and build something for so long that you become comfortable and complacent. When it is time to design and build something new it is difficult because you are transfixed on the old designs.
Thanks again Munro team. Others may have mention this before but Munro team is like Shelby & Miles from Ford vs Ferrari. Trying to do real engineering but try to play nice with Corporate, everything explaination have to have a positive spin to not ruff the feathers. Stay honest guys.
The parts count relates to MOVING parts, designed to move relative to each other. wearing each other out, needing complex lub. Wow, that inverter is NOT designed for volume production! It's like the designers were told "prototype only", but now they're moving it onto production with zero optimizing. Ford does not expect to need many inverters!
No, not many, they will make new versions for all model variants, of which they will have to have as many as GM promised, 30 by 2025... for total 1mio cars, lol, so few thousands each will do.
@@caseybv74 Ford CEO most likely pushed to market, hard to optimize when in production and this even would be in basic design not just moving operations around
The thing is even though the 500 plates are not moving relative to anything else. They still have to be assembled carefully. Just like an internal combustion engine needs to be assembled carefully. Just like anything else that's built by people, some are built better than others sometimes the design is better, sometimes the manufacturing is better. I drive a Toyota Prius which is a combination of internal combustion and electric and shares many of the components of all the other EV'S out there. It has a small battery, high voltage inverters and converters, coolant pumps etc. If you follow the Prius forums you will find out that the gasoline engine is the most reliable part of the car. At least in gen 2 cars. 2004 to 2009. You almost inevitably have a coolant pump, inverter or battery failure before the gasoline engine has any kind of serious mechanical problem. I suspect when electric vehicles start turning 10- 15 years old you're going to start to see some of the secondary support systems begin to fail. You lose a coolant pump and you're on the side of the road, same as in my prius. A little $50 12 volt pump can strand you. The difference for me at least currently is I can go anywhere and purchase that coolant pump and install it myself. And it's relatively easy to do. If I had the pump on hand I could probably do it on the side of the road if I really had to. Two hose clamps and three 10 mm bolts and an electrical connector. That's it. There's no doubt that electric vehicles are the future. But I suspect we're going to find out that they break down almost as much as internal combustion cars. Particularly if they are assembled badly, or the manufacturer decided to use a cheap crappy supplier for critical secondary systems like a coolant pump. Most of these electric vehicles depend on their air conditioning system to keep everything cool. They have a glycol loop like an internal combustion car but that glycol loop is passed through the air conditioning system to get rid of the heat. So you lose an AC compressor you're probably in limp mode.
At 24:00, would it be possible to have the MOSFETs soldered to the lower PCB as part of PCB manufacturing, and then rather slide the MOSFETs all at once into the cooling fin during mechanical assembly? It will still be a lot of fine alignment needed, but not 90 pins. That implies that the lower PCB would need to be the first electric/electronics part to be mounted into the housing. Soldering of the MOSFETs to the copper plate bus bars would then be after the lower PCB is in place.
IFord could use engineering managers with a big overview about how things could be done, and an organizational structure that would lead to the simplified design vision.
As an R&D Engineer who practiced the "Less is More" philosophy in Automotive, Military, Aerospace and Medical devices, that Inverter is what I would call a real Cluster**&k!😱 I wonder how much that would lighten your wallet to replace? Assuming you vehicle did not turn into a useless lump of Slag after the Inverter melted down and took everything else with it!😬 A definite Steep learning curve at work here. As long as they keep making timely progress. Before they are put out of business. Mike in San Diego.🌞🚀🎸🖖
That's the argument for VW removing it.. However.. VW's motors are FAR less performance oriented then Teslas... like A LOT. I sincerely doubt VWs method without the pump would provide enough cooling for the performance motors Tesla uses. Even on their non performance models, Teslas motors are WAY higher performance than anything VW is using
Tesla oil pump leaks are the best and the most finely engineered oil pump leaks in the automotive industry and legacy automakers are just haters because they can't make their oil pumps leak as efficiently and autonomously.
@@airheart1 well..in the base AWD version of the model 3 for example the front motor is 121 kw, the rear motor is 203 kw. In the ID 4 the (rear ) motor is 150 kw, so in the middle of the two. So it could be at least worth a consideration for the front motor in Tesla AWD cars.
@@tzarcoal1018 I suppose there's a point there. But I wonder what the RPM differences are that those motors are experiencing. Even the smaller front motor in the model 3 AWD may be at higher RPMs, requiring more cooling then the VW motor does. But Idk. Certainly something Tesla should look at.. if it's possible to remove a part.. that's Elon's thing lol
@@airheart1 The motors are direct drive. RPMs will be the same if the wheel circumference is the same. The difference in power output is small and doesn’t explain the need for oil filters on Tesla. Audi isn’t using them on the E-tron either.
This e drive was developed and built by Borq Warner and will likely be sold to other oems. The Lightning will use a Ford in-house designed and built motor assembled at Ford Van Dyke electrification center.
Hello Sandy and Ben, HOW ARE YOU?? Interesting episode today on the Mach-E motor, but I gotta be honest with you, it's way above my technical knowledge, but I'm sure many electrical engineers must love it!! Funny thing though, as I'm watching this video, my wife brings in the mail, and there's a box from Munro L ive with my new Munro baseball cap!! The tracking said it'd come Saturday, but it came a day early, THANKS!! It's a nice cap, black with mesh sides; it's very well made looking; it says "YUPOONG" on the label, I guess it's Chinese, I know Sandy has a lot of contacts over there, and while it appears to have Sandy's autograph on the visor, I assume it's silk-screened on, since Sandy must be WAY TOO BUSY to autograph everything he sells, but I'm happy to have it on my hat and will wear it with pride from Sandy Munro, whom I call the WILLIE NELSON OF THE AUTO INDUSTRY!!! Chas Orvs, Plainville, Ct.
I wouldn't be surprised if the inverter assembly involved a lot of jigs with pre-loaded parts. I also highly doubt that inverter assembly is anywhere near the vehicle assembly line. As in, probably another country.
Still.. why are the other manufacturers not doing it this way? Why are Teslas motors still more efficient and powerful? I think Ford could do with learning some things from the others here
![Felix "Unfair" | [Stray Kids : SKZ-PLAYER]](http://i.ytimg.com/vi/Oswujxm2Ag0/mqdefault.jpg)
The motor featured in this video is NOT a BorgWarner motor.
Hi, can you please expand on this? Is it still a Ford Mach E rear motor?
I was pretty bummed when my lab lost the business for the ford. We were all excited to have that iDM in the test lab and prepped for dyno performance.
I knew Sandy would make the joke on the 17th bolt - that there was no space for it... classic gentle and justified sarcasm we all love!
Sandy's sarcasm is wearing a little thin. His videos are becoming more showmanship than actually informative as he hams it up for his non-industry audience.
@@Miata822 If I was chief engineer over the Mach E, I would be showing this to the entire team. Plenty of meat there.
You want the industrial grade info, that's probably gonna cost you. Buy the report.
@@rockon7848 I have the i3 report. It's mostly part-by-part costing and source info.
The Mach E inverter is made by Denso. Ford will probably bring it in-house if they think they can improve it or cut cost. I was impressed by the use of SiC switching.
@@Miata822 So Ford threw a bunch of off the shelf components to get the Mach E out the door ASAP, they will re-engineering it if they live long enough.
Something like that?
@@Miata822 The info is still all here - the packaging has improved tremendously though
Can't wait to see the new Tesla carbon wrapped motor being featured on this channel.
it will be like defusing a bomb. The carbon fiber is so tightly wrapped that when you cut it the motor inflates twice its size!
Lol.... carbon wrapped rotors have been used in applications over 200k rpm for years now, nothing special to see here.
@@carholic-sz3qv The bigger thing is that it is being mass produced I suppose by a machine created by themselves. Will be interesting to see if there's anything else new with the motors too though of course. We've seen the 3/Y motors with Munro but never S/X. :)
@@carholic-sz3qv Except these Tesla engines are part of the fastest production vehicle ever built. To me that's kinda special.
@@alaricgoldkuhl155 really?! Stfu! The model is quick but not fast. Impressive things are buggati chiron(fastest car in the world) with 300mph, mercedes unimog with tousands configurations possible for any applications, a scania 770s.......
From what I have read Ford is not building these electric motors, BorgWarner is supplying them as complete modules. Love the tear down videos, keep up the great work.
"Code doesn't cost much, and it weighs nothing." Yes, indeed.
and writing a code is 1 time cost and can be used a any number of vehicle without additional cost
Program design is easy, code coding can be hard. Firmware is always changing since the hardware is improving and new features are added.
Code is one time expensive, and maintenance of code is not cheap. But no weight.
@@aitorbleda8267 If you write the class that handles the parking brake the right way, you do not have to maintain it, other than maybe improving it overtime. I'm petty sure making modifications to the code and much, much cheaper then producing all these parts for every single vehicle, plus - if you want to improve the behavior of the mechanical part that means you have to replace it in most cases.
Validating it and making sure it passes all the right standards however costs a fortune.
Ford didn’t design either the inverter or the motor on this vehicle. It was all done by BorgWarner, hence the non-Ford spec fasteners
@AnalogueKid2112. Then you probably know that both Ford and GM farm out the vast majority of their design work to the Detroit design infrastructure and beyond ("beyond" being for example the Chevy Bolt having an LG Drivetrain). You can tell that Tesla has control of every detail and are designing organically/systemically, there is design iteration being co-ordinated from a central locus, and even often manufactured in-house. US legacy manufacturers are full of teams that oversee parts design from a zillion vendors off-site and around the world. Fords look like design-by-committee, GMs look like Sears appliances, where contracts are let based on spec to the lowest vendor bid. Teslas are more like the classic band, Boston...where a really smart guy designs all his equipment, plays all his equipment himself and it works really well and sounds different from everyone else.
love the name!
@@johncahill3644 Spot on. This is how we ended up with the now "infamous" Six Liter Power Stroke. A really good engine that got butchered up in an effort to get it EPA qualified (not a bad thing in itself) by several committees. In so doing they nearly killed a legendary Ford cash cow.
I thought it was Bosch?
“ There are a lot of reasons they had for that, but I can’t think of one.” 😎
I think 🤔 that engineer was gonna do the 1 gram everyday reduction plan and he never got around to starting it, lol 😂 He was set for complete development cycle with that housing !
@@maples328 He was quite a fox designing it heavy so he could easily achieve the target of mass reduction and get boni on a regular basis. 😂
It's a big-block motor, see? ;)
@@itekani You said that perfectly
He can't think of one. Probably because he's clearly not an engineer.
6:48 "You don't really need that second washer ..." The washer is used because the shape of the under head of the bolt is typically slightly conical (not flat), designed for mating with harder steels. You need a flatter surface as found in the washer to make the torque reading more reliable on installation and not cut into softer materials such as low carbon steel lamination or aluminum.
15:40 Most castings, like most injection molded plastics, have draft angle. If they don't get a surface perpendicular to the spring surface, it will want to walk out or shift the coil, possibly breaking the silicon carbide transistors or the coil fins. I suppose they couldn't put the mating draft into the spring (probably poke yoke issues during assembly), thus the "standoffs".
Won't the standoff want to walk out (just like the spring would if the standoff wasn't there)?
I'm pretty sure that spacer at 15:40 would make sense if they had understood the order of assembly. The gate driver board with mounted FET modules is installed, with the FET modules slotting into the cooling plate assembly. Then the spring and wedges/spaces are installed to apply clamping force to the stack.
Either way, it is a shitty build lol
Large circuit boards subject to vibration tend to flex in not supported spans as a result of vibratory accelerations. This flexing can crack solder joints. Solutions are to use thicker PCB material (high cost) or provide more supports over the span. We did larger PCBs and vib testing will kill them if you don't have the needed support points. Brackets or support rails are a typical solution. I believe that the vibration environment of the car and associated motor is the reason.
I don't see how someone could challenge the intelligence of a design god that understands everything man-made that moves. He could've saved every OEM billions of dollars by using least amount of components to achieve highest level of design perfection. Thing about how much all these money can do to world peace.
resonant frequencies are a b1tch... a simple interaction between a couple of fans and a hdd can make a pc case vibrate so loud you'd think it will disintegrate. :)
also by the look of the board at 25:56 the fasteners appear to be thermal couplers and the plate that mounts on top is a large surface heat sink.
Could it be that the MOSFETs are soldered to the PCB before insertion in the hatsink assembly?
Yes, that's undoubtedly how it's actually done. I imagine the MOSFETs are attached to the PCB, and the entire control circuit side is assembled before being fed through the cooling plate with the TIM applied to the pre-applied MOSFETs. This makes it a very simple process that isn't very error prone.
The top side assembly could be largely automated then... one tech installed the first plate, sends it to a machine which handles all the welds, then it goes on to the next station... insulation is applied, next plate, back to the next welding station, etc...
@@looncraz Makes Sense! Still a lot of steps and welding and soldering during assembly time.
@@AndrejGobec Sure, but that's not much of an issue for this particular component since each step is basic enough you can train replacement workers easily and quickly and run several production lines at once. The numerous fasteners are the greatest issue, as always.
Exactly, I'm pretty sure Sandy got this part wrong. Never the less, the design of the inverter could much more elegant and simple.
@@looncraz The 3 welding steps they described probably go in different sequence - put both coper contacting sides at one time maybe manually, then machine welds it, puts the middle top part and weld it again.
*SANDY* don't like bolts... Awww... the *Chevy Bolt* was doomed from the start.. LoL
BEST! LOL
@@gunnar4200
Sandy- I hate bolts
Me- whys that?
Sandy- They always catch on fire
I'd never have guessed buying a Tesla 2 years ago would lead me to watching RUclips videos about screws...
Ditto, love my MS.
Fasteners are Fascinating.......lol.
@@markplott4820 and boos for screws
One of the important things that Tesla and SpaceX is doing which is completely over looked is inspiring people in all sorts of areas. It's getting young people interested in engineering, electronics, even construction with the drone videos of the factory builds and so on. My wife teaches primary school kids and often shows them videos of rockets and other such things and the kids love it. It's important to get those kids to want to learn but also injects some positivity in their lifes and future, which is often lacking.
How true! It is amazing how fascinated we have all become to see the competition. Amazing research for investors that is “free”
Looks like your team does not have enough power electronics and motor drives background since you focus a lot on the mechanical and manufacturing part. For a motor with or without oil cooling, the difference is the cooling efficiency which determines the size of motor under the given maximum power. So the Mach-E definitely has higher power density than vw ID. For a inverter, the most important indexes are power capacity and control accuracy. The relevant parts are power module, cooling system, current measurement and position sensor. Looking forward to seeing these information in the later e-powertrain tear down videos.
BEST episode yet!
Technical, engaging, humorous and with a few quotable gems.
Love it!
I love that the outside of the Ford inverter casting has "FoMoCo". I'm sure it's supposed to be "Ford Motor Company", but I immediately thought of "Fear of Missing Out Company" lol😂
Excellent !!
That has been on Ford parts for decades.
@@Miata822 These idiots have never been under a hood before.
Ford has engineered this car to have "a lot of opportunity for failure." Ford dealerships are very pleased because they see big $$$$$ repairs!!!
Imagine the joy on a dealers face, when someone comes in with a rattle in the inverter, and occasional cutting out. " Oh yeah, we get that a lot, someone's dropped a screw inside the inverter." That'll be $1849.63 please.
Have you ever heard of an EV inverter failure? I haven't, but I'm sure it must have happened somewhere. Parts like this are replaced, not repaired. The three copper strips Sandy complains about allow inverter replacement w/o opening the motor housing.
@@chrisheath2637 or a short - no one will hear that tiny rattle
This why I do not buy anything from Ford....only an idiot buy a electric car from Ford.
@@chrisheath2637 But before they repair it they will try to tell the customer that the noise is normal and a reboot is nothing to be concerned about.
Thank you so much for showing these insights here for free. There is so much any engineer can learn from that, even in totally different industries. Keep up the good stuff!
25:25 Why did they pick 16 standoffs? I would guess that it was to eliminate a vibration harmonic in the board.
That base part is so demanding to assemble that I expect even the robots are going to walk off the job,
🤣🤣
Sandy should give us a total count of threaded fasteners needed to build that assembly
Was that designed and built by a third party? The labor of assembly and all the extra parts!
@@billy-go9kx It's a Borg Warner IDU. B-W did the overall design, subcontracting out the design and part of the manufacturing for the motor and controller. The case and gearing was done at B-W, and the final integration was to Ford specs.
@@1djbecker You'd think that a Borg collective effort would be better than that.
4:49 I think people refer to moving parts, the rotor is moving but the laminates do not move individually.
^this 100%
Exactly what I was yelling at the screen lol. I understand his point. People do maybe oversimplify the making of an electric motor. But it does have massively less complexity and individual moving parts
The multitude of stampings in the body shell are all welded together and do not move individually. they are still in the count of parts as an inseparable assy.
Most people don't know about all of the independently moving parts in common machines.
How many moving parts are there in a hydraulic lifter?
What about the variable valve timing actuators in your favorite ICE engine?
And the throttle actuator?
You also never need to disassemble the laminates on the rotor for service. That's like saying babbitt bearings are multiple parts because they are layers of metal.
Thank you for the update and all the excellent work!!!
Why so many screws on PCBs? Mikeselectricstuff pointed it out : it's for avoiding mechanical resonance issues or prevent any flexing on these boards, BGAs don't like this at all. And this inverter assy is exposed to an extreme vibration environment.
Couldn't they have used snap in standoffs to do the same thing? I've been involved in a few product designs and we only used screw standoffs for prototyping. I can't think of a single case where we used them in production.
@@videcomp What are you prototyping?
The design team likely has a spec for the area and weight between circuit board supports for each environment. Since this is rigidly mounted to the drive unit, it's a high shock and vibration environment. Without the freedom to select a thicker, more rigid circuit board stack-up (which creates its own process change problems), the solution is more clamping supports.
@@6AM_YT Latest project is a smallish remote crawler with manipulator attachments for making repairs inside sewer pipes. I'm mostly providing guidance, experience and refining the requirements. My engineering student intern is doing all the heavy lifting. He's been working part time for me since high school and can now out program me in one language and is definitely faster using Fusion 360 for mechanical design. So he is doing most of the hands on work.. I actually quote Sandy to him and say "look we can leave that screw in for now but remember you'll have to figure out how to remove it from the next iteration." I like Sandy's "we call them un-fasteners" . We try to limit them to where a tech might need to access a part for troubleshooting, repair or maintenance. We try to avoid someone accidentally breaking an assembly because they can't figure out how to disassemble it, so we use "un-fasteners" because that is what they are for, not for assembly purposes.
There are many better ways of preventing flexing and fixing a pair of boards and a shield between them. Especially as you have a ridged cover on top of them. With proper design maybe 10 screws for both boards.
Another area of concern is the high number of soldered interconnections between circuit boards. I have seen solder joint failures in the form of “fatigue cracking” (possibly originating from a poorly prepared solder joint) on automotive circuit boards that could potentially occur here over time from motor/drivetrain vibration subjected to the circuit boards. It would be beneficial to reduce part count by integrating more of circuitry onto one or two boards at most.
Or use leaded solder again...
@@dangraff8467 They're either wave soldered or the parts have solder "paste" and are heat flowed, but they are flowed in one shot.
@@dangraff8467 You can weld certain things, but when you are attempting to electrically attach delicate components, it's not easy to contain the current or magnetic flux welding produces. It would also be extraordinarily time consuming and expensive. Plus, welding is impossible on ball grid ICs of which many are used these days in dense boards. Properly soldered circuit boards should easily last 30 years in a car environment from a physical standpoint (including shock and thermal stresses).
@@dangraff8467 The subject though was solder connections on circuit boards which was all I was addressing here.
@@dangraff8467 Thicker, more thermally stable circuit boards can improve long term resistance to solder related failures, but the old leaded solder was less prone to crack. Also, proper ventilation of the board and arranging the connectors so they don't get stresses on them which help accelerate solder failures. It comes down to the manufacturer and how much they want to spend to extend life.
these technical episodes are solid gold!
What's wild is, this sort of deep-dive, technical episode is really for me as an EV customer. I would never buy a Mach-E after seeing this. I wouldn't want the "many opportunities for failure." I like the idea of buying a clever and efficiently manufactured (lean design) EV.
Skewing magnets is done to rduce torque ripple because of magnet flux linkage to stator fin poles. Optimal total skew is always 1/2 * 360 / Stator_slots
degrees. It also smoothens harmonics from motors back EMF, but reduces your torque output by max 10%.
Excellent episode. Happy when Ben showed up.
This is funny and informative. Thanks for posting these teardowns!
32:30 I love your sense of humor Sandy!
13:10 i was told (FWIW) that we dont use the disk brakes because in wet weather int he winter or in really cold temps the pads can stick to the rotors and take while to thaw in the morning
Key difference to highlight with part counts is moving parts. Ie parts that wear.
the IGBTs used in this inverter are called dsc (double side cooling) which have been extensively adopted by Japanese suppliers such as Denso for Prius for years.
They are not IGBT's
@@vidznstuff1 These are IGBT's. Can confirm 100% that there are no SiC MOSFET power cards in production at Ford today.
Couldn't the MOSFETs be mounted to the PCB then inserted into the heat sink, before the welding (soldering) is done on the power side?
That's my guess, the fets are soldered to the gate drive board and they drop in from the top together.
On the other hand, the control board must be soldered in on the line so who knows.
Yep. Honestly, them thinking that board was installed after the FETS were welded in rather than it being populated and then installed into the heat sink really makes me reconsider the credibility of this channel.
@@petergabrielsson2013 The thermal expansion spring is installed only after the SiC MOSFETs are inserted between the heat sink plates, thus the separate bracket that they didn't understand in the beginning.
@@1djbecker makes sense
21:56 Is it possible they’re welding and pre assembling the mosfets and then sliding that entire assembly down or up into the cooler? If they did that process they could weld and pre assemble the electric components and bolt them together after they fact.
That's what i think too. Sandy just tried to assemble it in the wrong sequence.
They would solder all the mosfets to the PCB first, probably by dirt cheap wave soldering and then they insert it from the bottom into the uncompressed heatsink. They use a spunger to drop in the blank aluminium spacer and the Spring from the top then. Easy as pie.
Those boards get tested for sure with a fixture, pogopins, automated testing and the best would be to have the mosfets already soldered to it when testing so that nothing can get wrong in a later soldering stage. By the way, even this steel plate between those boards and all those nickel plated brass standoffs can be assembled beforehand as do the solder bridges between those two pcbs. It is one preassembled and tested unit before they insert it into the housing and heatsink plates.
I also thought about why they solder so much and do not use board to board connectors. One thing could be current capability, as a solder joint will always be better and more compact than any crimped connector. The other thing is deteiorating leaf springs inside the connectors. Also due to the vibration any gold plating in such a connector will shave off quickly and then it starts corroding the contact. Solderjoints can crack too, but there are solutions to alleviate excess stress like the fancy curvy Pins they used in the Mach e front Inverter.
A very large single PCB would be prone to flexing and cracking due to vibrations, but i would have preferred Rigid Flex PCBs over all those stacked soldered connectors. Also there are a lot of options of floating spring loaded board to board connectors, which are suitable for a high vibration enviroment (hirose, amphenol, etc.). But i guess this is where they try to save the cost.
This is what I’ve been waiting for. The heart of the machine
But, being from FUD motors, I was expecting a ICE engine..... .lol.
24:50 I am certain this pcb is mounted first and after that the mosfets are placed one by one and soldered. Or the mosfets are soldered on this pcb first and then this pcb is mounted sliding thos mosfets between the cooling ribs. Both ways are much simpler and are way faster as it will 'auto align' the parts.
Thank You for the updates! Excellent work!
Glad you like them!
Unless Sandy understands a design instantly, he implies there is an issue.
This may not be the case. What if there were reasons things were done a certain way? For example, I would expect those soldered connections to be more reliable than a connector, especially for those high voltages.
He is a perfectionist, he brings up valid points but in the end none of it matters, as long as the customer is happy, and the device functions. Nobody typically cares about how their car is assembled.
@@slayyou Until it starts having problems after the honeymoon is over. Triple the amount of coolant hoses, parts, brackets, clamps, etc. as the Model Y -- can't be easy to repair. As much as I like the Mach E's interior and driving experience, I'm going with an Ioniq 5 or a Model Y. Just as fun to drive and better engineered.
Maybe the mosfets were soldered onto the board first? Circuit board side assembly first then the side you started with?
This knowledge is worth so much money! Thanks for sharing freely!
I just saw this on my feed. I liked and now I'm watching.
Awesome! Thank you!
If you missed the Tesla tear down videos - I highly recommend you watch them when you get a chance. Informative, educational and entertaining.
@@videcomp Watched them all. That's when I started watching Munro.
About the inverter assembly, I think the Mosfets are first soldered in place on that PCB, then slide into the cooling thingy and the side on the PCB is assembled first, then flipped over and all the bus bars are put in place and welded.. This prevents to get these 90 or so little pins to line up with the PCB, which would indeed be a big headache for the operator assigned to this job.
It even could be that the PCB is already fitted with the mosfets from the place where the PCB is assembled, and this could be automated and makes sure the solder connections to the Mosfets are high quality and consistent.
Regarding the parking brake, in some markets (ie the EU) it is a legal requirement to have an independant mechanically interlocked parking brake on an automatic car, which an EV falls under, despite having no actual gears to change.
It also tends to be used to ensure the safety case for "un-intended movement" for which an EV is particularly vunerable as compared to an ICE. With an EV, if the battery contactors are in, then the vehicle can suddenly and un-expectidly move if the inverter applies torque. For an ICE, the faults required for the engine to start, fuel pumps to run, gears to engage and the vehicle to move un-expectidly are just about impossible to occur. Not the case with the EV......
And imagine your car is San Francisco and catches fire, and start rolling down the hill, because the software just burned ...
I was wondering about the regulatory ramifications of using a mechanical parking pawl.
@@Fix_It_Again_Tony Legislation doesn't demand that it's a traditional park pawl/park lock - many OEMs are now just relying on the worm-gear driven e-brake integrated into the foundation brake calipers.
Never saw Sandy this depressed talking about engineering... 😔
Good job, Ford! 😂
exactly my thoughts haha
I feel like sandy is too focused on cost reduction. Faulting Ford for having an oil cooler while praising others for not having one is nuts IMO. Give me the oil cooler any day. Or welding a high voltage electrical connection over a snap in. I’ll take the weld. Especially in an application with plenty of vibration. He knows his stuff, but this mentality is why we get cheaply built crap. It’s anti-consumer and pro bottom line for the company.
@@AustinFerguson the build quality of the Electric Motor in the Mach-E looks way more advanced and robust compared to the Tesla.
19:07 MECHA-MUNRO!!
Yep...I was about to post the same comment
Great show - camera and edit getting better to - Thanks
Great to hear!
"There was not enough room to get SEVENTEEN in there..." LOL! Sandy is RUTHLESS!
Also, holy crap - lining up the board with 90 pins below = bent pins on every unit 100%.
Definitely at least 25 dropped screws rattling around in every one of these...
Great work Sandy and Ben!
Sandy looks at motor: "I see nothing but heartache and grief."
It reminds me of why CPUs moved from having pins on the CPU to having pins on the motherboard. Bending those pins was extremely easy!
@@AudiTTQuattro2003 Agreed.
@@AudiTTQuattro2003 - The more parts and clutzy design the greater the chance of future problems.
@@AudiTTQuattro2003 what time is necessary? they have the competition right there next to it. You can SEE the differences. And the performance numbers are already known. Ford's motors are 'the best' at nothing I've seen measured
That was awesome. The bedazzled looks of the design choices when re-assembling the inverter :-D
Thanks you Munro Live Team!!!
The MOSFETs would have been soldered to the board first, then lowered into the cooler, the flipped to weld the tabs.
Thumbs up JUST for the Intro! I love the individual layout on the table stop motion part!
Nice video.. love watching. One thing is interesting though. You are assuming they don't mount the circuits to the silicon board first.. if that were the case there would be no problem aligning a few welded connections..
@3:45 Wow! Cool AF... I'd never seen that before. Thanks *Mr. Munro.* (I love it when I learn something new. 🤓)
Can't the MOSFETs be pre-soldered to the board and the whole board go in *first* into the radiator from the bottom, *then* weld the bus bars?
EDIT: Saw the second board...
They're laser welded. And yes it looks like the mosfet and Buspars could be pre-assembled, inserted and bolted down as a single unit.
Really enjoyed this. Thanks for putting it together.
8:58 I'm sure it's supposed to say "BorgWarner" for the name on the stator housing, but it really, really looks like "BongWarmer"????
That's what she said LOL.......
Good cold weather performance then?? 😁😉
I have a front seat, and I am ready to learn.
Watching SANDY, is like watching Professor PROTON on Saturday morning...... Lol.
“There’s a lot of reasons I’m sure they had for that, but I can’t think of one right now.” 😅🤣 8:49
EMI shielding could be indeed the reason for the countless screws in the inverter box. The high power impulses from the switching mosfets can have very high frequency content. To shield those high frequencies you need to have much surface-area to contact the shield parts together. Too little contact-area between the shielding parts renders the shield useless because of skin effect and high parasitic inductance. I saw that happen many times because of oxid layers on the aluminium which acts as an insulator. So in the first glance you think you have a wonderful farraday cage but with all the oxid on your aluminium parts you only have a partly closed cage which is not obvious. So you have to take care of it as an engineer.
Just an educated guess....
That also explains the fancy coated screws, thanks for the tip
Shielding for EMI can be a real headache, no?
I don't know if they used used a conductive gasket on the lid, perhaps a molded rubber one with a woven metal strip around the entire perimeter.
D. G-S, how do the other manufacturers get around that without the mess that we saw here? I'm sure there's a way to design this better, especially given that we've seen evidence of this being done.
If that’s the case why not use a RF gasket sealing arrangement?
24:30
Wouldn't the transistors be soldered onto the board before and then the whole stack is being put in between the cooling fins?
I reckon this is the first step and then you do the other side.
I agree with this. I don't think Sandy has the order of assembly correct.
i think so too. Nevertheless, I think the layout, which has to be worked on from both sides, is bad
and regardless.. why is Ford the only one doing it this way? It doesn't seem like, regardless of the order of assembly, that this is a very intelligent and efficient way of making this component and connecting it to the motor. Looks like if they took the time to see how others were already doing it, they might have come up with more efficient methods
I don't think so. Look at 16:22 when Ben inserts the slotted MOSFET holder. That component needs to be inserted from the side slot and if the top had been assembled before, then the MOSFETs would get in the way of sliding that component into the spring.
@@Zedus-rl9hp It's almost certainly put together standing on its side, so you can access the top and bottom without having to manually flipping it over.
Please, please, PLEASE!! A Sandy Munro + AvE teardown would be comedy gold of epic proportions. The best of engineering snark from both sides of the border. Please make it happen!
OMG, can you imagine Sandy and AvE together. I would pay to see that.
@@MegaPraystation if AvE would hold firm on never being on camera, we would only see AvE's hands. But I would be so happy just to hear the two discuss. I am sure Sandy and AvE would have an amazing exchange over something like the inverter shown here.
We would have to demonetise that video, as I think AvE would bring out the bar room Sandy, negative filter. 😂
You "never see" brazed plate heat exchangers on EV drive units then walk right over to a Tesla drive unit and lay your hand right on their brazed plate heat exchanger. (???)
Yeah, had a big WTF moment there for me too. Not sure what he was trying to get at there.
Ford's heat exchanger was inside the case (noticed from the short clips in the opening sequence). Tesla's is outside. That might be what he meant.
@@charlesball6519 In the second brief clip barely one second into the video you can clearly see it on the outside of the drive unit (upper left in portion of the image).
I love these teardown videos!
💪🏻😃😃😃😃😃😃😃😃😃😃👍🏻
"If I was that cover, I'd feel dejected." - Sandy Munro, the greatest Engineer Ford forgot about.
Also, did anyone notice the logo Ford had on the casing? "FoMoCo - Guess their FOMO is shining through in these hasty design decisions . . .
I see what you did there,brilliant.
I was hoping someone else would catch that!
Virtually every part in a Ford Motor Company vehicle is marked with "FoMoCo"
@@bradcavanagh3092 So you are saying Ford invented the Fear of Missing Out?
Classic. One of the best tear downs.
I'am a retired electronics engineer, when ever I bought a new pice of equipment and it was held together by screws, I would always open it up to see what's inside, since if the used screws they must want you to take it apart.
Wow, actual footage of teardown is nice touch
8:15 Perhaps it could have been made more compact, but they need to pass the connection through a watertight seal, and they did it with that component.
24:30 I was thinking they'd install the MOSFETs with the same pick'n'place equipment that populated the rest of the board, then drop the board into the inverter casing, slotting the MOSFETs into the heat exchanger. THEN you turn it over and do all the goofy welding.
Note also that the inverter is probably assembled in a separate plant, most likely in a low-grade cleanroom. Still a pretty wacky design, I'll give you that.
Doesn't the PCB, on the other side of the inverter, is added first?
I can believe that it can be mounted in the order that you guys show!
Yeah, there's no chance it was actually assembled in the order they showed.
It seems each sub component is designed by a different person or team. Then when they meet for the 1st time, to assemble, they have another team that tries to get the subcomponents connect with additional sub components (connectors).
Yes, while those others are in their intern meetings not disturbing the other engineering departments which are stupid anyway...:)
Like the 4 blind men holding a different part of the elephant.
Loved that exploded teardown intro , very cool .
That inverter is a work of creative artwork.
I think the comparison between the parts on ICE vs electric are MOVING parts. Those 500+ parts in that rotor are essentially one part, since they don't move independently.
don't forget to count those capacitors too 🤣😂🤣
Thanks for the show ! IGBT insulated gate bipolar transistor...
Wrong video, buddy - this doesn't have any
@@vidznstuff1 Mentioned in the video. They were earlier switching semiconductors, and now they are moving to silicon carbide Mosfets, which have advantages, but are more expensive.
I never tire of Sandy's humor ☺️
@Munro Live: Could you have the assembly sequence wrong? If the MOSFETS fall through the casting, they could be inserted from either side. It seems to me that if you started at 23:06, you could have previously coated the MOSFETs, inserted them individually into the circuit board, used a precision jig with slots on the sides that duplicates the casting slot spacing to keep them in alignment, and flow soldered them to the board in one step? That unit would be easier to insert in the casting from that side than inserting the individual MOSFETs and aligning the circuit board with all the pins and the casting.
The only thing that would sort of justify the standoffs and aluminum plate is if they act as a power bus or a barrier shield between the two boards. The welding is still a mess.
The EMI shield may be to isolate between the two PCBs as noise fron the one below is highly inductive and probably induces false logic signals into the one above.
Thank you Sandy! always looking out for the cover plate's feelings of inadequacy!
Circuit boards are extremely vulnerable to cracks caused by fatigue failure resulting from vibration. They need far more support than a cover.
Munro Live is the Gold standard to EV reviews.
Thanks Andy!
With 7000 miles on my Mach E, the vehicle is running flawlessly. (Phone as a key has had issues). With 9000 miles on my Tesla model 3, I’ve had two visits to Tesla service for minor issues. Rear view mirror fell off and the charging door wouldn’t open. I’ll reserve judgement until both have had 50,000 miles on them.
My friend just sold her Tesla model 3 last Saturday and bought the Mustang Mach E and she said it's so much better to her. She had hers for 9 months and had several issues and she had enough.
That's a reasonable way to assess them from the consumer level. But from an engineering level (and I am NOT an engineer.. so looking at it from a limited perspective).. cost analysis and engineering complexity.. Ford is up and down from one area to another. They got some things done well.. other areas seem to need a lot of refinement. Not to say Tesla is perfect lol.. but a lot of their issues is the fit and finish an quality control that they need to improve on.. as a fairly new company, you can expect these to naturally improve and mature over time. They already have drastically. As well as their under body design, going towards the mega castings. As for Ford. I am not so sure these engineering challenges just improve naturally. My concern is these are things pushed on these engineers based on Fords culture of operations. Silos.. bean counters.. parts bins.. management making decisions when they have no real interest in what's best for the customers. Not a fan of the MBAs making decisions on the guts of the cars.
Has the phone as a key issue been resolved with the Mach E? Also, what year is your M3? I've heard early builds had a lot of issues.
@@jodrhh yes the phone as a Key issue has been resolved to an extent. But every time there is an OTA update it seems I have to reinstall the phone as a key. My model 3 was purchased in December 2020 but is a 2021 model year. The Mach E was purchased in February 2021. We wanted to replace our 2013 infinity jx35 and the obvious choice was the model y buy we decided on the Mach E after test driving both.
23:30 Wouldn't it make more sense to put the circuit board on first, so you plug the MOSFETs in and they hold the MOSFET in place for the welding and assembly?
I'm not sure what I like better: "cranky" Sandy or "snarky" Sandy. We got both in this episode!
I was hoping for BLOTTO Sandy, that would have been fun ......... lol.
Like them both as he needs to poke the Ford team to do better, and the fact that they have turned their backs on many of the things he implemented when he was there, well, I don't think he has much sympathy for them.
The inverter was good fun, thanks :D
You're welcome!
This shows a term I call " Engineering Inertia". When you design and build something for so long that you become comfortable and complacent. When it is time to design and build something new it is difficult because you are transfixed on the old designs.
wut, lol? That coaxial axle setup is extremely advanced. Tesla's geartrain is from a steam engine.
Thanks again Munro team. Others may have mention this before but Munro team is like Shelby & Miles from Ford vs Ferrari. Trying to do real engineering but try to play nice with Corporate, everything explaination have to have a positive spin to not ruff the feathers.
Stay honest guys.
The parts count relates to MOVING parts, designed to move relative to each other. wearing each other out, needing complex lub.
Wow, that inverter is NOT designed for volume production! It's like the designers were told "prototype only", but now they're moving it onto production with zero optimizing. Ford does not expect to need many inverters!
No, not many, they will make new versions for all model variants, of which they will have to have as many as GM promised, 30 by 2025... for total 1mio cars, lol, so few thousands each will do.
@@praero551 They won't make any profit that way.
I actually work on those inverters. They are made by Denso and we are adding lines, but we just don't have enough people right now.
@@caseybv74 Ford CEO most likely pushed to market, hard to optimize when in production and this even would be in basic design not just moving operations around
The thing is even though the 500 plates are not moving relative to anything else. They still have to be assembled carefully. Just like an internal combustion engine needs to be assembled carefully.
Just like anything else that's built by people, some are built better than others sometimes the design is better, sometimes the manufacturing is better.
I drive a Toyota Prius which is a combination of internal combustion and electric and shares many of the components of all the other EV'S out there. It has a small battery, high voltage inverters and converters, coolant pumps etc.
If you follow the Prius forums you will find out that the gasoline engine is the most reliable part of the car.
At least in gen 2 cars. 2004 to 2009.
You almost inevitably have a coolant pump, inverter or battery failure before the gasoline engine has any kind of serious mechanical problem.
I suspect when electric vehicles start turning 10- 15 years old you're going to start to see some of the secondary support systems begin to fail. You lose a coolant pump and you're on the side of the road, same as in my prius.
A little $50 12 volt pump can strand you. The difference for me at least currently is I can go anywhere and purchase that coolant pump and install it myself. And it's relatively easy to do. If I had the pump on hand I could probably do it on the side of the road if I really had to. Two hose clamps and three 10 mm bolts and an electrical connector. That's it.
There's no doubt that electric vehicles are the future. But I suspect we're going to find out that they break down almost as much as internal combustion cars. Particularly if they are assembled badly, or the manufacturer decided to use a cheap crappy supplier for critical secondary systems like a coolant pump.
Most of these electric vehicles depend on their air conditioning system to keep everything cool. They have a glycol loop like an internal combustion car but that glycol loop is passed through the air conditioning system to get rid of the heat. So you lose an AC compressor you're probably in limp mode.
thanks for your efforts! amazing to see how its all build inside. Super good information, no idea how it all works but very interesting.
23:05 “and for my next act, I’ll be flipping a 3 egg omelette single handed!” Love your show!
At 24:00, would it be possible to have the MOSFETs soldered to the lower PCB as part of PCB manufacturing, and then rather slide the MOSFETs all at once into the cooling fin during mechanical assembly? It will still be a lot of fine alignment needed, but not 90 pins. That implies that the lower PCB would need to be the first electric/electronics part to be mounted into the housing. Soldering of the MOSFETs to the copper plate bus bars would then be after the lower PCB is in place.
IFord could use engineering managers with a big overview about how things could be done, and an organizational structure that would lead to the simplified design vision.
HIll-Rom hospital bedside table as a bench top lol. I use the same thing. Great video today. Thanks!
As an R&D Engineer who practiced the "Less is More" philosophy in Automotive, Military, Aerospace and Medical devices, that Inverter is what I would call a real Cluster**&k!😱
I wonder how much that would lighten your wallet to replace?
Assuming you vehicle did not turn into a useless lump of Slag after the Inverter melted down and took everything else with it!😬
A definite Steep learning curve at work here. As long as they keep making timely progress.
Before they are put out of business.
Mike in San Diego.🌞🚀🎸🖖
About Model 3 Oil pump. Mine had oil leak and replaced under warranty.
That's the argument for VW removing it.. However.. VW's motors are FAR less performance oriented then Teslas... like A LOT. I sincerely doubt VWs method without the pump would provide enough cooling for the performance motors Tesla uses. Even on their non performance models, Teslas motors are WAY higher performance than anything VW is using
Tesla oil pump leaks are the best and the most finely engineered oil pump leaks in the automotive industry and legacy automakers are just haters because they can't make their oil pumps leak as efficiently and autonomously.
@@airheart1 well..in the base AWD version of the model 3 for example the front motor is 121 kw, the rear motor is 203 kw. In the ID 4 the (rear ) motor is 150 kw, so in the middle of the two. So it could be at least worth a consideration for the front motor in Tesla AWD cars.
@@tzarcoal1018 I suppose there's a point there. But I wonder what the RPM differences are that those motors are experiencing. Even the smaller front motor in the model 3 AWD may be at higher RPMs, requiring more cooling then the VW motor does. But Idk. Certainly something Tesla should look at.. if it's possible to remove a part.. that's Elon's thing lol
@@airheart1 The motors are direct drive. RPMs will be the same if the wheel circumference is the same. The difference in power output is small and doesn’t explain the need for oil filters on Tesla. Audi isn’t using them on the E-tron either.
This e drive was developed and built by Borq Warner and will likely be sold to other oems. The Lightning will use a Ford in-house designed and built motor assembled at Ford Van Dyke electrification center.
Probably with "RIVIAN" embossed on the casing (😉)
Makes sense Borg Warner components have always been mediocre at best.
Yeah I was looking for BW commentary.
Why should anyone believe someone who thinks it's BORQ Warner?
Hello Sandy and Ben, HOW ARE YOU?? Interesting episode today on the Mach-E motor, but I gotta be honest with you, it's way above my technical knowledge, but I'm sure many electrical engineers must love it!! Funny thing though, as I'm watching this video, my wife brings in the mail, and there's a box from Munro L ive with my new Munro baseball cap!! The tracking said it'd come Saturday, but it came a day early, THANKS!! It's a nice cap, black with mesh sides; it's very well made looking; it says "YUPOONG" on the label, I guess it's Chinese, I know Sandy has a lot of contacts over there, and while it appears to have Sandy's autograph on the visor, I assume it's silk-screened on, since Sandy must be WAY TOO BUSY to autograph everything he sells, but I'm happy to have it on my hat and will wear it with pride from Sandy Munro, whom I call the WILLIE NELSON OF THE AUTO INDUSTRY!!! Chas Orvs, Plainville, Ct.
I wouldn't be surprised if the inverter assembly involved a lot of jigs with pre-loaded parts. I also highly doubt that inverter assembly is anywhere near the vehicle assembly line. As in, probably another country.
Lots of extra jobs... UAW probably happy.
The Mach-E is assembled in Mexico, but maybe the inverters are assembled in a UAW shop. Somehow Ford still got invited to the Bidden/UAW EV-party.
I believe the MOSFETS would be soldered to the pcb in a jig first.
Still.. why are the other manufacturers not doing it this way? Why are Teslas motors still more efficient and powerful? I think Ford could do with learning some things from the others here
I can see Munro growing from exclusively breaking down vehicles to any piece of engineered technology for all the tech nerdz out there ;)
This was intense. Sandy you were visibly tired of all the parts and pieces.