[Long road to e-kart] Practical information about making high power circuits?

[Ciel]

Hello peeps, Ive been learning electronics from the internet (ikr) on and off for a month now, and Im set an “ultimate goal” so to say to build a 5/10kW electric kart? golf car? some kind of LEV with two pmsm motors and a custom controller/driver. But Im sure I have plenty to go before I should be working with either 100+V or A. I was also insistent on using sodium ion cells for some reason so there’s also challenges that come with that(72V pack goes from 95V max to 36V cutoff lmao, but ofc I don’t need to always extract everything)

I know basic electricity from studying physics in HS, and some design patterns and parts from strolling around the net, including a great series about BLDCs from Jantzen Lee, which taught me a lot of higher level concepts about controlling BLDCs(well, PMSMs but theyre similar), but there seems to be little in the way of more practical information about handling higher power currents through the pcb or how to connect to the thicc bolt terminal things, probably because it’s much harder to get right just for a video. I did find an EV inverter teardown video from DENKI OTAKU and that gave me some good ideas, but the details are still pretty faint. Im also slowly going through MIT’s 6.002 Circuits and Electronics(2007) on OCW though that’s more fundamental(and has countable number of pixels), maybe I should move to 6.622 Power Electronics (2023).

So in a way Im asking for two things here, for one Im looking for some resources to learn about the specifics of dealing with high power, safety and execution wise. Like Im looking for MOSFETs and they barely go above 100A from TI and seems to be limited to 200A by the leads from Infineon so Im not sure how to get some more safety margins, maybe the way is to parallel them? But that seems like a different set of wormholes. Overall I just don’t feel… comfortable? confident? with power electronics yet as I am with lower power, like I don’t know how to design a pcb specifically but I can’t be *that* much off, I feel like I can always get it working even if I need to hack it, at least I know what to hack. But with high power, I’m just clueless, like there are things that I have zero idea on what to do like connecting the components, I feel that I am missing still some crucial pieces of information to make it all “click”.

On the other hand, I would also love y’alls opinion on what I can do to prepare myself up to this big project. I got a gimbal bldc already because I wanted to make the SmartKnob, maybe I should do this first? What should I do next?

Also I should mention that I haven’t gotten a power supply or oscilloscope yet(other than an USB oscilloscope and a low? current .5A multi-voltage board) but I’ll get them soon when I get a space in a few months(space constraints in my home). I do have a decent multimeter and soldering setup though if that counts for anything. And since I mentioned this, equipment advice is also welcome lol.

Sorry in advance if anything was hard to understand, Im very bad at writing long passages and it can become decoherent.

[jwet]

Since you got no responses I'll try.

Your question is like saying to a musician that you want to write a symphony but you just need some help with the notes.  I think having an eventual goal to build an electric cart is good goal.  Here are perhaps some ways to learn.  It will be a long road.

Look at similar equipment and see how they're wired an built.  The fancy name for this is reverse engineering. 

You might buy a broken cart or kiddy car or electric wheelchair and get it going.

There is a good book called "Practical Electronics for Inventors" by Shertz.  Its in the right area for your needs.

As far as high current wiring, approach it from ohm's law that you got in HS physics.  V=IR, P=I^2R,R= rho*L/A (rho resistivity is 1.7e-8 for copper).  To move 100 + amp currents, you need 50 mm^2 area wire (7 mm diam).  This is from standard tables from the National Electric Code. You can calculate what the total power dissipated in this wire would be with those equations.

Your question is very difficult to answer but I guess it comes down to dedicate yourself to the problem and learn. 

[Ciel]

Thank you for the response, I know that I am far from my goals and I am inching(centimetering?) away at it. I think I will get a controller and disassemble it, it can also come in handy for testing, I had been thinking about it for a bit but it’s quite expensive for a nice one. I was planning on buying a small wheeled robot for a more robotics focused project of mine as well and I was planning on taking a peak inside.

Regarding the high current part, yeah I know I need really thick wires but that’s easy to do, the more burning question I have is how to get them onto a PCB without obliterating everything, high power mosfets/gan fets like to come in tiny packages cause parasitic capacitance and loop something that I forgot the name of, so there has to be a transition from “thicc wire” to pcb traces which uhh, R=ρ l/A doesn’t have the best A… and the 2221 standard tell me I need 30cm traces(iirc and it’s said it’s not applicable at this scale) so that’s probably not gonna work out, what is that missing link of “getting the huge amount of current onto a pcb, into the mosfet, and out of it? How do I ensure the traces dont turn into copper soup? Either my wording or idea is wrong because somehow I had been able to find zero information talking about this, at most mentioning in passing.

Also thanks for the book recommendation, Ill look into it.

[jwet]

If you take apart some car audio stuff or a hefty DC to AC inverter- you can see how they do it.  You use very beefy PCB and very wide tracks.  Board materials come in different copper weights- standard JLPCB stuff is 1 oz (weight/square foot)- this works out to 35 microns or about 1.3 mils thick.  But it goes up from the there- I've done work with 10 oz and I think there is larger- this 10 oz stuff would .013".  A 2 inch track of this would have have an area of .026 in^2,  that will do 100 amps if you go through the numbers or look at ampacity tables.  Sometimes you'll see the high current area flowed over with extra solder- lead is a decent conductor- 10x worse than copper on rho but decent.  This is common in audio stuff.  Often they will screw the board down to copper bus bars to get off the board and act as input terminals.  In cheap stuff, they use big ring lugs with large nuts, washers and bolts to the board.   There are also connectors for this stuff that you'll see on tow trucks or electric forklifts- 200 amp is pretty standard.  Solar installs have proliferated these high current connectors and there are some pretty nice ones.  Its just ohms law.  Milliohms turn into watts at these currents (squared).  You can look at the specific heat of copper and see what 5 watts would do to a cubic inch...  Physics all the way down.

There are tons of tables and guides for all this stuff, you don't have to go back to first principles but its not a bad idea to derive it once at least.  I googled and found Storm Power that has nice resources on line.  Have fun, get off the computer, go buy some junked power stuff and get some burns, that's is how you'll learn.

https://stormpowercomponents.com/technical-library/ampacity-and-specifications-tables/copper-alloy-c11000-busbar-ampacity-chart/

Use the thanks button, its how I measure my worth to the world.

[Konkedout]

I have been design engineering electronics for almost 50 years, and I have a couple of observations:

1) I recommend starting with lower power.  Not sure what sort of thing to recommend but if you want to do electric locomotion then maybe do an R/C toy car.

2) The greatest schematic diagram in the world will be garbage if the pcb layout is not good.  A good pcb layout is critical and its importance cannot be overstated.  In order to do a good pcb layout, you really need to understand what is going on.

Bypass capacitors generally should be the first components placed/connected to many ICs, but newbies will typically postpone or ignore that.

Check out any application notes (previously LT but now ADI) discussing "hot loops".  I am sure that TI and other manufacturers also discuss the issue but they probably do not call it a "hot loop".

Those are just the first two items that come to mind dealing with circuitry and power.

[Smokey]

Buy at e-kart kit if you want an e-kart. 

Just for a frame of reference, it would take a team of professional power electronics engineers quite some time (and money) to design and build what you are describing and have it work reliably.  Every part of this is hard.

[jwet]

Of course you're right- I don't think he really wants an E-Kart- he wants to dream about making one.  It really isn't about what makes any sense in a rational sense, this is aspirational.  I won't say it out loud, but its very likely, that this dream will never be realized.  That's not the point either.  Its just about thinking about this thing with a Sodium Battery that does things.  Its beautiful.

I have a dream to make an electric boat- I could buy one, likely better than my imagined object but it doesn't matter much either.  It won't happen but I won't admit that.

We are on earth to dream and strive- most dreams are never realized- this is ok.

[voltsandjolts]

search "formula student electric"

[eTobey]

The way to success start with a small step.

If you have some electronics knowledge, start with a small project. Parts are cheaper, and not that dangerous, if they blow up. (still dangerous with certain battery types).

[Ciel]

I have been design engineering electronics for almost 50 years, and I have a couple of observations:

1) I recommend starting with lower power.  Not sure what sort of thing to recommend but if you want to do electric locomotion then maybe do an R/C toy car.

2) The greatest schematic diagram in the world will be garbage if the pcb layout is not good.  A good pcb layout is critical and its importance cannot be overstated.  In order to do a good pcb layout, you really need to understand what is going on.

Bypass capacitors generally should be the first components placed/connected to many ICs, but newbies will typically postpone or ignore that.

Check out any application notes (previously LT but now ADI) discussing "hot loops".  I am sure that TI and other manufacturers also discuss the issue but they probably do not call it a "hot loop".

Those are just the first two items that come to mind dealing with circuitry and power.

Thank you for the insight! I'll be looking out in terms of PCB design.

I think I mentioned in a previous reply but I was planning on getting what is kind of a bigger RC car for a separate reason anyway, it'll probably be a lot of quality learning

Do you mind linking to these "application notes" so I know what kind of documents I should be looking out for? Thanks

[Ciel]

Of course you're right- I don't think he really wants an E-Kart- he wants to dream about making one.  It really isn't about what makes any sense in a rational sense, this is aspirational.  I won't say it out loud, but its very likely, that this dream will never be realized.  That's not the point either.  Its just about thinking about this thing with a Sodium Battery that does things.  Its beautiful.

I have a dream to make an electric boat- I could buy one, likely better than my imagined object but it doesn't matter much either.  It won't happen but I won't admit that.

We are on earth to dream and strive- most dreams are never realized- this is ok.

Buy at e-kart kit if you want an e-kart. 

Just for a frame of reference, it would take a team of professional power electronics engineers quite some time (and money) to design and build what you are describing and have it work reliably.  Every part of this is hard.

I did look a bit into FSAE/Formula Student info, and I know it's a ton of work to get an actually good kart and/or with compliance to join races and stuff. But my goal wasn't to make a good kart anyway, go-kart is just the form factor, really I just want to make some kind of high power thing and go vroom XD. Of course jumping to it is probably not the best idea, but I do aim to realize it no matter how unusable is turns out, just like a lot of other expensive toys I got lol. How does the saying go? You don't grow up, your toys only get bigger :p.

[Smokey]

You may find that electronics of this power level do not exist on a "continuum of functionality" that goes linearly from not-working to working with all manner of partially-working in between.

Roughly speaking, Power electronics at this level are either working or blowing themselves up.

[Ciel]

You may find that electronics of this power level do not exist on a "continuum of functionality" that goes linearly from not-working to working with all manner of partially-working in between.

Roughly speaking, Power electronics at this level are either working or blowing themselves up.

Good to know, thank you for the advice