2 ms·
Downconverting voltages is surprisingly easy - a single bank of MOSFETS and an inductor is all you need. A 2MW convertor made of [1] comes out to less than 10
by londons_explore 7y ago
Downconverting voltages is surprisingly easy - a single bank of MOSFETS and an inductor is all you need. A 2MW convertor made of [1] comes out to less than 100 grams (excluding cooling, but that's fairly small at ~100 watts).
There are also a lot of reasons to increase voltages in EV battery systems. The current ~400V systems need thick heavy internal wiring, and for all electric (ie. no mechanical) braking and ultra fast charging, they are insufficient.
[1]: http://www.ti.com/lit/ds/symlink/csd19537q3.pdf http://www.ti.com/lit/ds/symlink/csd19537q3.pdf
- baybal2 7y agoHave you seen voltage rating on that thing? You can't reliably handle 1500V+ voltages in solid state electronics without purpose made IGBTs, or exotic semiconductor materials. Your EE knowledge doesn't seem to be more than cursory.
- londons_explore 7y agoIt has a maximum voltage rating of 100V, so you would need 100 in series, with leak resistors to prevent any one avalanching. This is how most modern high voltage switches work, because IGBT's are far less flexible. Take apart a tesla Model 3 motor controller and let me know what you find...
- baybal2 7y agoTo begin with, Tesla's "motor controller" is an inverter. The charger on board is AC-DC. DC-DC It's an old, but still not completely solved problem: imagine your stack open, but for a single FET, and your resistor is a little bit off. It instantly pops, and the "tuning" for the rest of it is now completely off, and it will cascade further In case of Tesla, they do use the exotic semiconductor materials I mentioned above. SiC goes to 1500V nominally You need to learn electronics engineering beyond wiring Arduinos There is old, but still quite good and relevant explanation why stacking FETs suck on edaboard https://www.edaboard.com/showthread.php?269261-Is-MOSFET-in-series-is-possible https://www.edaboard.com/showthread.php?269261-Is-MOSFET-in-... There is an even older, but still excellent paper published by Philips called "Power Semiconductor Applications" where they explored nearly all trick to make FET stacks to work reliably. Short: it's near impossible to do it reliably with off the shelf part, and it's better to not to risk if you deal with 600A+ currents. And if are determined to do all tricks necessary, IGBTs or III-V materials will still win on simplicity and cost
- Kirby64 7y agoPeople have taken apart the Tesla Model 3 inverter and we know what they've found. Here's a hint: it's not a stack of 100V MOSFETs in series. Depending on the model (SR+, Performance, etc), it's either SiC MOSFETs or IGBTs. And there's only one in series, not 10 or 100. Also, please do me a favor and never design power electronics ever...
- Kirby64 7y agoYou understand that the device you linked will blow up if exposed to over 100V right? You're insanely oversimplifying what it would take to convert 10kV down to a usable voltage for use in EVs. Anything that can convert 2MW will be the size of a fridge, at minimum. If you could save a bunch of copper cabling by running everything at 10kV, people would be doing it. 800V is already pushing the limits of what most folks can do with technology we have today. As a reminder, 10kV can arc close to 2 inches in free air (depending on humidity, of course), which means you need to insulate the hell out of it anywhere it's exposed. That means motors are pretty complicated to use at 10kV. By comparison, 800V only arcs 0.1mm (0.004 inches) max.
- t34543 7y agoYes, and even a tiny penetration of the insulation will destroy the dialetric strength of the material, posing a substantial risk to humans.
- nlfwhulsdhouv 7y agoEvery step of your analysis is several orders of magnitude too optimistic. The only thing sensible here is conceptually going to higher voltage. Loss in these systems is dominated by switching, not conduction. Are you just using the Rds(on) figure? It's best to work back from efficiency for back of napkin analysis. At 99 % efficiency (very optimistic for such a large step down ratio), it's 20 kW of loss, not 100 W. That's a lot of power to pull out. (1) Rds(on) Cgs product for a 100 V FET is a lot better than a 10 kV SCR or IGBT. (2) The inductor for a 10 kV, 2 MW converter will be massive, lossy, and hard to insulate. (3) That MOSFET that can only handle a few 10's of W of dissipation before it melts off the PCB, even with forced air/water cooling. Times like these remind me that when people speak so authoritatively on HN they're usually full of shit.