Proposed improvements to SiC MOSFET power converter technology overcome existing challenges:
Transistors, which are devices that control or amplify electrical signals and power, are one of the most ubiquitous components of modern electronics. The most widely used transistor is known as the MOSFET, which stands for metal–oxide–semiconductor field-effect transistor. MOSFETs have been in use since the 1960s and have typically relied on silicon to act as a semiconductor. The latest iteration of the MOSFET uses silicon carbide as a semiconductor, known as SiC MOSFET. This has many advantages, but it has not been widely adopted for medium-voltage power conversion. This is because of several challenges associated with SiC MOSFETs that researchers hope to solve by combining novel technologies.
SiC MOSFETs offer many advantages over traditional transistors. Compared to the current technology, SiC MOSFETs can meet the demands of modern electronics by improving efficiency and power density. However, if you were to just switch out the current technology for SiC MOSFETs as is, there would not be enough of a benefit to justify the transition. In order to get the most out of SiC MOSFETs and expand their use across a wide range of applications, researchers employed novel control technologies and strategies to improve how the SiC MOSFETs work in medium-voltage applications.
[...] "For medium-voltage power conversion, 10 kV SiC MOSFETS have inherent superiorities, such as high breakdown voltage, fast switching, high temperature operation, and low specific on-state resistance," said paper author Slavko Mocevic, a researcher at the ABB Corporate Research Center in Raleigh, NC in the United States.
[...] By overcoming challenges like electromagnetic interference, high switching frequency, fast voltage transitions, and the need for high-voltage insulation, the SiC MOSFET technology can be more broadly applied to medium-voltage converters. Looking ahead, Mocevic said, "The immediate next step is to improve performance and fully understand the behavior of this converter to ensure stable operations in all situations for all targeted applications. The ultimate goal is to develop a family of medium-voltage circuit networks that utilize SCC and ICBT control that can fully utilize the power processing capacity of SiC devices. This will effectively tackle the lack of circuit solutions currently barring their adoption."
Journal Reference:
Design of a 10 kV SiC MOSFET-based high-density, high-efficiency, modular medium-voltage power converter, (DOI: 10.23919/IEN.2022.0001)
(Score: 0) by Anonymous Coward on Tuesday May 17 2022, @09:54PM (13 children)
solid state tech is magic :D
for the SiC, i would want them for induction cooking?
as for replacing transformers, say 22kV to 220V, well i just "see" the same real benefits as "smart electric meters" present: non for you and lots for whoever wants to control you.
.
i am sure in some fairytale reality, SiC are operating fusion reactors powering shields and warp drives and replicators, but here in our reality, they prolly enable trustfunds and bonds based on fleecing electrical users...
good luck anyways. also, in the silicon camp btw: nitrogen and silicon. spinning magnetic fields still need to spin on something.
(Score: 4, Informative) by Muad'Dave on Tuesday May 17 2022, @10:12PM (12 children)
They're used in Tesla power inverters [pntpower.com].
(Score: 0) by Anonymous Coward on Tuesday May 17 2022, @10:35PM (11 children)
oh, okay. cool.
maybe they're used when big Amps need to flow and survive their own inefficiences (temp)?
it IS facinating to imagine that its a single atom or very small group (?) of atoms that conduct and switch these large currents ... the equivalent mechanical relay is much bigger and sparks during opening/closing?
(Score: 5, Interesting) by anubi on Wednesday May 18 2022, @12:44AM (10 children)
My biggest fear is lessons learned a long time ago about solid state devices, inductive loads, fast switching transients, and the propensity of a solid state device to fail ON.
I haven't unlearned that yet. I fix a lot of broken stuff, and still come across failed power transistor devices, failed ON, with subsequent cascade device failures.
Even to this day, I still feel it ain't safe until I see a good physical air gap between me and the power grid.
I am presently designing a little Arduino to control an old cooktop/oven a neighbor tossed. It's electronics failed. Didn't like the way it worked anyway. But the rest of it is really nicely made.
It is way to easy to turn it on or change it's setpoints. Way too easy. Even a two year old can do it. And will.
I wanted interlocks, so I would have to know a bit about the device to even turn it on, even know a bit of math, to keep inquisitive kids at bay.
And once the controller is satisfied I know what I am doing, it will engage two electromechanical White-Rodgers industrial contactors ( same part I am using to control the glow plugs in my diesel van ) to supply 120/240 V to the heating element controllers..
But even with the latest SCR/TRIAC technology, I am very afraid to put solid state devices, exposed to line transients, in control of things that would lead to disaster should they fail ON. I mean what is a solid state device, designed for, say, 600 volts going to do if some big breaker opens somewhere, lightning strikes the overhead wire, some load dump somewhere, sends a whooping multi-kilovolt, maybe kilojoule, pulse my way? Will it simply fuse my PN junction? I know what will happen in the White Rodgers contactor. It will arc. Most likely extinguish after the transient event, but I expect the downstream SCR/TRIAC controller to fuse shorted upon receiving the energy that arced across the contactor.
If I still have enough intelligence left in the controller to successfully turn the White Rodgers contactors on, I should have sufficient logic still working to detect the SCR/TRIAC failure and immediately shut down, I will probably even use my favorite little fuse trick of deliberately blowing the control power fuse so subsequent attempts to power up will fail. The fuse will remember it has been blown. I learned that little gem from a microwave oven safety interlock. If you succeeded in opening the door with the magnetron ON, it would deliberately short out the transformer, blow the fuse, and the whole thing dead as a doornail until serviced.
They seem to have come a long way with MOSFETs and nondestructive transient avalanche technology, but my previous experiences still has me wary. What if this thing fails ON?
"Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
(Score: 0) by Anonymous Coward on Wednesday May 18 2022, @01:31AM (2 children)
Add it to the circuit board.
(Score: 0) by Anonymous Coward on Wednesday May 18 2022, @01:34AM (1 child)
Never mind. It seems you've thought about fuse scenarios. I am wondering then why you are so concerned about it if you have a fuse mechanism in place.
(Score: 2, Interesting) by anubi on Wednesday May 18 2022, @02:21AM
I like to design my own stuff because I flat do not like what is offered.
In most cases, the mechanical guys did great. In my case, beautiful stainless steel cooktop and oven.
But the controller, in my mind, sucked big time. Bad design. Unsafe. Hard to maintain. Not reconfigurable to what I may want it to do. Nonstandard parts. No documentation.
Looked to me to be deliberately designed to provide a continuing revenue stream for copyright and patented things that will be used to enforce planned obsolescence.
I've already had to deal with corporate employment, so I at least know why so many corners were cut, so that the permission-giver I was organized to work under could justify a huge salary for his leadership skill. With what passes as business skill in USA, no wonder we have other nations build our stuff. We excel in printing money, buying, selling, renting, and granting permission/ licensing for others to do something.
Oh God, I hated to design crap.
"Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
(Score: 4, Interesting) by RS3 on Wednesday May 18 2022, @05:58AM (6 children)
What if it fails ON? You get a nicely preheated oven, saving you time when you're hungry for that frozen pizza. Win-win, no? :)
Fully agree with you on solid state stuff. I too do much component-level repair and design. I would use triacs for temp control, and contactor for safety. If Arduino (or whatever) sees power coming out of triac when it's supposed to be off, then kill contactor and sound alarm. And use contactor to disconnect everything from mains when oven not in use.
You might want to use "snubbers" to damp switching kick. Some heating elements can be quite inductive. Some are counter-wound, reversing the wind every 10 or so turns to dampen the magnetic field.
Big contactors have arc dampers built in.
Varistors (MOV) might be a good thing too, connected after fuses of course.
(Score: 2, Informative) by anubi on Wednesday May 18 2022, @07:55AM (5 children)
All those are very valid components for a safe design. I will do all sorts of stuff to mitigate power spikes. Another one that gets me is radiation artifacts occasionally flipping bits in computers, but I can normally mitigate that by careful design and having the controller and the device ( usually analog in nature ) supervise each other. That's what analog watchdogs are for. If the processor won't wake up, the analog takes over to insure the system fails safe.
I design both hardware and software in order that I can design both to supervise the other.
I had my early engineering training at the Chevron oil refinery in Mississippi. They indelibly marked me for life on the importance of failsafe design.
I just have this thing about anything that can ignite fires, or cause harm to anyone, be supervised at all times by a qualified human. I never know about flammable things carelessly left on a cooktop.
I even have another repurposed oven sitting outside where I keep my older Diesel van batteries in it as I continue to use them on continuous float charge to power lighting circuits. The oven being just a fireproof box. Just in case one of the batteries decide to explode.
I will probably even put a key lockswitch on the oven just to make damn sure I never multitask and abandon a live oven to go somewhere. I've done some really stupid things, and when I build something like this, I will take the extra care to make damn sure the really stupid thing I did won't happen again. Thankfully, all I did the first time was reduce to ash a perfectly good roast because a neighbor came over and one thing led to another and I flat forgot about the roast. That's how one loses their house!
Which brings me right back to my original comment about high power solid state devices nearly always failing fused ON, being such concern to me.
"Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
(Score: 0) by Anonymous Coward on Wednesday May 18 2022, @08:03AM (4 children)
Oh, one more thing, a failed ON oven controller isn't just preheated, it's fullbore ON! And likely did this completely behind my back!
(Score: 0) by Anonymous Coward on Wednesday May 18 2022, @06:39PM (3 children)
i don't understand all the comments.
however, everything can fail.
one lazy evening i heard a noise that sounded like a crash coupled with a "strain" and when walking to the front gate, indeed, some pickup truck had crashed smack into the middle of the pillar of the property wall. the strain sound was from the rebar doing its thing.
one of his "sport" rims had failed going around a slight bent and he ended up in the wall.
until i got there he had back out and tried to "flee the scene". he got about 50 meters. then got stuck.
anyways, at my location i fear for my electric motors more then anything else.
i got a solid-state relay and a "grid-quality-monitor". the monitor can do 10A directly but the socket it is monitoring can do 16A (reset-able fuse) so i had to get the solid-state relay.
i get mostly brown-outs, that is a loss of one phase on the 100kva wye-transformer doing 22kv to 230v/50Hz duty.
there's still voltage on the lost phase (and a lightbulb gives "brown" light) but reduced and giving less then 170 V to a motor that has a label of 220V it makes up with more amps draw which maybe the wires don't like so much. so now i have that for a fridge and a washing machine ...
[220-230/50Hz]
solid state relay: http://www.tend.com.tw/eng/product1/tssr1-01.htm [tend.com.tw] (total miserable website. obviously bought it in a shop)
grid-quality monitor: https://wipelectric.com/w-op2/ [wipelectric.com]
i guess a shitty low voltage grid makes stuff like this more profitable: https://www.marketwatch.com/investing/stock/kce?countrycode=th&mod=mw_quote_tab [marketwatch.com]
(Score: 1) by anubi on Wednesday May 18 2022, @10:47PM (1 child)
Apparently, you are running three phase stuff.
And flakey power.
Stuff like that is why I got so interested in Arduino.
Those motors, while extremely efficient and reliable, do have an Achilles heel about their power feed.
One phase goes screwy, the motor becomes unbalanced, and burns out. You did the right thing: power quality monitors.
You probably need the solid state controller for both it's preciseness and its almost limitless number of switching operations. However I always try to put in the old style electromechanical contactor ahead of everything else just to guarantee to me that I can kill all power if all hell breaks loose....which is one of the things that will happen if you lose a phase or a phase becomes screwy ( happens when something else on that phase malfunctions and injects fault currents from other phases into it ). Your power quality monitor is for seeing that.
You may also need a heads-up signal to shut other things down before cascade failure occurs.
From the looks of your post, you are aware of what can happen, and the existence of power quality monitors. You know about the wolf. All these problems are unique to your unique situation. Keep thinking about it. From one engineer to another, your situation is very similar to mine, and you are investigating and solving exactly the same way I would.
"Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
(Score: 0) by Anonymous Coward on Thursday May 19 2022, @04:53PM
thx for reply.
my house has a 4-wires coming in, so 3-phase 220V/Hz.
however my fridge is just single phase and the washing machine too. if i understand correctly, even single-phase motors will die prematurely if powered by "sub-standard-voltage"... if one 22kv line (there are 3 lines) feeding the transformer is lost (each one line has its own 22kV breaker), then the 220V coming out won't be 220V but maybe 76V/167V/220V (or was that for two lost phases?)
for "bigger" motors, like say a 0.75kw water pump i have the same device but connected to a mechanical relay (and a timer and a manual switch) that has a buy option to add a conform "thermal protection" where one can set the "running current" with a screw driver. this thermal protector after the relay is slow thus motor starting inrush-current will not trigger it. however, the voltage can be a-okay but for some reason something got stuck or is grinding inside the impeller and the motor starts to draw more current, this is where the slow "thermal protector" comes into play...
it also just occurred to me that these "grid quality monitors" or "voltage monitors" could be used to turn on/off grid connected solar-inverters (solar.dc-2-grid.ac) that are being "duped" by a battery inverter to function during a real grid blackout.
the problem is that with battery inverters that are a "voltage sources" and provide "pure sine-wave" the connected solar inverter thinks that the grid is up and running.
so in the case of a sunny day and nobody is home and the grid died, the solar inverter pushes out current with no where to go but into the battery.
but a battery has a limited storage capacity and the battery inverter has no way to tell this to the solar inverter: "please stop giving energy, i am full".
if the battery inverter is not completely stupid, the outputed voltage should rise gradually when being charged ... i am not sure maybe from standard 220Vac (50% full) to 245Vac (95%full).
with the mentioned "voltage monitor" one can set the "abnormal voltage" threshold to say 244VAC, give signal to relay (solid or mechanical) that disconnects the "stupid" but eager grid-tie solar inverter.
in same manner, everything now runs on battery only, it gets discharged, reaching maybe 222Vac (53% full or whatnot) and the "voltage monitor" now considers the voltage to be "normal" again and reconnects the solar inverter.
alas, now the sun has gone down :D still better then coming home to find burned ruins :P
(Score: 3, Informative) by RS3 on Thursday May 19 2022, @02:30AM
I started to write this earlier, and browser crashed and lost my typing.
Anyway, I work in a smallish food processing factory and most things run on 480 VAC 3-phase. 3-phase motors are awesome, but their one Achilles heel is, like Anubi wrote, the phases and voltages need to be clean and stable. If one or two phases lose some voltage, the motor will draw more current, and may overheat.
Most of the motors which are controlled by a contactor (big relay) have an additional motor protector module. It's a very sensitive overload cutout, and also senses phase direction and balance. If 1 or 2 phases drop below some point, it'll shut off the motor.
That said, many of the motors are driven by a VFD, or Variable Frequency Drive. It rectifies and filters the 480 VAC resulting in ~670 VDC, and then a 6 transistor "bridge" circuit synthesizes a sine wave to run the motor. They typically can vary from 0 Hz to 120 Hz, but can go much higher if you override the limit. Of course you have to be sure the motor and mechanicals can withstand the speeds. The VFD usually "knows" the motor's load, power factor, iir heating, etc.; in other words, it can determine if the motor might be overheating due to the characteristics of the current waveform drawn, "back EMF", etc.
You can run 3-phase VFDs from 1 phase power, so you can run a 3-phase motor even if you only have 1 phase lines power. In USA, most commercial and industrial buildings are powered with 3-phase, but most residences are 1 phase (240 "split": 120-0-120 ).
(Score: 2, Interesting) by Anonymous Coward on Wednesday May 18 2022, @12:34AM (5 children)
1. Advantages of SiC transistor vs Si Transistor.
2. "medium-voltage convertor" - seems like a lingo for power-grid field that even many EEs won't know.
TFS don't make much sense to anyone outside those with background in solid-state physics AND power-grid/energy transmission.
I suggest that the editors post subs on topics they have some clue about so they can actually "edit" with some competence, instead of just pushing whatever is at the top of queue.
(Score: 3, Interesting) by RS3 on Wednesday May 18 2022, @06:03AM (4 children)
I'm okay with the articles and submissions here. I don't know why people complain- it's free, and quite easy to skip over things I'm not so interested in.
But I sort of agree about TFA. I'm an EE, understand some physics, power-grid, etc., and the article is pretty bad. It talks about SiCs and overcoming their obstacles, but doesn't say what the obstacles are!
Only takeaway for me is they're saying more research needs to be done to understand SiCs, their pros and cons, what might be needed to improve them, their implementation, etc.
(Score: 4, Informative) by Rich on Wednesday May 18 2022, @10:09AM (3 children)
This whole modern power switching/converting is a science of its own. Normal people don't really get there unless they have to deal with driving high power brushless DC motors (microstepping also counts). I came across the topic when I looked at what's inside the driver box of my little milling machine.
There are about three high-power technologies that outperform classic MOSFETs. SiC, GaN, and IGBT. Each have their own advantages and disadvantages. E.g. (in general) SiC is considered fast, while IGBT and GaN can deal with higher loads, but IGBTs are slow, and GaN won't do high voltage. By now there are probably developments (see TFA) where one technologies advances into the field of the others. I don't know the details too well, but I imagine they might even be combined.
From a quick glance (not following up), I'd guess the article is about SiC advancing into IGBT territory.
(Score: 2) by RS3 on Wednesday May 18 2022, @03:00PM
Thanks and yes, makes sense. I like power stuff, but I'm not designing in that world and I'm not up on the latest details. One thing in TFA that I hadn't heard of before: integrated capacitor-blocked transistor (ICBT). Very interesting.
TFA and discussion reminded me (slowly) about "wide band-gap semiconductors" [appliedmaterials.com].
(Score: 0) by Anonymous Coward on Thursday May 19 2022, @01:52AM (1 child)
> This whole modern power switching/converting is a science of its own.
This ^^^
I'm getting an education on it from some practicing pros--my present job is part of a team designing a test cell that will include a couple of 1000+ KW motors (~1400 HP). One of the design decisions is whether 690V is a better choice than "medium voltage" which in this case seems to be something like 4KV(?)--pros and cons to either choice.
My expertise is on the ME side but this big power control is fascinating to learn about.
(Score: 2) by RS3 on Thursday May 19 2022, @02:49AM
Wow, I didn't know you could run a 1,000 KW motor on something as low as 690 V. So the amps are like 833 / phase?
If so, that's a tough one. That's a lot of amps to switch. But 4 KV is a lot of volts... I'd be curious what your company ends up with.
(Score: 1, Informative) by Anonymous Coward on Wednesday May 18 2022, @01:37AM
Maybe some aspects of this transistor are new, but SiC itself is not.