SoylentNews is people
SoylentNews
Surpassing Silicon: Paper-Thin Gallium Oxide Transistor Handles More Than 8,000 Volts:
People love their electric cars. But not so much the bulky batteries and related power systems that take up precious cargo space. Help could be on the way from a gallium oxide-based transistor under development at the University at Buffalo.
In a study published in the June edition of IEEE Electron Device Letters, electrical engineers describe how the tiny electronic switch can handle more than 8,000 volts, an impressive feat considering it's about as thin as a sheet of paper.
The transistor could lead to smaller and more efficient electronic systems that control and convert electric power — a field of study known as power electronics — in electric cars, locomotives and airplanes. In turn, this could help improve how far these vehicles can travel.
[...] Tests conducted [...] in March show the transistor can handle 8,032 volts before breaking down, which is more than similarly designed transistors made of silicon carbide or gallium nitride that are under development.
"The higher the breakdown voltage, the more power a device can handle," says Singisetti. "The passivation layer is a simple, efficient and cost-effective way to boost the performance of gallium oxide transistors."
Journal Reference:
Shivam Sharma, Ke Zeng, Sudipto Saha, Uttam Singisetti. Field-Plated Lateral Ga2O3 MOSFETs With Polymer Passivation and 8.03 kV Breakdown Voltage, (DOI: 10.1109/LED.2020.2991146)
(Score: 2, Insightful) by fustakrakich on Thursday June 04 2020, @06:25PM (3 children)
"The higher the breakdown voltage, the more power a device can handle,"
Yeah, they're half right. How much current can it handle?
La politica e i criminali sono la stessa cosa..
(Score: 2) by captain normal on Thursday June 04 2020, @06:39PM
My first impression also, potential is easy, current is the hard part.
"It is easier to fool someone than it is to convince them that they have been fooled" Mark Twain
(Score: 3, Interesting) by VLM on Thursday June 04 2020, @07:49PM (1 child)
I was looking at a datasheet for a MJE18008 recently (long irrelevant story, but its a half way decent switching power supply transistor) and "a couple amps" is no biggie for modern transistors. And a couple amps at a couple KV is a lot of watts.
The real killer problem as I see it is the old power = volts squared divided by resistance formula.
So you swap out a kilovolt class MJE18008 and solder in this vaporware lab curiosity ten-kilovolt class transistor and take your MJE18008 design and 10x the input voltage because "what could possibly go wrong here?"
So with volts squared, ten times the input voltage means superficially you gotta dump a hundred times the heat. That's gonna HURT in a TO-220 package. I think that's what fustakrakich is getting at. Sure, you can switch 8x more voltage than COTS products, but how you gonna cool that little bastard when it wants to glow like the surface of the sun?
The real "fun" with kilovolt class transistors to me would be RF amplifiers. Of course the contemporary COTS MJE18008 has a gain bandwidth product of like ten mhz, so you're not gonna make a 5 KW UHF TV transmitter outta that dude (or a 10x rated voltage) any time soon.
Or to paraphrase that quote about the JAWS movie and "gonna need a bigger boat" we "gonna need a bigger heatsink".
Still, tech advances and you're gonna see stuff like this in the more exotic tier of VFD and so forth sooner or later.
(Score: 2) by Muad'Dave on Friday June 05 2020, @01:49PM
It's the same with existing transistors. The power dissipation in saturation is driven by drain current and the drain-source resistance Rds(on). The article doesn't mention the Rds(on) parameter, so we have no way of knowing the amount of power that will be dissipated as heat given a particular Vds and Ids combo. There are existing MOSFETs with very low Rds(on) values.
this beast [unitedsic.com] is rated for 1200V Vds and Id of 120A. The Rds(on) is an amazing 8.6 milliohms while passing 100A. P = I * I / R, so that dude is only dissipating 86 Watts! If the Vds is 1200V in cutoff, that thing is switching something like 120 kW.
(Score: 1, Informative) by Anonymous Coward on Thursday June 04 2020, @06:36PM
GaAs is the future of semiconductor, and always will be.
(Score: 2) by VLM on Thursday June 04 2020, @07:31PM (1 child)
Over my gen-x lifespan vfd tech has improved from exotic aerospace stuff to commodity for LV work.
I'm pretty sure I'll live long enough to see powerco transformers replaced by MV rated VFDs. Interesting time to be alive....
Of course on the other side, if you thought copper/iron xformers were sensitive to EMP / lightning, imagine early generation VFD replacements getting wasted and possibly literally vaporizing on the pole, LOL. Or imagine the results of a short out looking a lot like a Star Trek bridge explosion in every house served by the VFD.
Still amazing to imagine maybe in 2030 replacing a giant barrel distribution xfrmer with a little shoebox VFD outputing 240V to my house, LOL.
(Score: 1) by anubi on Thursday June 04 2020, @10:49PM
Yeh, I wonder how resilient it is to inductive kickback. I have seen all too many really big devices done in by flyback pulses from stray inductance.
Some power FETs will just avalanche and recover if possible. While other devices suffer irreversible and permanent degradation.
Stray inductance is a bitch.
"Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
(Score: 2) by Unixnut on Friday June 05 2020, @08:27AM
The drive electronics in current BEVs are not that big compared to the rest of the car. I mean, from memory the largest part is the battery, followed by the motors themselves. The Electronics are usually sandwiched in a little area where they can be adequately cooled.
So even if this Transistor reduced the size of the drive electronics, it would not make much difference, as say, using a more energy dense form of energy storage compared to current battery tech, or developing lighter/smaller electric motors of the same power rating.