[HN Gopher] TI expands internal manufacturing for gallium nitrid...
___________________________________________________________________
TI expands internal manufacturing for gallium nitride (GAN)
semiconductors
Author : sandwichsphinx
Score : 46 points
Date : 2024-10-24 21:31 UTC (4 days ago)
(HTM) web link (www.ti.com)
(TXT) w3m dump (www.ti.com)
| alephnerd wrote:
| GaN semiconductors are heavily used in military and space
| applications like Radar, Lasers, and Power ICs.
|
| The US has recently been subsidizing the development of an
| alternative supply chain for GaN and other similar semiconductor
| in Japan (eg. the TI announcement), India [0], Phillipines [1],
| Australia [2], etc
|
| [0] - https://www.state.gov/new-partnership-with-india-to-
| explore-...
|
| [1] - https://www.csis.org/analysis/new-trilateral-chapter-
| united-...
|
| [2] - https://www.ussc.edu.au/expanding-the-national-security-
| inno...
| ninju wrote:
| Its also quite popular with commercial electronics
|
| https://www.belkin.com/products/product-resources/gan-charge...
| saturn8601 wrote:
| I can't wait for the day that we can get GaN replacement
| adapters for all of our old vintage electronics. Its such a
| cool technology. We have SSD replacements for spinning disks,
| HDMI and OLED mods for video and now the final frontier: old
| terrible non switch mode (or even early switch mode) PSU
| replaced with GaN supplies.
| nsxwolf wrote:
| Can't we do that now? There's already a cottage industry of
| replacement power supplies in the vintage computing and
| video game sphere, I would imagine GaN products will
| dominate that space eventually.
| saturn8601 wrote:
| Seems like there is a lack of power supply design
| expertise there, maybe due to certification requirements
| or possibly just the margins on a replacement PSU is much
| less than something like a flash cart or HDMI solution. I
| personally wouldn't dare to design a PSU: I would
| probably burn my house down.
|
| My hope is that certified designs that target various DC
| voltages will cover a lot of use cases.
|
| EDIT: Oh sorry I misread your comment. You are claiming
| there is a large selection already? I don't see it as
| often as I see other 'mods'. Are you speaking about any
| particular vintage computing/gaming community in
| particular? For consoles all I see is junk coming out of
| China. I've been burned too many times by that.
| rodgerd wrote:
| That's part of it - GAN-FETs are becoming very popular in the
| audio world (e.g. the Peachtree GaN power amps).
| gtvwill wrote:
| GaN has been getting HUGE in the audio realm, less so in
| western companies but in Asia in the PA realm its been the
| go-to for a few years now. Admarks AD442(4x4200w @8ohm) and
| AD60(2x6000w @ 8ohm) are a great example.
|
| 1U, Compact and more power than you can poke a stick at and
| fairly reliable. Also their pricing dumps on the offerings
| in the west (Both well under $2000 AUD). Just be warned you
| want MINIMUM 240v 15A or 32A single phase power circuits
| for these and probably run them at 1/4 or 1/8th rated
| output for ultra low THD.
| whalesalad wrote:
| Wild. I recall back in the early 2000's my dad was working at ITT
| Gilfilan (I think they are defunct now -
| https://en.wikipedia.org/wiki/ITT-Gilfillan) and the latest
| hotness was Gallium Arsenide radar technology. Pretty quickly it
| seems we leveled up to Gallium Nitride.
| preisschild wrote:
| Yeah the latest and greatest radars (Raytheon AN/SPY-6) is
| using GaN instead of GaAs (like the older AN/SPY-1 did) too
|
| Really cool stuff
| kjs3 wrote:
| I always remember GaAs as the tech most responsible for killing
| the Cray-3/Cray-4.
| lysace wrote:
| > at its factory in Aizu, Japan.
|
| Which is great since it's not in China, which the world's
| democratic countries needs to work on uncoupling from since it's
| a) a totalitarian regime, b) part of the weird BRICS network.
| Remember that the R stands for Russia.
|
| We should aim to keep high-tech manufacturing in democratic
| countries. Automation is typically what makes this possible in
| new factories.
| muxator wrote:
| With all its present and past limitations, Brazil is a
| democracy. I do not know India enough, but it is not a regime
| either.
| lysace wrote:
| I did not claim the opposite. Although I do question the
| morality of partnering with Russia in 2024.
| varjag wrote:
| It's worth remembering that the origin of BRICS was the
| Goldman Sachs portfolio for emerging markets. For some
| bizarre reason the term became loved by Western far left and
| then took a life of its own.
| saturn8601 wrote:
| Whats your beef with BRICS? There should be competition in the
| world against the US dollar. It keeps the West more honest if
| there was someone nipping at their heels. Yeah its a mess
| backed by some unsavory countries but at least they are trying.
| Think about everything in your life. There is always at least
| 2-3 options. Why don't we really have that at a country level?
| We got the US, the rest of the west is essentially just vassal
| states for them and then we maybe got China but they may
| collapse before they amount to anything. Thats it.
| Etheryte wrote:
| While all of this is a heavy tangent to the actual article
| we're discussing, I'm not sure if Russia, China, Iran et al
| are exactly the kind of competition we should be fostering
| and cheering on.
| frankharv wrote:
| I have to wonder what the world would look like if USA did
| not go along with Chinas entry into the WTO.
|
| Why a communist state was afforded WTO status is puzzling.
| Look at the situation now. Slick Willie did that.
|
| Press release about how bad we fscked up:
| https://ustr.gov/about-us/policy-offices/press-
| office/press-...
| mattgrice wrote:
| Let's be honest here, there was a long period of
| negotiation that spanned the Reagan - Clinton admins as
| well as approval from congress required. Everyone said it
| was a golden opportunity to open up a huge new market for
| US goods. I was doing a concentration in Chinese History
| in the late 90s and was flabbergasted at the very
| mainstream idea that this was going to be a win-win. I
| guess at the time they thought Russia was going to be a
| liberal democracy, too.
|
| I suspect a lot of powerful people knew that it would
| hurt US workers, US industrial strength, were neutral on
| the latter, positive on the former, and very positive
| about profits (during their lifetime).
| James_K wrote:
| Quite frankly, why should they stick to WTO rules? It is
| an organisation that seems entirely beholden to Western
| interests, to the degree that the US has decided to
| simply stop it from doing anything by blocking judge
| appointments because they didn't want to follow their own
| rules. The only conclusion I can draw from this is that
| the WTO is essentially a feckless and weak organisation
| that exists to muscle smaller countries in line with US
| interest. No one powerful care about its rules, so I
| doubt that China's succession made much of any difference
| at all.
| londons_explore wrote:
| I prefer bad competition to no competition.
|
| They still keep the good guys more honest than they'd be
| with no competition.
| notatoad wrote:
| > Whats your beef with BRICS? ... Yeah its a mess backed by
| some unsavory countries
|
| well, that?
| throw0101d wrote:
| > _Whats your beef with BRICS? There should be competition in
| the world against the US dollar._
|
| Well, there's the EUR and JPY, and to a certain extent the
| GBP.
|
| As for BRICS:
|
| > _Pretty straightforward really. You combine Brazil 's
| history of monetary stability, with Russia's respect for
| property rights, India's domestic tranquility, China's
| financial transparency, and South Africa's investment
| opportunities - and hey presto, you've got a new global
| money_
|
| * https://twitter.com/davidfrum/status/1665053372402081792
| alephnerd wrote:
| The impact of BRICS is really overstated, and is only really
| discussed on Reddit and HN (because Reddit).
|
| In action it's been blunted because of China and India's
| rivalry, with each country vetoing expansion of nations that
| are fully lopsided one way or the other.
|
| It's not different from the SCO or APEC in that sense.
|
| If you want to learn further about blunting in IR I highly
| recommend reading Rush Doshi and Oriana Skylar Mastro's
| research.
|
| Alternatively, this classic scene from "Yes Minister" also
| pretty much explains this strategy -
| https://youtu.be/ZVYqB0uTKlE?feature=shared&t=107
| James_K wrote:
| Better take manufacturing out of America then, if you want to
| keep it in democratic countries.
| dvh wrote:
| GaN diodes have higher voltage drop (1.4-1.7) than Schottky (0.3)
| or even normal diodes (0.7). How can they be more efficient?
| connicpu wrote:
| Probably not for the diode properties, but the transistor
| properties. I'm not an expert in this field, but some quick
| googling says GaN transistors (compared to silicon) have faster
| switching speeds and lower on-state resistance. I'm sure
| there's other tradeoffs, but those are great properties for
| high power, high frequency circuits.
| magicalhippo wrote:
| Here's what ST says[1]:
|
| _The very high electron mobility of GaN material allows
| devices with very low on-resistance and exceptionally high
| switching frequencies, which are key advantages in the design
| of next generation power systems, especially those for electric
| vehicles and renewable energy applications._
|
| Regarding the band gap[2]:
|
| _It has a wide bandgap of 3.4 eV and an electron mobility of
| 1,700 cm2 /Vs. Comparatively, Silicon sits at 1.1 eV and 1,400
| cm2/Vs. GaN's inherent properties thus result in a higher
| breakdown voltage and lower on-state resistance, which means
| that the component can more efficiently handle greater loads
| compared to a similar-sized silicon device, which then leads to
| a lower bill of materials._
|
| [1]: https://www.st.com/content/st_com/en/about/innovation---
| tech...
|
| [2]: https://blog.st.com/powergan/
| londons_explore wrote:
| You aim for no current to ever flow through the diode. Any time
| current might flow through the diode, you should either switch
| on the transistor (so the same current flows through the
| channel), or switch some other transistor in your circuit to
| block the current (ie. a series transistor with source and
| drain switched).
| adrian_b wrote:
| They are more efficient at high switching frequencies, because
| they have very low switching losses.
|
| When a diode is not switched, or it is switched at very low
| frequencies, then the losses are given by the voltage drop when
| the diode is on and by the leakage current when the diode is
| off.
|
| These 2 parameters cannot be improved simultaneously, so some
| diodes are better when used at low voltages and high currents
| and others are better at high voltages and low currents.
|
| However when the switching frequency is increased, which is
| desirable in power supplies and in amplifiers because it allows
| the use of smaller inductors and capacitors, reducing the size
| of the equipment, the switching losses increase proportionally
| with the frequency and over some frequency they become higher
| than the static losses.
|
| When such high frequencies are reached, the most efficient
| diodes become those with the smallest switching losses, which
| currently are the gallium nitride diodes. The GaN diodes and
| field-effect transistors have enabled the much smaller chargers
| that exist now for laptops and smartphones, in comparison with
| the traditional bricks, which for many small computers were
| bigger and heavier than the computers.
|
| GaN diodes and transistors are the best for switching power
| supplies and switching amplifiers with working voltages from
| about 100 V up to 500 V or 600 V, which is enough for equipment
| powered by household voltages.
|
| For the higher voltages that may be needed in industrial and
| automotive applications, silicon carbide diodes and transistors
| are the best.
|
| For voltages under 100 V, silicon diodes and transistors remain
| the best.
___________________________________________________________________
(page generated 2024-10-28 23:00 UTC)