Why no talk on Gallium Nitride GaN transistors/power electronics??

Logic11

1 kW
Joined
May 2, 2022
Messages
343
Gallium nitride electronics could drastically cut energy usage in data centers, consumer devices.
"...GaN transistors have at least one-tenth the resistance of such silicon-based transistors, according to the company. This allows for much higher energy-efficiency, and orders-of-magnitude faster switching frequency — meaning power-electronics systems with these components can be made much smaller. CEI is using its transistors to enable power electronics that will make data centers less energy-intensive, electric cars cheaper and more powerful, and laptop power adapters one- third the size — or even small enough to fit inside the computer itself..."

Embedded GaN charger with E-bike battery​

"...announced the new GaN charger for e-bikes embedded with the battery...GaN devices provides power electronics designers with new levels of power density, performance, and cost-effectiveness. Gallium nitride (GaN) is a wide-bandgap, next-generation semiconductor technology that has become key for development of advanced power electronics. It operates up to 20x faster than silicon and provides up to 3x the power or 3x the charge in half the size and weight of silicon devices..."

  • Creation of 2DEG: By growing a thin layer of aluminum gallium nitride (AlGaN) on top of a GaN crystal, a strain is created at the interface that induces a compensating two-dimensional electron gas (2DEG). This 2DEG is used to efficiently conduct electrons when an electric field is applied across it.
  • Efficient Conduction: This 2DEG is highly conductive, in part due to the confinement of the electrons to a very small region at the interface. This confinement increases the mobility of electrons from about 1000 cm2/V·s in unstrained GaN to between 1500 and 2000 cm2/V·s in the 2DEG region.
  • Superior Performance: Gallium Nitride’s high conductivity produces transistors and integrated circuits that feature higher breakdown strength, faster switching speed, higher thermal conductivity and lower on-resistance than comparable silicon solutions.

I would have thought GaN would be all over this forum, with people modding current chargers and controllers etc for a start and much talk on new ctlrs and chargers etc..
But this 'Game Changer' it's only mentioned in passing and infrequently.
So what's up with that!?

Is it patented up the wazoo and expensive as hell, or in everything already and I've just now crept out from under my rock!?? :)
 
The problem is GaN like SiC semiconductors has advantages but using them is complex, you can't just swap them in where you had silicon and devices need to be designed around them, a process that is difficult. They are in many ways much more difficult to use than silicon for lots of very complex reasons. I've seen a few attempts at GaN controllers but I don't know that it's really worth the effort most of the time. Both GaN and SiC allow you to do things better but those things aren't always very relevant for us. With GaN you can make things very small, but our controllers are already small enough. You can make them more efficient, I think often by running a much higher switching speed so that would be in a charger but who cares enough about charger efficiency to spend more than twice as much probably on one. And I don't know that the advantage with GaN would be as significatnt in a controller as it would be more efficient but I think to fully take advantage of it you would need to increase the switching freqency which requires changing everything else. SiC can operate at much higher temperatures and voltages, doesn't seem that relevant for our uses. In the end if you need more power or less resistive losses you just add more mosfets.

It's not to say that more GaN devices won't come along but they will mostly be commercial scale production of power supplies where the design effort can be paid off through production scale. We might see some GaN power supplies that are at the voltages we need to charge things, I guess the GaN USB supplies could already be used for some lower voltage things. GaN controllers I think are aways out except for maybe some one off builds by skilled designers.
 
I tried to make a 48V asynchronous motor inverter using bare die GaN (with a ball grid array under), they're incredibly difficult to solder onto the PCB, and are expensive

the footprint isn't practical, and is so tiny
 
I'd think the issue is marginal gains. Saving some percentage of the small amount that is currently being wasted doesn't add up to all that much as a practical matter for most ebike situations. There are certainly some where it would matter but not many.
 
[/QUOTE]
The problem is GaN like SiC semiconductors has advantages but using them is complex, you can't just swap them in where you had silicon and devices need to be designed around them, a process that is difficult. They are in many ways much more difficult to use than silicon for lots of very complex reasons. I've seen a few attempts at GaN controllers but I don't know that it's really worth the effort most of the time. Both GaN and SiC allow you to do things better but those things aren't always very relevant for us. With GaN you can make things very small, but our controllers are already small enough. You can make them more efficient, I think often by running a much higher switching speed so that would be in a charger but who cares enough about charger efficiency to spend more than twice as much probably on one. And I don't know that the advantage with GaN would be as significatnt in a controller as it would be more efficient but I think to fully take advantage of it you would need to increase the switching freqency which requires changing everything else. SiC can operate at much higher temperatures and voltages, doesn't seem that relevant for our uses. In the end if you need more power or less resistive losses you just add more mosfets.

It's not to say that more GaN devices won't come along but they will mostly be commercial scale production of power supplies where the design effort can be paid off through production scale. We might see some GaN power supplies that are at the voltages we need to charge things, I guess the GaN USB supplies could already be used for some lower voltage things. GaN controllers I think are aways out except for maybe some one off builds by skilled designers.

Thx guys.
So in summary;
They are better but difficult to work with and the gains aren't big enough to make putting up with the difficulty of using them to build circuitry worthwhile..?

I did find this while looking to see if they where in any E-bike products yet;


"...Compared with the traditional flyback controller which consumes the transformer leakage inductance as heat, the AHB controller not only recycles the leakage inductance but also realizes zero-voltage switching of the main MOSFET and zero-current switching of the synchronous rectifier, reducing switching losses and heat generation....

...AHB combines the advantages of the flyback converter and the asymmetric half-bridge, utilizing the excitation current to achieve MOSFET ZVS (Zero Voltage Switching) and the resonant current to achieve secondary rectifier diode ZCS (Zero Current Switching)..."

The bottom bits are all the manufacturers that have adopted the above.
 
Last edited:
1 kW GaN reference design for pedal-assisted e-bikes
GaN transistors and integrated circuits allow increased power density in electric motor drive applications. Optical layout enables oscillation-free output waveforms and clean current reconstruction signals. EPC presents a reference design for motor control of pedal-assisted bicycle


 
Last edited:
OK, that's interesting. Eliminating the need for a heatsink and reducing size and weight can certainly have benefits for ebike design. My first comments were mostly with an eye on improved efficiency. That said, when I look at the sizes of Grin's PhaseRunner and BaseRunner controllers, I wonder how much size benefit there really is. I also noted that in the demo the max power supply amperage was around 5A @ 36v or less than 200 watts. I'd like to know what the design looks like at much higher levels.

I would be interested to know what @justin_le and others who understand electronic design far better than I do think about GaN. Though @scianiac may very well have covered the major pertinent design issues.
 
IMO a driving force will be the low resistance cells that are coming out. Example the Molicel 60B that is mentioned in the post below:

 
Last edited:
Back
Top