Re: LMG1025

On Wed, 09 Jul 2025 07:32:08 -0700, john larkin <jl@glen--canyon.com>
wrote:

On Wed, 9 Jul 2025 10:58:13 -0000 (UTC), piglet
<erichpwagner@hotmail.com> wrote:
>

john larkin <jl@glen--canyon.com> wrote:

On Tue, 8 Jul 2025 18:36:01 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:
 

On 2025-07-07 20:11, john larkin wrote:

It's a TI GaNfet lowside gate driver. Insane specs.
 
It's a nasty little leadless package. There's an even faster BGA
version.
 
I wish there was a higher-voltage version, for driving SiC parts.
 

 
Yikes, nine bucks in onesies.

 
That's in the noise floor for the products that we have in mind. But
it will cost a lot more than the fets we will be driving.
 
What's interesting is that the schematic on the data sheet seems to
show a p-channel GaN fet... on chip!
 

 
Pretty cool driving  5V into 220 pF in 650 ps, for sure.
 
Cheers
 
Phil Hobbs

 

>
I took that to be a regular silicon p-ch fet. Although the chip is intended
to drive GaN devices isn’t it itself regular silicon- does the datasheet
say it is fabricated from GaN?

>
But the performance is amazing, if it is silicon.
>
My alternative GaN gate driver would be a *lot* of Tiny Logic gate
sections in parallel. Som of those are about as fast.
>
We are also considering using a BUF602, 1 GHz bandwidth and 350 mA
out, but it would need weird positive and negative supplies.
>
We'll try all the above.
>
I really want a fast SiC gate driver too. That would need close to 20
volts swing. Most SiC drivers are isolated (which we don't need) and
have huge prop delays, and probably correspondingly high jitter.
>
My summer interns are playing with fast high voltage switchers on the
theory that a product might emerge somehow.

I does sound like people are putting several semi types on one chip.

https://compoundsemiconductor.net/home