HFU HF Underground

Technical Topics => The RF Workbench => Topic started by: OgreVorbis on October 07, 2019, 2256 UTC

Title: IXYS-RF better than GaN?
Post by: OgreVorbis on October 07, 2019, 2256 UTC
I just came across these mosfets that are designed specifically for class D or E amplifiers. I knew about SiC and GaN being good options, but I didn't know they made fets for this specific purpose. Do you think they would be better than GaN? The RDS on doesn't look as good as the Gan, but what do you think? Maybe it's just older technology being advertised for this purpose?


http://ixapps.ixys.com/DataSheet/IXZR08N120-120A-120B.pdf (http://ixapps.ixys.com/DataSheet/IXZR08N120-120A-120B.pdf)
http://ixapps.ixys.com/DataSheet/IXZR16N60-60A-60B.pdf (http://ixapps.ixys.com/DataSheet/IXZR16N60-60A-60B.pdf)

They're kind of expensive, but I might try them at some point.
Title: Re: IXYS-RF better than GaN?
Post by: redhat on October 07, 2019, 2331 UTC
Both datasheets date to 2009 and use conventional mosfet technology.  All the places I checked list them as obsolete or NLA.  At close to 2000pF, they will be hard to drive, and at only 8A of continuous drain current output power capability will be limited.  Modern SiC devices are much better in almost all of the above metrics, and are available from a multitude of vendors at low cost.

Title: Re: IXYS-RF better than GaN?
Post by: OgreVorbis on October 07, 2019, 2356 UTC
That was my suspicion, thanks!

Right now I use C3M0280090D which has 280 mOhm and 150pf

I am looking at GaN:
41mOhm and 1500pf
https://www.transphormusa.com/document/datasheet-tp65h035ws-650v-gan-fet/ (https://www.transphormusa.com/document/datasheet-tp65h035ws-650v-gan-fet/)

60mOhm and 1000pf
https://www.transphormusa.com/document/datasheet-tp65h050ws-650v-gan-fet/ (https://www.transphormusa.com/document/datasheet-tp65h050ws-650v-gan-fet/)

So they have lower RDS on, but much higher capacitance. So what is more important RDS on or capacitance? Would they be better?

Thanks :)
Title: Re: IXYS-RF better than GaN?
Post by: redhat on October 08, 2019, 0014 UTC
It all depends on your application!  If you want maximum power, you need to consider voltage and current ratings over the range of temperatures you expect the devices to operate at, and also whether you can safely dissipate and get rid of the heat you expect to generate.  The Achilles heal of these devices will be their power dissipation rating.  If you want to operate at 9 MHz or less and at power levels of under 500W carrier, conventional SiC devices are the cheaper way to go.  Above 10 MHz and under 500W, I would expect GaN to have an advantage.

The GaN devices you pointed out are designed for power conversion applications, just like most of the SiC stuff.  The only thing I worry about is what happens when you run the GaN devices at much higher frequencies (10 MHz) and how they behave.  You have two mosfets in the case to worry about now, and I don't know what it takes to make them avalanche or behave unexpectedly, kinda like conventional fets and their body diodes when used in third quadrant (full bridge class D) operation.

http://www.aosmd.com/res/application_notes/mosfets/Power_MOSFET_Basics.pdf (http://www.aosmd.com/res/application_notes/mosfets/Power_MOSFET_Basics.pdf)

Title: Re: IXYS-RF better than GaN?
Post by: Stretchyman on October 29, 2019, 0733 UTC

6V drive.

I have them working at 14MHz.