NCP81074A driver use

[N4LTA]

I am planning to start doing some work with some GaN FETs and plan to use NCP81074A drivers. I have one question. The spec sheet shows R1 and R2 at the source and sink output connecting to the FET gates. I assume these are low value resistors that may be used to control rise and fall time? Any explanation and suggestion for a typical value would be appreciated before I start making smoke.

Thanks

Pat

[VE3ELQ]

Pat,
The 2 outputs on pins 4 and 5 provide versatility to individually control turn-on and turn-off time of the FET so your assumptions are correct.
For most efficient class D operation of a GaN in a transmitter the FETs needs to switch as fast as possible so just parallel pins 4 and 5 then connect them directly to the gate leads as close as practicable.
BTW the LM5134 drivers work just as well, at least in my projects anyway.  Leave pin 2 open.  Happy building.

http://www.ti.com/lit/ds/symlink/lm5134.pdf

73s  Nigel

[N4LTA]

Thanks Nigel

I had hoped you would chime in.  I have some GaN and SiC fets to experiment with. I have a copy of your experimental 20 meter schematic and am trying to do something with two FETs similar on 75 meters to start. Thanks for the LM5134 tip.

Pat

[M0VRF]

As stated, just tie the legs together and drive directly.

On 40m the Cree C3M0280090D make the drivers quite hot but have never noticed any issues. Any higher in frequency you'll have to use the GaNs which are simply amazing and easily driven with 6V (I use 8V). The SiCs however will need 12V or more but at 12V the drivers run quite hot as stated so I've settled for 12V at the expense of a few %.

Personally I'd stick to GaN for any frequency above 5MHz, keeps everything Cool...

Under 2MHz the C2M0160120D are fine and easily driven, not tried them on 80m but the C3Ms will be fine..

Worth mentioning that on full spectral analysis the square drive gives rise to parasitics around the 300MHz region. Slowing the gate turn on with ferrite or low value Res goes some way to cure this but not fully. Still investigating that one..

JB.

**Edit** was using and elliptical LPF with a top coupling cap to notch out the 2nd harmonic, added another pole without a top cap and the parasitics ahave disaapered from ther O/P, Marvellous!

[N4LTA]

Thanks

I may use the Cree C3 SiC FETs for my 80 meter RF deck. I have some of those.

Pat

[G4ZOW]

What GAN fet's and drivers are suggested for use on 40m or higher bands for 350W+ carrier so plenty of headroom?

David 

[M0VRF]

TPH3206PSB.

Youll need lots of volts and I've yet to find a cheap 150V PSU.

I guess the simplest method is to get a 110V isolating transformer and DC the O/P, bit of a lump mind.

I was ONLY looking for 100W carrier and using any 48V 10A 'el cheapo' SMPSU, that's not a problem.

@ <2A5 per device @ 150V you're close to 1500W peak for four devices so that 's the way to go as Nigel and other have.

My 'product' had to be under 2KG's and be small(ish) using a standard diecast enclosure somewhat limiting my options.

Still get 95% eff at the relatively low voltage and 1MHz B/W.

14MHz is quite possible too, haven't tried any higher.

J.