Update: Experimental 4 Silicon-Carbide MOSFET 1kW Transmitter on the air

[John K5PRO]

Class D versus Class E on ham bands:
Is frequency agility comparable between them? I though D could tune around with very little change in performance (at least over 0.5 MHz) whereas the shunt C in E had some frequency dependency. I might have to read the papers again. One thing for sure, when the heat sink isn't hot anymore, but the RF ammeter is upscale, thats real good engineering.

[steve_qix]

In my experience, frequency agility (at least over an entire ham band) is not a problem with either D or E as long as the load impedance is reasonable.  The shunt C in a class E amplifier is a relatively non critical value as long as all of the shunt capacitors are the same in a multi-module setup.  In my own setup, I go from one end of the band to the other.  The only problem I have occasionally run into is a lack of tuning capacitance at the low end because I needed to use a larger capacitor to start with and didn't have the right one in my junkbox

[ka1tdq]

I have an FQA11N90 in my 60 watt, single FET CW rig.  It's class C.  With the higher transconductance figure and better capacitance numbers, would I expect any significant improvement by swapping the FET with one of these?  I'll have one in the mail tomorrow just in case.

Jon

[WD5JKO]

With all the Talk about SIC FET's, Class E versus D, etc...Why not discuss 40 meters, or even 20m?

There seems to be movement by Bruce K7DYY to market a 80-40 Meter version of the Super Senior. His web site has not changed (other than a new audio processor) , but look what Janis AB2RA wrote about it...starts about half way down.
   
http://wireless-girl.com/Projects/AMTransmitters/K7DYYtransmitter.html

Jim
Wd5JKO

[steve_qix]

Not sure about 20 meters - the gate self-resonance frequency is probably near to the band (maybe a little higher), but 40 will work with the new MOSFETs.  Someone just needs to take the time to build it     Unfortunately, I am not yet retired, else the 40 meter rig would be on the air already.

I have a 16 MOSFET 40 meter transmitter I use up there whenever I get on that band.  It's 10 years old and due to be updated with a new design....using Silicon-Carbide MOSFETs, of course!

[VE3ELQ]

With all the Talk about SIC FET's, Class E versus D, etc...Why not discuss 40 meters, or even 20m?

Jim
Wd5JKO

I dont want to hijack Steves thread but a few weeks back I did a 7.4 mhz test with two of the 10a FETs.  I re-configured a bridge deck for push pull that had IXDD604 drivers. The driver current was quite high and the little non heat sinked dip8 chips got really hot but it did make some RF. It was not suitable for 40 M as is. Ive been consumed chasing an annoying HUM in my 80 meter rig for the past month that defies logic so never got back to it.  The 10A SiC FETs switch about 10 times faster than the 11N90 FETs which do work on 40 so with the right drivers they should do well.

73s  Nigel

[ka1tdq]

I replaced the FET and tweaked power up to 80 watts.  I was able to go to 100 watts, but it seemed to be generating too much heat.  

Jon

**And just got a nice 599 out of Springfield, MO from NN0D.  It works.

[steve_qix]

I replaced the FET and tweaked power up to 80 watts.  I was able to go to 100 watts, but it seemed to be generating too much heat.  

Jon

**And just got a nice 599 out of Springfield, MO from NN0D.  It works.

Hmmmm... don't know enough about the operational parameters to be able to respond properly.  Do you have a scope pic of the gate waveform?  That will tell QUITE a lot!

Regards,   Steve

[ka1tdq]

I don't want to hijack the thread for this, but just a few comments...

Actually, I haven't looked at the gate waveform for this.  I need to borrow a scope to do so and I will in the not-too-distant future.  I was using a single FB43-1020 core and I did note that it was getting pretty hot.  I replaced it with two in series and it runs cooler. The FET gets warm, but not hot.  The duty cycle of the DDS VFO is set for 50%, but I haven't verified it by looking at the waveform.   

Also, Vdrain is regulated 48 volts.  I'm not sure of the current draw, but 80 watts output.

When I was soldering this new FET in, I was very careful.  At $34 each, they're expensive 1" squares. 

I'll post pictures of the gate waveform on another thread sometime.  It'll be a few weeks before I get my hands on the scope.

Jon

[steve_qix]

This is an update to the experiments using the C2M0040120D CREE SiC MOSFETs.

Things are working very well on 80 meters.  These MOSFETs appear to be practical for this band for daily use transmitters.  A few thoughts:

- Gate drive is VERY important.  I am driving the gates with 20V of RF on 85 meters.  This is the manufacturer's recommend value.  It may be that a slightly higher drive voltage than 20V is possible without risk of device damage, because the gate resistance attenuates the drive to a degree.  With 20V of drive, the heatsink BARELY BREAKS TEMPERATURE !  The RF amplifier is incredibly efficient - quite a bit more efficient than standard MOSFETs.

- Each of these SiC MOSFETs replaces 4 FQA11N90s.

- Price/Performance is similar to FQA11N90s when you consider the additional drivers, bypass capacitors for drivers - and other driver components, additional machining, etc.

- Construction complexity is significantly reduced

Definitely worth your time in some experiments!!

Regards,  Steve