The AM Forum

AMfone's Online AM Handbook => Transmitters => Topic started by: KE1GF on September 24, 2005, 12:22:39 PM



Title: The KE1GF Integral (160-80-40) Driver, 2FET SEPP Class-E PA
Post by: KE1GF on September 24, 2005, 12:22:39 PM
Hey Crew, I've been talking for some time now about a new Class-E PA that I've been working on. It's a 200Watt Single Ended Push-Pull PA with an integral 160-80-40 Driver (CARRIER), she'll do 1250+Watts CW continuous. This is a very cool deck, I'm presenting it here configured with shunt caps and an output network for 80m operation in native Switcher CAD III form.

It takes the standard 45Volts carrier (for use with a QIX Class-H or PWM modulator/PS) and 135V for CW operation.

Even if you're not at the least bit interested in building, this is a fun thing to tinker with! It's a nice way to learn how to use (FREE) LTC SPICE 3 with the very intuitive Switcher CAD III GUI. Which can be downloaded no hooks attached from: http://www.linear.com/company/software.jsp It's pretty cool to tinker with the simulator and fiddle around with circuit without touching an iron or investing in parts or drilling and blasting, etc...  :D I'll let you in on a little secret too, there are tube models included with the SwCAD III package.  8)

I've provided a complete package for you for opening with Switcher CAD, all the required component models are included zipped and unzipped... Along with some screenshots if you don't want to download SwCADIII to open the stuff and just want to see the schematics. I strongly encourage you experiement, see how and why these things work  ;)

http://users.wpi.edu/~smithw/KE1GF_2FET_TTL-Drive_SEPP_E-PA_SwCADIII/

(http://users.wpi.edu/~smithw/KE1GF_2FET_TTL-Drive_SEPP_E-PA_SwCADIII/KE1GF_2FET_200W_Class-E_Schematic-SMALL.jpg)

(http://users.wpi.edu/~smithw/KE1GF_2FET_TTL-Drive_SEPP_E-PA_SwCADIII/KE1GF_2FET_200W_Class-E_Simulation-SMALL.jpg)

Enjoy!


Title: Re: The KE1GF Integral (160-80-40) Driver, 2FET SEPP Class-E PA
Post by: WA1GFZ on September 26, 2005, 01:05:37 PM
Bill,
I wonder why you decided to transformer couple the stages rather than dc coupled stages from the totem pole to P_P gates???  fc


Title: Re: The KE1GF Integral (160-80-40) Driver, 2FET SEPP Class-E PA
Post by: KE1GF on September 26, 2005, 04:19:44 PM
Frank, I like to DC couple them like I did in my first design, but I ran into a few problems. DC coupling is the best cause you always have command of the gate, some previous stage gives out, you can design the logic like in my first to latch the sink the gates, which is a good thing. But I ran into problems:

1) When I modulate the final FETS, the gate C swings a bit causing some differences in the amount of current induced into the flyback, which can be dealt with using a very low RDS(ON) FET, like the IRFZ14. The slope of the trapazoid varys a bit, making the final not obey square law power, exactly. But when I use a transformer it's not so bad, so a little cleaner, you get some power stored in the core. Which gets lost as heat, but its there.

2) It's difficult to layout the DC coupled design, all the driver stage's sink FETs have to be grounded directly where the next stage is grounded, also the bypass caps and so on, a real layout nightmare. With transformer coupled you get more space between the stages, because of the size of the transformer which doesn't have to be physically mounted on the heatsink.

3) There isn't complex timing logic involved with tranformer coupled trapazoid drive, seeing that the waveforms have a zero crossing point, shoot-through current is eliminated inherently, which is a big plus. It makes the amp more like an amp and not a state machine.

4) With transformer drive I'm able to run seperate parallel windings for the first stage of driver chips and I don't have to worry about small timing differences. You're not supposed to piggy-back the driver chips specifically because of this reason, I've seen it done in photographs with long lead lenghts between the driven FET and the drivers. The reason why the driver chips don't catch on fire when this is done, to the best of my observation is because of the stray inductance in the lead lenghts.

Also I've found some cheap driver chips that should be able to make this thing work on 14MHz, I've been told that the FQA11N90 will go up there with sinewave drive, we'll see. I'm thinking that the gates will vaporize, dunno...

UPS should be here soon, It might be round file material, it might be worth its weight in gold.. The parts are cheap, not like I'm playing with $100 IXYS DESI parts.

My AN/USM-425 has problems, I'm in the market for a new (to me but used) phospor scope, who's a good dealer that won't sell me a cheap ebay piece of crap that's gonna die in 3 months? Or do you think it's worth fixing?


Title: Re: The KE1GF Integral (160-80-40) Driver, 2FET SEPP Class-E PA
Post by: WA1GFZ on September 27, 2005, 12:39:04 PM
Bill,
I like transformer coupled stages for fault isolation so failures don't cascade.
This is on top of everything you said.


Title: Re: The KE1GF Integral (160-80-40) Driver, 2FET SEPP Class-E PA
Post by: KE1GF on September 27, 2005, 02:25:09 PM
Bill,
I like transformer coupled stages for fault isolation so failures don't cascade.
This is on top of everything you said.


It'll be interesting to see how it plays... wish me luck :)


Title: Re: The KE1GF Integral (160-80-40) Driver, 2FET SEPP Class-E PA
Post by: KE1GF on September 27, 2005, 02:29:45 PM
Bill,
I like transformer coupled stages for fault isolation so failures don't cascade.
This is on top of everything you said.


It'll be interesting to see how it plays... wish me luck :)
I have a DV camera now so, I'll be able to put it on the deck while stress testing it and see what goes up first, should be fun.


Title: Re: The KE1GF Integral (160-80-40) Driver, 2FET SEPP Class-E PA
Post by: WA1GFZ on September 28, 2005, 09:29:53 AM
Bill,
The transformer should work fine if you do it right. Both my rigs are  trandformer coupled.
AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands