[600MRG] Fwd: Class E Keying
Brian Pease
bpease2 at myfairpoint.net
Fri Feb 12 10:19:29 CST 2016
Any change in the input duty cycle would require a corresponding change
in the tuning of the "tank" circuit (changing the drain to ground
capacitor) to maintain class-E operation and avoid current spikes and
reduced efficiency.
Modulation or on-off keying should be done by varying drain voltage.
There is a linear relationship between DC drain voltage and RF output
voltage, which allows clean Amplitude Modulation (AM)
I use an Ultimate-3 to drive a homebrew 100 Watt P-P Class-E MOSFET amp
on 630m. It is based on the theory found in
http://www.ka7oei.com/mpm_class_e.html . I simulated the design in LT
Spice. My drain waveform looks correct. For gate drive, I use a MOSFET
driver IC capable of 9 Amps peak current, driven by a 50% square wave.
The gate waveform has fast rise & fall times. My efficiency is >90%
with most of the loss in the tank and low pass inductors. By varying
just the drain supply voltage, I can operate from 1W to 100W.
On 2/12/2016 8:22 AM, Hans Summers wrote:
> Hi Jim
>
> Class-E is inherently non-linear by design. Attempts to modulate the input
>> remove it from Class-E.
>>
>> Although certainly not perfect RC shaping is better than no shaping at
>> all. More sophisticated PWM modulation can no doubt be accomplished to
>> achieve different PA waveforms.
>>
> Yes agreed, RC shaping is better than nothing...
>
> The following papers by NA5N are very informative on Class-E:
>
> http://www.nrao.edu/~pharden/hobby/_ClassDEF1.pdf
> http://www.nrao.edu/~pharden/hobby/_ClassDEF2.pdf
>
> According to that, the best way to control the power output is by varying
> the duty cycle of the input square wave. Messing with the bias on the
> driver is a not unreasonable way to achieve something similar.
>
> Oh Class-E seems such a delicate thing - everything has to be just right -
> it seems to me that pretty much anything causes it to drop out of Class-E!
> I wouldn't be at all surprised if dropping the PA voltage will cause it to
> escape from Class-E operation. On the other hand maybe that doesn't matter
> too much for a few milliseconds during key-down/key-up.
>
> 73 Hans G0UPL
> http://www/qrp-labs.com
> http://www.hanssummers.com
> _______________________________________________
> 600MRG mailing list
> 600MRG at w7ekb.com
> http://w7ekb.com/mailman/listinfo/600mrg_w7ekb.com
>
More information about the 600MRG
mailing list