Ok Jahan, tnx.
I do not use an output transformer in my E PAs, rather a simple coil (which
becomes the biggest part of the PA on 137kHz when losses should be less :) ),
i mean a LC transformer.
OK, i understand what you mean about the harmonics. But are the harmonics such
a problem? I use a standard 4 pole lowpass filter and harmonics weren't ever
been a real problem. Even without a LP filter the wave looks astonishing sinus
like ;), at least at Z=R=50 Ohm.
Additional complexity will be on the driver side (?) but that is an interesting
concept! I have never thought about that. Are there no problems with coupling?
I mean, the phase has to be exactly 180deg and so you have to choose exactly
the same Cs, Ls and so on. What about the variation of the parameters of the
FETs and Ls and Cs due to temperature rising (if it is not changing equally),
manufacturing and so on? Isn't that a bigger problem than the harmonics?
What are you planning, how much power on wich band? And which FETs do you use?
IRFP series?
73 es gl :)
Stefan/DK7FC
________________________________
Von: [email protected] im Auftrag von Johan H. Bodin
Gesendet: Di 12.01.2010 12:51
An: [email protected]
Betreff: Re: AW: LF: PA0 into Alaska 600m
Hi Stefan,
yes, a single ended class E amplifier would give the same power with
both FETs in parallel but, due to symmetry, the pushpull design has the
advantage of cancelling all even order harmonics by design. 3f is the
first harmonic to worry about so the lowpass filter can be made simpler.
The added complexity of pushpull is minimal  one more capacitor and
one more DC feeding choke. And, yes, an output transformer in case you
didn't already use one for impedance matching in the single ended case.
By the way, the amplitude of the first harmonics of a single ended class
E amplifier (without lowpass filter!) is 0.51/QL (2f) and 0.08/QL (3f)
with respect to the carrier (1f) amplitude. QL is the loaded Q of the
series resonant tank, XL/Rload. Example:
If QL = 5.1, a reasonable value, the amplitude of the 2f harmonic will
be 0.51/5.1 = 0.1 compared to the carrier i.e. 20dBc.
In a perfectly balanced pushpull circuit, the 2f harmonic will
disappear so the first harmonic problem will be 3f at 36dBc which is
"16dB easier" to filter and further from the carrier as well :).
73
Johan SM6LKM

Stefan Schäfer wrote:
> Hi Johan,
>
> What are you planning? This sounds like QROO? Why using a pushpull class E
> stage when a single stange can do it?
> I am interested! :)
>
> Stefan/DK7FC
<<winmail.dat>>
