Well AM does not require EER. We have used EER techniques for decades to
trick AM transmitters to transmit various things other than AM. EER or
envelope elimination and recovery in the old analog days this was done
by by taking an input RF signal containing any type of modulation and
splitting it into two paths - one into a hard limiter and the other path
an envelope detector. The clipped signal contains phase information and
goes through the RF chain which can be class C, D or E. The envelope pah
goes to an amplitude modulator. The difficult part was to perfectly
align the two paths for for identical phase across the signal bandwidth.
SSB is easily possible if the modulator is capable of going down to DC.
SSB implementation of EER goes back to the 1950s.
EER and I/Q are close cousins but not exactly the same. Digital
processing makes this all quite easy. It is quite common to implement a
quasi feedback system to develop a pre-correction scheme. Generally a
sample of the output is taken and compared to an input signal that is
delayed enough to be time aligned with the output sampling delay. A look
up table is developed from the result. None of this is really new or
that difficult.
My personal preference is to avoid class E. I think it is too difficult
to use across a wide band and even when used on a single frequency it
has higher incidental phase modulation which works against EER. There
are many other methods to accomplish the end result. Some have pulse
width modulated the RF carrier but that requires special techniques to
maintain the ability to preserve phase modulation. Any saturated means
of amplification will require output networks of some sort.
It would also be possible to take a fairly standard AB or even Class A
amplifier and drive them with a pre-corrected signal that is digitally
derived in an FPGA or DSP. The advantage would be a wideband
amplification system that with the right pre-correction could eliminate
most RF filtering. This would eliminate band pass filters used in most
SS rigs.
A composite transmitter would get around FCC restrictions on standalone
Linears.
On 5/3/12 6:35 PM, Dan Mills wrote:
> On Thu, 2012-05-03 at 18:20 -0400, W2XJ wrote:
>> A good portion of this has already been done. Check the HPSDR group.
>> Also look at the ADAT. The final hurdle is to do this at the legal limit.
> Yea, I know about the HPSDR project.
>
> AM is really the easy case as there is no phase modulation component, so
> EER is really simple and class E (as long as you don't want easy tuning
> across a wide band!) rules.
> SSB is far more tricky as there is significant phase modulation
> component in play, and unlike AM the envelope goes through zero
> amplitude, so polar loop is really tricky.
>
> A sort of halfway house would be to export the I/Q pair from an
> essentially conventional rig (maybe along with a constant amplitude
> carrier reference), this would allow an almost conventional amp to take
> advantage of whatever efficiency and linearisation opportunities were
> available while still retaining fallback to 'simple minded amplifier'
> functionality.
>
> For me, 'legal limit' is somewhere around 450 - 500W (400W at the
> feedpoint by the rules), makes things a bit easier.
>
> Regards, Dan.
>
> _______________________________________________
> Amps mailing list
> Amps@contesting.com
> http://lists.contesting.com/mailman/listinfo/amps
>
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|