[hpsdr] Penelope's progress

[Phil Harman]

I now have the Alpha prototype of Penelope running and producing RF out.
Lyle, KK7P,  did a great job of laying out the PCB and kitting for the 
prototypes -
many thanks Lyle.

My experience with the prototype, and the experiments on the  preselector
(Alex) that I have been doing with Graham, KE9H,  have resulted in a change 
in
architecture for Penelope (and Mercury).

The reason for this is as follows.  We had intended to allow the use of  the
LPFs from Penelope as front end filters for Mercury and/or QSD etc based
receivers.

In testing various filters for Alex what I found was that for receivers with
a potential IP3 of  +50dBm one needs to be very careful as to  the choice of
inductors used at the front-end if one is not to significantly degrade the
IP3 performance.

I now fully appreciate why the German Hilberling PT8000A  receiver uses a
massive 2" diameter  T200-2 toroidal inductor in its preselector.

See the photo of the preselector here

< http://www.hilberling.de/produkte/produkte_pt8_t9.htm >

 Even with such a large core I can detect a slight degradation in  IP3 
performance when using a
T200-2 inductor as apposed to an equivalent air cored inductor.

The bottom line is that whilst we could  make LPFs for Penelope that will
provide adequate harmonic attenuation, given the size constraints of the
Penelope PCB, the IP3 performance of a high performance receiver will be
noticeably degraded if we use these filters on receive also.

We have found the same IP3 issues with the HPF's and LPF on the Mercury
front-end - the small SMD inductors are simply not good enough for a 
receiver
with such a high IP3 performance.

Rather than compromise these designs we have decided to remove all output
filtering from Penelope and the input of Mercury and instead put the filters
on a separate Atlas bus board.  By dedicating a full board to this task we 
can
fit inductors of the appropriate size.  We can also provide for the usual
antenna switching  and connectors e.g. A/B, separate receive only antenna, 
separate
antennas for dual receivers  etc.

Even without any output filters at  0.5w out the spectrum
of Penelope meets the FCC specifications, mostly due to the fact there is no 
hardware
mixing in the signal path.

Carrier and  unwanted sideband suppression also appear to be excellent -
the advantage of doing the signal processing digitally and as close to the
antenna as possible.

In addition, we have used only about 30% of the EP2C8 FPGA. There is enough
room left to process the microphone input, via the ADC  that is provided on
the board, and create I & Q signals.

With the addition of an LCD or other display, tuning knob and various push
buttons, connected to the Atlas DIN connector we have a stand-alone exciter
that could  be used independently of an Atlas bus.

We are about to revise the PCB to incorporate the things we have learnt from
building the Alpha boards. I will keep you informed as to progress.

I will update the Wiki shortly to reflect the changes we have made.

73's Phil...VK6APH