Here is my latest project. I call it The Frankenstein because of the two BNC connectors that come off the side of the DDS oscillator box -- they look to me like the bolts on Frankenstein's neck. The square waves from the DDS LO also seemed to evoke Frank's bolts. There may be other similarities. We'll see.
Here is the idea: Phasing, Direct Conversion, Image Rejecting receiver based largely on the R2 design by Rick Campbell KK7B as presented in the January 1993 QST.
I'm using an AD9850 with an M0XPD Kanga board and an Arduino to generate the quadrature LO signals (you can see the square waves on the 'scope in the background). I'm using the software of Richard AD7C; this, combined with the divide-by-4 scheme on the Kanga board, puts the upper limit of reception at 7.3 MHz. That's OK for now.
When I first fired up my AD9850 box I was dismayed to find that the square wave quadrature output was no longer there. I was about to give up and get anther shield board, but this kind of surrender bothered me. So I started troubleshooting and isolated the problem to the /4 chips. My soldering of the surface mount chips was, well, a bit dodgy, so I changed to a tiny soldering tip and reheated all those tiny little pads. Hooray! I fixed it.
The receiver will be built mostly on a PC board that Pete made for me back when he was trying to convince me to build a fourth BITX receiver. I am pleased to put the board to use. See below.
Yesterday I soldered on the two SBL-1 mixers that will form the heart of this receiver. I realized that the very robust quadrature square waves from the Kanga board might be robust enough to fry the sensitive little SBL-1s. Sure enough, I measured about 17 dbm coming out of the Kanga board. I threw together two roughly 10 db resistive pads. These should prevent the SBL-1s from releasing their smoke.
I hope this receiver will be four receivers in one:
1) Standard DC receiver.
2) Binaural Receiver! Groovy, stereo CW that floats around in your head, man!
3) I-Q receiver that can be fed into the sound card of the computer for DSP, panoramic display, etc. I promise not use it to find fault with the signals of homebrew SSB rigs.
4) SSB image rejecting receiver for easy, Direct Conversion SSB listening without the burden of having to listen to the other side of zero beat.
There is already a lot of soul in this new machine: Kanga board with the design my Paul M0XPD, PC board made on Pete's $250,000 CNC machine, and all of it on an actual breadboard (from Italy, I think).
Rick Campbell and Peter Parker have commented on the allure of phasing rigs. There is something very attractive about them. There is a cleverness in the way this design exploits the phase relationships between sidebands to allow us to null out the unwanted side of zero beat. It took me a while to really understand how this is done -- once I understood it, I really wanted to build a rig that would make use of this principle.