Foreword by Colin, G3SBI
I asked Martein if I could write a short foreword about his investigation of the suitability of the Analog Devices AD9910 DDS chip as the local oscillator in what has become known as the holy grail receiver or transceiver.
You will know from this web-site that Martein is a very good research engineer. Not only does this need technical knowledge, it requires determination, logical thought and an attention to detail. These seem to be qualities that are in short supply in RF engineering today.
Using homemade test equipment and his trusty W&G SNA62 spectrum analyzer he has identified a design problem in the AD9910 which will probably also affect the AD9912 and other recent DDS chips by ADI. Having an amateur radio friend in ADI he told me the right people to talk to in the high speed DDS group in North Carolina. There was a great deal of reluctance to accept Martein's findings. Eventually they were able to duplicate his results and they were very grateful to Martein because the same mistake will not be made in their new designs.
The remarkable thing is that it has taken an amateur radio state of the art receiver project to identify this and it has not been found by Analog Devices big customers with all their high tech test equipment.
The bad news is that it is unlikely there will be a chip revision, even though the fix is simple unless there is pressure from a major customer. The good news is that it may bring forward the development of a family of parts, the first one being a 5GHz 10 bit DAC device but there are talks of 16 bit DAC devices at 2.5GHz. Its also likely that Martein will get some early versions of these chips to test.
Analog Devices is a company I have a lot of respect for. The high speed DDS chips are a major technical achievement and are useful to our hobby as Martein will show. During chip development the design team probably paid too much attention to results from their high tech measuring equipment when by design it rejected a device parameter they should also have been looking for.
About the DDS Local Oscillator Update and how it was made possible
A lot of time and effort has gone into the evaluation of the AD9910 DDS. Especially because of the complications encountered in reproducing the otherwise excellent phase noise performance possible with this chip.
I'd like to thank Colin, G3SBI for the many constructive discussions we have had during this project and also for his support together with Andrew Moss from the Daresbury Laboratory in Warrington UK in arranging the ultra-low-noise Wenzel oscillators that would have been out of my reach otherwise. My thanks also go to Mark Sumner from AOR-UK, who has been helpful as always in organizing parts from Mini-Circuits. Last but not least my thanks go to David Brandon, Paul Kern and Chris Muriel from Analog Devices for supplying me with sample chips through priority mail!
Impact on Dynamic Range
Sideband Noise Measurement
Back to Receiver Overview
Back to the TOC