SDR Discussion

This week I thought I might talk about software defined radios and compare them with the conventional transceiver we all know and love.

Radios in the past 50 years or more have been designed with multiple frequency translation stages to provide good selectivity and good image rejection across all of HF, and for the last 15 years or so on VHF and UHF with the same receiver. The more coin you have to throw at a radio, the better image refection you get, quieter frequency translation resulting in a more sensitive receiver, sharper filtering, then we get down to smarter features like voice keyers, pan adapters, multiple receivers and the like. The biggest cost of these radios is in the filtering and things like phase noise in oscillators and mixers.

Prior to synthesised radios, we had radios like the FT-101 and the TS-520 which were the staple diet for a new ham in the 70s and early 80s. These had much simpler designs and in some ways probably performed or outperformed the newer radios, most likely because of their simplicity (less moving parts to create noise in the receiver). Prior to those radios, many hams built their own radios which in the 50s and earlier, was the norm, rather than the exception.

Then came the software defined radio. Initially these where simple devices, often using a sound card on a PC to carry out the grunt work. These are generally direct conversion radios, i.e. no frequency translation. Straight from RF to decoded audio! Not having multiple stages of frequency translation means no mixing noise, no phase noise in oscillators etc. So how do we do this? We sample a chunk of spectrum, break it up into as small chunks as we can, then we process that data and turn it into audio. 10-15 years or so ago, the hardware to manage this was so expensive you may as well keep to the high end radios like the IC-7700s and the like because the gear used for high performance software defined transceivers were a lot more expensive. This technology was almost exclusively the realm of $40K plus radio test equipment and Military radios.

As computers became more and more imbedded in our society and became more and more powerful, the hardware required for a more functional software defined radio started to come down to something that allowed its use in modern ham rigs. We started seeing software defined receivers coming on to the market for around a few hundred or less that worked amazingly well! The PC became an integral part of the radio, and with computers in general becoming more and more imbedded in the shack and connected to radios so you can add frequency and mode to your log book automatically. It seemed quite natural now for the PC to “be the radio”.

Flex Radio came out with the first commercially available software defined transceivers. The PC still did a large chunk of the hard yacka. They had a lot of potential but you almost needed a high end gaming PC to run one of these radios, and probably cost more than the radio itself! Flex were bringing out new versions and becoming more and more relevant. Others like Apache Labs and Elecraft entered the same market and spiced it up a bit. Apache Labs made the Anan series of radios and took a large chunk of this market from Flex, as did Elecraft. This spurred Flex into bigger and bigger projects, before bringing out the 6000 series and now the 8000 series radio which had the majority of the grunt required to run the radio, imbedded in the radio, meaning you could run the radio with a $200 second hand PC now.

Then Icom and Yaesu entered the Software Defined game as well. They came in with a slightly different approach. They put the same concepts behind what looks like a normal frequency synthesised radio, but with amazingly smart user functions on an intuitive touch screen user interface. These new radios like the IC-7300 and the FTDX-10 were now competing performance wise with the big boys, but with a very cheap price tag, being around half or even less that of the entry level Flex Radio. When I was in High School and a new ham in the late 70s and early 80s, The FT-101 was the most popular radio around. It was the model T Ford of radio. Functional and everyone could afford one. Now the IC-7300 seems to have filled that spot. 

And to add to the fun of this newer technology, you can hear mutterings at club nights etc “you wont catch me buying one of them computer radios”. I only want real radios in my shack! They say this as they pick up a mobile phone which has been using SDR technology for a decade or more, and testing their radio on a second hand $3000 HP-8920 radio test that has used the same SDR technology for 30 years now.

It’s a fact now that the cost of building a conventional transceiver to the same quality as a modern SDR is so much now that its just not commercially viable any more. I don’t know of any radio manufacturer, amateur or commercial, who are still making radios in the old fashioned way. Love it or hate it, the Software Defined radio is here to stay and chances are, whether you like it or not, your next brand new radio purchase is likely to be an SDR!

Its not all doom and gloom though. The possibilities are boundless with what can be done when the major design task of a radio is in software. Integrating logbooks, DX clusters, the audio chain (TX and RX), smarter more functional filters etc make for a lot of fun on its own. It means that instead of replacing your radio every 5 years or so, you keep the hardware platform and just keep upgrading the software! With the price of radios coming down, more people can get into the hobby, which will make it more fun for the rest of us. Personally I cant see a down side anywhere.