|Subject:||Re: [Discuss-gnuradio] Practicality of DC-light RF front-end?|
|Date:||Fri, 17 Dec 2004 20:03:11 -0800|
|User-agent:||Mozilla Thunderbird 1.0RC1 (Windows/20041201)|
David Bengtson wrote:
Marcus D. Leech wrote:I've been thinking about the practicality of a "DC-light" RF receiver front end, for radio astronomy in particular, but also for general use.It seems to me that something like the MAX2116 (or similar direct-conversion satellite receiver) having a very wide tuning bandwidth (over 1Ghz), couldbe used as part of a very wideband receciver front end.
The big problem with this approach is the required receiver dynamic range. 100 dB or more isn't really out of the question for a general receiver front-end dynamic range, so you'll need 16 to 18 bit of effective digitization to make it work. This is do-able for a 10 to 20 MHz wide bandwidth, but not at this point in time.
This much dynamic range would be needed (actually, a bit more) to "digitize the antenna" and recover weak signals amongst strong interferers.
The good news here is that the MAX2116 is a direct-conversion architecture that goes directly from a 950 MHz to 2150 MHz RF to a programmable 4 MHz to 33 MHz bandwidth analog quadrature IF output. So in Marcus' proposal, the ADC would only need at most 80 MSPS, and there are many that do this with 14 bits of resolution and 75 db SNR (such as the LTC1750). The LTC1746 in Marcus' block diagram is a little under powered (25 MSPS) for the MAX2116 at full bandwidth; you'd need to limit it to about 10 MHz "baseband" out of the MAX2116 in this scheme.
Anyway, it's worth pursuing more; I'd be happy (for my application) with a DC-1GHz receiver by using a single up-conversion to the MAX2116 input range and a 4MHz quadrature analog output. You could do a lot with this (like feed it to a USRP). At this point, the performance limiting aspects are the pre-selector and mixer linearity, not the digital side of things.
|[Prev in Thread]||Current Thread||[Next in Thread]|