Hi Gian, This is what I had in mind! You could go one step further than just the 20 kHz filter and also have narrower IF filters for once you had 'zoomed in' on the signal in question, much like the Orion has a bank of selectible 'roofing' filters. No matter where the signal of interest was on the bandscope before zooming in, the computer-radio could shift the LO so that it is centered in the narrower filter once one zoomed in. When you unzoomed the computer could remember the original LO and switch to it and show the original bandscope again so that as the user the whole thing would appear 'seamless'.
I think this is the way to build a very high performance 'SDR'; by starting with a very strong analog front end, and a set of appropriately chosen IF 'roofing' filter widths, and then using the power of the computer to make it all work seamlessly, so the 'user' gets the best of both worlds. You are correct that I of course assumed the I7SWX H-mode mixer would be the one used when I did my gedanken-experiment, as I wanted to use the best design possible ;) I hope that one of the currently very active SDR programs makes use of some of these concepts. It seems to me that without the switchable filters noted above, the hardware, while very interesting and fun to use casually, won't make a viable contest radio [at least for me] unless the strong signal handling capability is 50 dB or so better than current hardware. Thanks again for the very cogent post, Gian. Hopefully others will read and appreciate ;) 73, W3SZ Roger Rehr http://www.nitehawk.com/w3sz Quoting Giancarlo <[EMAIL PROTECTED]>: > Hi Roger, > > Vy nice to hear from you ... I believe you are throwing not a stone > in the pond but a lighted match in the petrol.... hi > > I can share one of my ideas ... SDR still has a lot to pedal before > it can respond to all requirements of a contester and VHF-SHFers > looking for bouncing signals where IMD is important. > > My few million $$$$ idea is to have one conversion to a "not too > expensive commercial filters" IF, as an example: 10.7 MHz, usually > with 20kHz BW. From 10.7 MHz or the like you will need a strong QSD > or possibly an I-Q Mixer (you could select Pericom PI5C3125 and > 74AC00 decoder (see Marco, IK1ODO QSD trial) or a SR40 V7 like mixer) > converting to I-Q baseband. > > You will need some intelligent front end control. The system will > look at the widest possible bandwidth, when zooming into the > interested signal, the IF filter will be inserted and the DDS or PLL > oscillator will be tuned to the signal. Your BW now will be 20kHz in > place of 100 or 250 or expected 400 kHz. A good audio card like Delta > or better and your SDR software will give you all the power of > filtering. > > You could have a similar set-up for HF up to 50MHz with the best > front end mixer guess what ??? A 2 or 3 transformers H-Mode mixer > followed by a good diplexer and post mixer amp driving a good quality > xtal filter (8 to 10 MHz) followed by a QSD or I-Q Mixer and what > necessary . In HF you may or may not need to have a very large BW to > monitor, maybe on 28 and 50 MHz yes. > > I have tested a similar set up for HF and it works very nicely. I did > not make any real measurement but I can send you a Winrad screen > print tuned on 7MHz > > These are ideas good or bad but you can still send $$$$ hi. > > Maybe I am the one throwing the lighted match in the petrol > thank .boooom !!! > > BTW, I have replaced the second mixer with an I7SWX 2T H-Mode Mixer, > and re arranged 2nd IF gain in an FT1000MP ... during a "strong" SSB > contest with one splattering station at -3.5kHz and a clean one at > +2.5kHz they never missed any signal (low) they could hear... no > bubbling. > > 73 > > Gian > I7SWX > > > --- In [email protected], Roger Rehr <[EMAIL PROTECTED]> wrote: >> >> Hello, All, >> >> Bob and Alberto's comments on the 'quality' of sound in a DC-SDR >> provide me to ask a generic question which I have been pondering > for a >> while now. This forum has a lot of talent so I am asking it here > first. >> >> What was said about the exceptional 'clarity' of signals derived > from >> direct conversion SDRs has been my experience as well, whether the >> radio was a simple homebrew design, a SoftRock, an SDR14, or > even 'The >> Time Machine'. It has also been true whether I was using Linrad, >> PowerSDR, Winrad, Rocky, KGKSDR, Spectraview, etc. as the software >> backend. The 'old-style-radios I used for comparison were FT1000MP > Mk >> V and Elecraft K2 [and before that IC735 and IC970, but these last > 2 >> are not really in the same class at all]. >> >> So the sound effect is not limited to one type of hardware or one >> particular piece of software. I agree that it likely relates to > the >> lack of crystal filters, etc. Unfortunately, in some situations > these >> filters seem to be a 'necessity' and that is the reason for this >> email/question, as I ponder which course to take as I upgrade my >> [contest] station here. >> >> The problem that has made me unable to go totally SDR, and for > which I >> really wish I had a solution, is that I do weak signal contest work > [50 >> MHz to 24 GHz] but live in an environment where during a [VHF and > up] >> contest I have more than 1 full-legal-power-on-multiple-bands > station >> operating within a few miles of me [one a multi op]. The extremely >> high RF levels from inband signals from these stations trash the > front >> end of any radio [SDR or not] that I have had available to me that >> doesn't have a narrow roofing filter [1st IF filter] to greatly >> attenuate the inband signal [as long as it is outside the roofing >> filter]. Given that the multiop stations usually have the lower > four >> bands [50-432 MHz] continuously occupied at 1.5 kW output, this is > a >> great problem on these bands. On the bands above 432 MHz it is of >> course not a problem. My FT1000MP has a 4 kHz roofing filter > instead >> of the stock 12 kHz filter, and the improvement this gives under > these >> contest conditions is tremendous, as long as the strong stations > are >> not within 4-5 kHz or less of where I want to operate. The K2's >> variable bandwidth crystal filter also does a very good job and > before >> I put the narrow roofing filter into the FT1000MP the K2 ran rings >> around it in terms of strong signal immunity. The SDR's fold up in >> this environment with spurious signals, increased noise, etc. >> >> I wonder if a hybrid approach such as something with an > architecture >> like the Orion I or II might offer the best of both worlds in terms > of >> performance in the environment described. The Ten-Tec Orion has a >> conventional front end and what is arguably the best roofing filter >> arrangement of any commercial conventional radio and follows that > with >> digital processing. Might this be the best way of achieving >> outstanding, 'bullet-proof' receiver performance? >> >> One of the great attractions of the SDR radios from an operational >> standpoint is the fact that they have wideband displays that from a >> EMEer or contester's perspective are extremely useful for showing > just >> 'what is going on' away from the operating frequency and on other >> bands. During a contest I constantly monitor bandscopes on all of >> 50-432 MHz simultaneously for this reason. The roofing filter > approach >> of minimizing the signal that gets thru to the processing stages > seems >> diametrically opposed to the use of a wideband display and 'doesn't >> fit' with the direct conversion model. >> >> One could achieve perhaps [for a station with my requirement] the > best >> of both worlds with an Orion-type receiver with conventional front > end >> with very good performance and narrow roofing filters, and stealing >> tiny bits of time to step the LO to sweep the band and generate a >> bandscope waterfall that way. Or if one designed the radio with > the >> 'usual' 2 receivers, using the second receiver hardware [when not > in >> dual-receive mode] to do the same. >> >> Of course with each of these designs, best-possible performance of > the >> RF preamp [if used], First Mixer, and Post-Mixer amp is needed, and >> these stages must be designed and implemented so that they are not > the >> limiting factors to strong signal performance. >> >> So my question is this. Is it possible to design a direct- > conversion >> SDR with wideband display that will under the strong signal > conditions >> that I have described offer the same protection from overload from >> strong signals while trying to listen to weak [and other strong] > ones >> that architectures such as the Orion II or the modified FT1000MP > for >> example provide? Will the 'best' implementation of a DC SDR always >> suffer IN THIS REGARD in comparison to the 'best' implementation of > a >> receiver using narrow 'roofing' filters? >> >> I ask this question now because [1] I just finished a contest > weekend >> and all of these issues are [too] fresh in my mind, and [2] I am in > the >> process of deciding whether to move for my contest station to [1] >> multiple Orion II's [operated using N4PY software so all can be >> operated from a single computer screen] or [2] Multiple >> SDR-1000/2000/etc, or [3] a combination of the two. It might for >> example be optimal [for a VHF contest station] to use an Orion for > each >> of the lower 4 bands where strong signal performance is paramount, > and >> then use an SDR for 903 MHz and up. >> >> Thanks in advance for your collective wisdom, and >> >> 73, >> >> Roger Rehr >> W3SZ >> Roger Rehr >> W3SZ >> http://www.nitehawk.com/w3sz >> > > > > > > Yahoo! Groups Links > > > > Roger Rehr W3SZ http://www.nitehawk.com/w3sz
