Blimey, Steve, you're starting to resemble an OF sometimes.

Even if you just use your computer to do nothing more than hook to the net, you can miss a rare one if you're booting up when a spot first shows up, too.

There are about a half a dozen things you can do with an SDR that are simply impossible with analog hardware, and every one of them makes contacts possible that would be impossible otherwise. Terry W0FM has some nice examples of this on EME that make the point perfectly.

There are a lot of shortcomings to SDR technology so far -- it's still in its infancy -- but sniping at the concept is merely silly, about as silly as sniping at SSB because a WWII surplus receiver had trouble with it.

73
Frank
AB2KT

<http://groups.yahoo.com/group/FT897/message/5232>
<http://groups.yahoo.com/group/hfpack/message/27656>
<http://groups.yahoo.com/group/hfpack/message/27666>

73
Frank
AB2KT

"As I understand it, amateur SDR designs still utilize traditional analog superhet circuits to receive and down-convert the signal..."

That is not correct.

The SDR-1000 uses front-end analog bandpass filtering, but that is for the purpose of antialiasing. The QSD does downconversion to the audio passband, conversion to complex form, and lowpass filtering in one step. The reason it downconverts to something other than pure baseband (0 Hz) is that soundcards roll off their frequency response towards DC. Thus the QSD downconversion is aimed at the flattest part of the soundcard passband instead. The remainder of the downconversion is done entirely in software.

The USRP performs broadband sampling. For some applications it uses a broadband tuner as a first stage, but the final channel downconversion is carried out in firmware.

There are good reasons *not* to sample the antenna directly. The chief reason is that A/D converters suitable for that task are exremely expensive. They need to have very high sampling rates *and* very high dynamic range. Even the broadest-band HF amateur applications are still relatively narrow by general standards, so expensive broadband sampling is basically a waste of money.

The USRP from Ettus Research is a much broader-band device than the FlexRadio SDR-1000. However it has a much lower dynamic range than the SDR-1000, and as such is more appropriate for high SNR applications at V/UHF, making a good fit with the kinds of signals that populate those bands. Weak signal work at HF of any kind is generally much narrower, and requires more dynamic range.

73
Frank
AB2KT

FWIW I got a Kulduckie half-wave for 2M at Dayton and put it on my FT817. The improvement in SSB performance over the stock 817 rubber duckie was nothing short of astonishing: it took the 817 from being useless with the onboard antenna to being quite acceptable for local area SSB on 2.5W.

It would be nice if the 70cm version were as much of an improvement.

73
Frank
AB2KT

You're right, the third pin on the Molex is for hard keying.

Someone kindly gave me one a couple of years ago. Rather than try to hunt down the right Molex, I simply brought the power leads outside to an external connector, and brought out a pigtail with an RCA jack for the keying. Works fine.

The RF sensing works too, but you'd probably want hard keying for CW.

73
Frank
AB2KT

An inverted L, running up one of the trees and across to the other, even hanging down a bit from the other tree, might do better than you think. Of course you'd need some radials too. It could be quite a bit better than some of the "flagpole" antennas out there now, especially if you could manage to put a remote autotuner at the base of the inverted L. I'm actually running something very much like that right now as a temporary measure until the weather clears up, and it works very well on all bands 80-6.

However, in the city, your real enemy is going to be RFI and EMI. Just the noise from automotive sources is pretty bad. For that and a number of other reasons, your best bet might be a tuned magnetic loop. The MFJ versions are reported to be pretty good. They're also fun to make yourself.

73 and good luck! I love working citydwellers.

Frank
AB2KT

Thanks all. I suspected it wouldn't matter much but it's always better to have experienced advice.

The exact situation is a flat roof (pitched only a few degrees) on a large dormer in the rear of the house. If the total height is less than 12 feet above the roof, it's next to impossible to see from street level in front. You know how some neighbors can be.

73
Frank
AB2KT

I'm getting ready to mount a set of yagis on the house. There are three of them -- 3 el for 6, 5 el for 2, 11 el for 432 -- all homebrew, total weight about 7 lbs, total wind area about 3.5 ft. In height, about 35 ft up.

The question is, what's the minimum height above the roof I should try to achieve, given a need to be a little stealthy? The 6m will need to be on the bottom, even though that's not optimal.

73 and TIA
Frank
AB2KT

As things stand right now, it's a reasonable bet that HF digital voice will remain a niche item.

Bad SNR and selective fading are the constant nightmares of HF digital in general. *Without* heroic error correction, the average quality you can expect from digital voice on the HF bands won't be any better than poor, fringe-area cellphone. *With* substantial error correction, quality will be better, but you can expect very significant latencies. Even then, the kind of corruption you'll see isn't nearly as graceful as with analog voice.

The situation is very different at VHF/UHF, where it's mostly line-of-sight, rarely point-to-point, and the (lighter) error correction can be carried out as the data is reshaped at a repeater.

I personally enjoy the digital text modes from a technical point of view, but as an operator find them a major pain. They're sluggish. You feel like you're developing cobwebs between overs. Digital voice is going to have that problem in spades. It's amusing to think that digital voice might find its niche among AM boatanchor enthusiasts, who are accustomed to holding the floor forever on high power, and have little interest in operating agility.

Each has its place.

I leave the little Ten-Tec by the big old manual tuner that my amp feeds; it's good for sharpening up a match before loading the amp.

My Autek analyzer is priceless for pruning and tuning a new antenna, cutting stubs, and things like that, but it's kind of fragile, and (like the all the other analyzers, apparently) it gobbles up batteries like there's no tomorrow.

The Palomar bridge is ideal for managing a portable multiband antenna in the field. It's nearly indestructible, it runs forever on a 9V battery, and I find it plenty convenient when coupled with a portable rx like on the FT-817.

If you had to choose only one, the analyzer would surely be the choice. But it doesn't render the others obsolete or pointless by any means.

Yup. The Ten-Tec bridge is a less-flexible version of a more typical noise bridge like the MFJ; it's set for 50 ohms pure resistive impedance. Pretty handy nevertheless.

Interesting. I'm really curious about the lore behind the 165' length. Do you suppose if you added a 34' leg to the ground side, it would pass for a "double-W3EDP?"

I believe some of the computer programs used for rig control can be programmed to display band information and limit operation to appropriate modes. But it sounds like your gear might not be computer-controllable...

The gray wires supplied with the MP-1 are kind of a nice gesture but definitely not the most effective solution to that counterpoise problem. It's easy to improve the performance of the antenna by replacing the supplied counterpoise with something more substantial. And even if you merely replace the 4 wires with ordinary hookup wire, you won't be losing anything noticeable in the way of performance. So you might as well go for better performance.

Almost any sort of wire will do. Almost any sort of connector will do, for example the equivalent connectors from RS, but even separate alligator clips on each counterpoise wire are satisfactory. Just use *lots* of wire, as much as you can manage. Using, say, 6 or 8 pieces of #26, insulated and stranded, each 15-20 feet long, will give you a much better antenna to play with.

However, if you are in a position to elevate the antenna and the radials -- like operating from a balcony, for example -- you'll find the best situation of all is to use one or two single counterpoise wires tuned for the band of interest. But that's a very different situation from being near the ground, where you do need a lot of counterpoise metal, and it's detuned by proximity to the earth.

K5LXP has it about right. I've settled on the 350Hz CW filter after having had the 500 and 250 filters in the rig for a few months.

The 1.9kHz SSB filter isn't really worth much, although there's still one in my rig. The PBT is much, much more helpful in reducing SSB QRM.

Even better is a decent audio filter. When using the rig at home I have an old MSC SCAF filter on the audio out which occasionally makes the splatter on 17m tolerable :-)

OTOH I don't work HF voice much anymore, so the voice filtering is not a big issue. The CW filter is, though.

73
Frank
AB2KT