Adventures in SDR

Tweetbutton tweet

k7ip Wed, 07/22/2009 - 04:29

The first thing that comes to mind when you see the Softrock Lite II kit is, "How are my giant fingers ever going to be able to build that dinky thing?" The second thing that comes to mind is, "A $13 dollar receiver can't be all THAT great!"

We'll see how it goes...the reports from the Softrock Yahoo group are VERY enthusiastic about the quality and performance of these things, and the great DSP software that goes with them.

I'm still concerned about the mutant fingers issue. My time with L-3 taught me the value of having a good stereo microscope and rework soldering station to deal with this virus-sized circuitry. I acquired a pretty decent Hakko ESD station a few months ago during my foray into repairing my daughter's Dell laptop, but spending half a large on Bausch and Lomb optics is still a bit out of the budgetary galaxy. It's closer, though, and is fairly close to the acquisition of a good drill press in tool realm importance.

Secondary issue, now deciding priorities between this and the preamp project. I'm guessing the preamp will win, there are no toroids to wind in that kit.

7/23 update

Wow. I have been watching Tony Parks' site: for updates, as he is the kit supplier for most/all of the Softrock products. His inventory his been very thin lately, and I happened to hit the site this morning just in time to score the complete V6.3 RXTX kit with two sets of BPF/PAs (One was 30/20/17m, and I can't remember if the other one was 80/40 or 15/12/10m) before all his kits were again exhausted. There has been a lot of chatter on the Softrock group about the upcoming 6/2 converter for the RX, so I've been paying close attention for that reason. Call me crazy, since my project list grows every day and I haven't really produced anything tangible yet! Anyway, the kits have been out of stock for a couple of weeks, so this seemed like a golden opportunity.

I wonder if anyone has used a webcam for a kind of projection microscope?

7/27 update

Because of out of town company this weekend, I didn't get a chance to do any real project solderwork. However, I was very pleasantly surprised by the arrival of the V6.3 RXTX kit on Saturday. Since I ordered it Wednesday afternoon, I'm impressed by Tony's response. I was also impressed by the fact that through some typical bonehead blunder on my part, I had accidentally ordered two identical PA/BPF kits. A short email exchange with Tony has replacement parts coming - free of charge - to convert one of the 30/20/17 kits to an 80/40m kit. This is 60's era customer service...why can't I get that kind of satisfaction from Comcast, or Verizon, or US Bank, or (insert your least favorite money sink here)?

One thing I did get done was some additional PowerPole upgrades and adaptations. Most went pretty easily, but my need to cram #12 wire into a 4 pin Cinch-Jones blade connector totally exhausted my extensive expletive vocabulary. Fortunately it became time to go to the airport just before I turned the whole connector into a puddle of smoldering plastic on the dining room table. I was also reminded that trying to use a 30 watt iron to solder connections while having a ceiling fan blow directly on the work is an exercise in extreme frustration!

7/28 update

What is with the heat? No soldering iron work yesterday, just sweltering.

08/17 update

Back to work on the Softrock Lite II for 30m this weekend. No construction, but I got the inventory finished and am anxious to get started. Boy, those parts are small. I spent a lot of time under the magnifier hood this weekend! And let's not minimize the importance of a GOOD high intensity lamp in all these efforts.

Tony's kit was complete, and even included a couple of spare parts. He really does run a quality operation out there. I can see that there's going to be a continuing need for 3.5 mm stereo jacks in my future. The Softrock is so tiny that the antenna and audio jacks will probably dwarf the rest of the receiver. For photos of a complete unit, go to and scroll down to the bottom of the page. On your screen, the unit is about four times actual size.

Along with the SDR inventory and finishing off the rebuild of the sound card interface, I also managed to get the last of the power pole adaptations made. One power cable for the RFC 2m brick and one for the 10 amp bench supply did it. I could probably convert the tracker and TM-V7 cables as well, but those connectors are already wired into the power harness, so it would be kind of a waste.

08/24 update

I apparently have ADD. A couple of weeks ago, I decided to also order a Softrock Lite II for 20 meters in addition to the one I have for 30. I inventoried the parts and decided to get started on that one this weekend rather than the 30 meter version because I can actually hear stations occasionally on 20 and it might be more useful in testing and troubleshooting to be able to hear something besides enormous amounts of noise. So...I got the power supply built (whoopee, five components) and part of the LO. And had my first experience with surface mount caps. Those things are REALLY REALLY TINY. True to form, the first one I tried to put on the board sprang on its own out of the tape strip and disappeared, apparently into a parallel dimension. Good thing Tony includes an extra with each of the chip cap values. True to form, I misinstalled the VERY FIRST component on the board...Geez. Fortunately, I discovered the error almost immediately and was able to continue without further issues. Being thoroughly embarrassed, I doubled the amount of double checking I was doing on component placement. Duh.

Getting the first one on the board took me about 20 minutes. Not because it was particularly hard, but because every little tremor of the hands is magnified when you are trying to hold those tiny suckers down with a pair of tweezers while you poke at it with the soldering iron. It's also very clear that you need to secure the board so it doesn't move around at all. The third hand I use is great for positioning, but holding down the caps exerts just enough pressure that the whole assembly moves. Next construction session, the board goes in the mini bench vise for attachement of the surface mount components. Progress this weekend was pretty slow, but I think it will go faster with better technique, and after all, this IS learning, isn't it?

09/08 Update

With the long holiday weekend, the soldering iron was very active, and I completed the Softrock 20 m kit. Soldering the SMT caps and chips gets much better with practice, and thanks to the video links on the WB5RVZ construction site. The Hakko 936-12 soldering station that I use is still a tad large, and I had better luck with my old 60/40 .032 lead solder than I did with the roll of 2% silver bearing .015 fine wire that I got specifically for this kit. I'm going to need to figure out if the solder needs more heat because of the silver, and whether the tin/lead tinning of the tip may have something to do with it. It seems that the joints flowed a little better with more heat than I thought would be needed, too. The iron temp (if the calibration is to be believed) is about 680 F, somewhat higher than the solder melting point is supposed to be. RF Cafe says 60/40 should be 368 F and the 2% Ag solder should be 354 F.

One of the more frustrating parts of the kit - for nearly everyone, I hear - is the winding of the toroids. It's not that it's hard, but corralling that #30 wire and keeping the turns reasonably spaced seems to require no less than four ulta-miniature hands. Two huge and jittery ones often get in the way!. However, I can also see that this technique will improve with practice. One tutorial suggests holding the turns in place on the toroid with a small slice of hot melt glue melted on the toroid using a heat gun. I don't have a heat gun, but found that Ace sells one for about $20. My next tool acquisition, I guess, since there's now a stack of Softrock kits sitting in the parts bin waiting to get finished.

The actual kit itself is pretty high quality. The Hakko medium sized iron has the smallest conical tip available, and getting it positioned on the tiny solder lands takes some persistence. If you position the uncut through hole lead to that the eyelet is exposed as much as possible, it's fairly straightforward to get a good connection.

The instructions the WB5RVZ has on are superb. The only complaint is that a couple of the current measurements with a current limiting resistor seemed to be misprinted. With the current limiter out, all the test points and interim measurements were well within acceptable tolerances of the measurements indicated. I had to skip the actual detection of the LO and divider signals since I didn't have my hf rig along during construction.

I also borrowed a Sound Blaster Live! 24 bit USB sound card from N7IPB to use since the Gateway mic jack is not stereo. This is kind of where the project effort ended, as I didn't really have time to get everything adjusted properly before it was time to put the tools away. The Softrock group mailer says that there's a configuration that needs to be done to set the card to 24 bit sampling rather than the default 16. I need to research and do that, because it appeared that the Rocky spectrum was only 24 kHz wide, a symptom of that adjustment.

Good stuff to play with next weekend at the Summer Gathering.

09/14 Update

Oh, WOW.

That says it all. The Summer Gathering had some really great presentations, but one of the hits for me was the demo of the Flex 3000 and PowerSDR software. SDR is redifining ham radio, and very quickly. The Flex is a great rig with a lot of capability, great features and a nice slick package. It's a winner!

But the OH WOW really comes from the Softrock. A version of PowerSDR is built especially for it, and I can tell you, it is unbelievable. That little .9 x 1.5" board has the heart of a Collins 75S3 lurking inside it. Comparing it to the Flex isn't really apples to apples, but since the same software powers both, a comparison is unavoidable. I didn't make any objective tests of sensitivity, but I can tell you that BOTH the Flex and the Softrock were LOADED with signals on 20m. Here is the totally astonishing (but not really) part. Since both use essentially the same software, there is no difference in the selectivity. Let me say that again...there is no difference in selectivity between the Flex and the Softrock. Why? PowerSDR. What a totally AMAZING piece of software it is. At one point, I was using the 25 hz filter on a single CW signal, totally intelligible, with NO ringing whatsoever. No Collins filter ever did 25 hz, yet this software makes the $13 Softrock a better receiver than anything Collins ever built in the way of mechanical filters.

Now, I do have to admit that the antenna I had available up at Valley Camp was a couple notches above the average - a G5RV 100' up in the trees, but even taking that into account, the Softrock is an astonishing performer.

On to the RXTX-XTALL!