Saturday, March 17, 2007

SDR-IQ First Impressions

Right. So what have we got here then? First of all an exceptionally small PCB, measuring around 9 cm both ways. With a USB connector, a BNC Antenna connector and an RS-232 port. The thing is powered via the USB cable. More about the physics on RFSpace's homepage.

My main concern was my local Loran C station. 14 km away, 250 kW ND, and my North America beverage pointing more or less at it. I had reason to worry. The Mediumwave band was literally filled with Loran C noise, while the other receivers (fed from the same splitter) had no problems - or at least not more problems than they've always had. Oh dear, oh dear. Shortwave sounded rather ok though. So how do we get rid of the noise? Well, reducing amplification is one simple way of dealing with the problem, so I reduced the IQ RF-amp with 10dB. That helped a lot. But made the receiver rather deaf (see sensitivity measurements below). Now, I have a couple of Loran C remedies at hand. One is a "Loran C Notch Filter" manufactured by the USCG Yard, recently bought from Dallas Lankford. It was exceptionally effective, and I could turn on the IQ RF amp again. But the filter itself has considerable attenuation, around 20dB on lower MW and 10dB on higher MW, so there is need for amplification. I put a 11dB preamp into the circuit, and it worked quite well.

Then I remembered that I have in fact another 100 kHz/200 kHz notch filter around, made by Swedish DX-er Stefan Wikander several years ago. I replaced the USCG Notch with Stefan's. The result was excellent, and with no attenuation. So, at least for the moment, the Loran C noise problem is solved.

So then. Sensitivity. How good is it? To be honest, I was disappointed. With full internal RF and IF amplification on, I found the AM, 6 kHz, 30% modulation sensitivity to be -101 dBm, or 2 uV. I only measured 1400 kHz and 10000 kHz but the results were identical and I don't expect the figures to vary a lot.

Audio quality: Very good.

Selectivity: Versatility beyond anyone's wet dreams. Up to 12.9 khz in SSB and up to 20 kHz in AM. Very, very effective, even compared to similar bandwidths in my IC-7000. Unfortunately no Passband Tuning, no Notch Filter, and a Noise Blanker that I haven't found anything to test on yet.

User interface: No, no opinion yet. There is a learning curve to the software, Spectravue, and I'd like to see what it does and does not before I make a judgement. Not totally intuitive though, but it seems to have a lot of power inside for configurations and customising for the user.

The IQ is easy on the PC. The PC I use is an almost 4 years old Toshiba Satellite with a 2.0 GHz Celeron CPU and 512 MB RAM. I thought it would prove to be too weak, but it appeared strong enough to record a 150 kHz span of RF data. I couldn't get it to take 190 kHz though but actually I didn't expect the 150 either.

Enough said so far. During the 4-5 hours I've used the IQ, it has not powered down once, or disconnected or made any other unexpected twists. That is reassuring. The picture shows a "Screendata" centered on 680 where a very potent KBRW Barrow AK resided. On 675 we see an equal-strength NRK Røst, and my semi-local Radio Rossii Murmansk is the strong one on 657. I have made my first RF Recording, a 10 MB file (containing a few seconds of a 150 kHz span).


4nradio said...

Hi Bjarne,

Nice initial comments; thanks for posting your thoughts. I have Walt Salmaniw's new SDR-IQ in the shack here; I am constructing a proper enclosure for the device and will be delivering it to him whenever we meet next.

I haven't yet powered up his SDR-IQ to see how it performs here. However, I was surprised to see the rather high noise floor between signals on the SDR-IQ at your location. My SDR-1000 here in suburban Seattle USA typically registers a -130 to -135 dBm noise floor between signals on its calibrated display. I wonder if your Loran C station is still causing problems despite the notch filtering?

I'll be interested to follow your further SDR-IQ observations. When I get Walt's powered up I'll let you know what I find. Also, I plan to be at a coastal DXpedition in 2-3 weeks, and will probably have Walt's SDR-IQ along for comparisons. 73, Guy

Walt Salmaniw said...

For obvious reasons, I'm very interested in hearing more about how the IQ works in the real world of dxing! I'm hoping to use my IQ as a kind of "RF vacumm" during interesting openings at my cottage on the Queen Charlotte Islands, an area of generally low noise.

Anonymous said...

Hi Bjarne;
Nice to hear about your results. I agree with you about the dynamic properties & sensitivity. If anyone has compared or measured values to SDR14 we should be pleased to hear of that. The noise floor really depends a lot of the QTH & antenna. I have tried mine at low noise rural qth with BOGs showing NF down to -140 while at my home with amplified ALA100 shows a NF typically as high as -110 dB...
73s Tarmo (some tests also at pudxk blogs)

4nradio said...

Hi Bjarne,

Tarmo's comments about a low noise floor down to -140 was encouraging. On my SDR-1000 blog, a number of the screenshots over the months show a noise floor in the -130 to -135 range. The best I've seen is about -140 while DXing with a Beverage antenna at a very quiet location on the Washington coast. Here's one blog entry showing a couple of screen captures with a -130 to -135 floor between the signals: These were taken while DXing from my home QTH. All of this considered, I think you too will see (and hear!) a very quiet SDR-IQ when you are in a better location with various antennas. Best wishes, Guy

Anonymous said...

Hi all;
I forgot to mention that my -140 Noise floor with BOG was with SDR-14. I wonder if the noise floor of IQ is a bit higher, am I right? As fas as I know the NF of SDR-1000 is "best" ever seen in these sdr-rx´s.
73 tarmo