FlexRadio Systems SDR-1000 Notes

Comparing the SDR-1000 to the Icom Pro III and Elecraft K2

Anyone considering the Flex Radio SDR 1000 might be interested in reading about others who have purchased this radio. Here are my notes and comments.

Author's Background

• Alan, WB6ZQZ, Extra class, licensed since 1967.
• Other rigs - Icom 756 Pro III & II, 746Pro, 746, 761, 706, 703, Elecraft K2, Yaesu FT897, Kenwood TS520, etc
• Education - BSEE, MSCS
• Professional - Design/Build Control and Data Acquisition Systems - combining Electronics Instrumentation and Computing into Systems
• Other Interests - Pneumatic Antenna Launching, Camping, HFPack, HF Mobile & Portable

Notes

November 2006 - First Experience

Ordered Flex Radio SDR-1000, w/ 100w internal amp, USB parallel adapter, Edirol FA-66, Heil Handi-mic, Griffin PowerMate Knob, Elecraft XG-2 Receiver Test Oscillator. Everything came in good shape in about a week (from 3 sources).

Cost of SDR-1000 - The radio is about 1300$US, add the sound card, mic, parallel/USB interface, and a computer, and it is about 3K, or the same price as the Icom Pro III. You can't just use some old junker computer for this, it requires good performance and recent operating systems. Linux support is coming, but this is basically an XP box for the moment. One thing that bothers me is that, for $1300, the hardware you get is a lot less sophisticated than an equivalent cost Yaesu or Icom. I would like to see some significant improvements in the radio hardware, but more on that later. The Software that you get is pretty amazing and makes this thing work.

On the SDR interior packing box I find a large warning sign - "Avoid Frustration... READ the Quick-Start Guides and the operator's manual OR THE RADIO MAY SELF-DESTRUCT!!!!" Note that I have read all the manuals over while waiting for the stuff to arrive, and I attempt to read/follow the guides as I set up the system.

First problem - When I ordered the Mic from Flex, I had read that it was recommended to cable the mic directly to the Edirol (Edirol quickstart guide), and so I did not order the Flex Handi-mic to SDR cable. Turns out that now I have no cable at all for the mic, and that Flex does not make one like that (mic to Edirol). Major Documentation problem.

So now I have choices. Order the Flex cable and run mic audio through the SDR, forcing a balanced audio system to be grounded on one side (not a great idea but it does work), or make/buy a cable.

Make power cable for SDR. Wonder where the fuse is? None in the cable, none visible. They supply a gold-plated banana plug to mate with the rig's binding post/banana power jack, but this unit sticks out a long way and has exposed conductors sticking out. Instead I take a pair of 1/4" to #12 ring lugs, install them on the supplied #12 red/black cable, cut and twist the cut rings so I can get them onto the binding posts, flatten them back out, and tighten down the binding posts. I cut the cable to 18" and applied Anderson Powerpoles to the other end. Using Astron RS-20A linear power supply for now.

Initially using a new AMD 64 bit laptop for the computer. It is about 1.6 Ghz and has 1 Gig of ram. Install the Edirol drivers. Edirol works. Try it with PowerSDR 1.6.3 software. At 192K sampling the CPU load is 100%. Drop to 96K Sampling. CPU around 50%. Connect the SDR, Edirol, USB, Laptop, Headphones (to Edirol). Re-Configure PowerSDR software. Lots of noise but no signals. No effect from connect/disconnect antenna. Relays working. Puzzling. Ken WB6MLC suggests startup order may be the problem. Yes. Restarting PowerSDR makes it work. Strange, you would expect that when selecting radio type, parallel type, etc. it would re-initialize the hardware. Apparently it only does a complete enough job when the PowerSDR is started initially. After that it assumes perfect synchronization with the hardware. In my experience this is very dangerous assumption for a control system to make... (Control Systems are what I do for a living). It leads to lots of time wasted and unhappy users...

So now I have a receiver. Frequency calibration is off. Use WWV at 2.5 to recal (can't hear the higher frequency signals well enough tonite).

Listen to a local CW net on 80m. Noise level pretty high. No help from blankers. Noise Reduction helps a little, maybe. Set cw filter to 25 hz. Pretty nice. One problem with tight CW filters (even if they don't ring) is that all the noise is now the same frequency as the signal. NOTE it turns out later that the filters don't necessarily work as displayed So you can select 25 hz but find that it is really only capable of 100 hz. The impossible filters should gray out...

MDS - Minimum Discernible Signal - Comparing to ProIII. This on 80m on a noisy band, so not a great comparison. One of the CW signals in the net was very weak here. Can't tell any significant difference between SDR and ProIII on that one.

SDR receive audio sounds slightly cleaner than ProIII. Probaby less distortion due to fewer stages.

Tuning with the Griffin Knob - The Griffin Knob is very pretty and has (their terms) a cool blue led. It does look nice. Has good feel, though it could use a finger dimple and perhaps a rubber skirt... The problem is the knob gets digitized into windows events like keyboard or cursor motions. So it is quantized a bit more coarsely that I would like. It is nice to use for fine-tuning SSB or CW.

Binaural mode is neat. Very neat sound. Filters not available?

Fan Noise - the fan is in front of the rig, and the filter will collect dirt and display it for visitors. With the fan in this position the noise is maximized. Not a great design choice. The noise level is noticeable but not loud, but adds to band noise. It does not speed up or slow down, so it is much noisier than the ProIII when the Pro is quiet, which is most of the time. When the ProIII fan cranks up it is similar to the SDR, but this is not frequent or for long.

Cables Aplenty, and the stock cables are waay too long

Computer Issues - getting a bit of crackling, which I understand is from computer latency. Have to work on that. (fixed by faster cpu)

Powering from Battery Bank may be a problem, power on/off protocol is complicated. Need an external 12V switch to remove DC from the rig, according to the instructions. Other rigs do not have this problem.

Power Consumption??

Transmitter Duty Cycle - poor - now only up to 40%. Not comparable with other rigs in this price class.

Frequency Stability not great

Parallel Cable to PC really a pain. USB variant better, should have USB to parallel adapter inside of SDR instead of the way it is done now.

Mic Cabling/configuration not well documented, problematic, hard to know what to order when buying

Tune Metering - can select SWR in tune mode, not in other modes. Would be nice to select SWR in Tune mode and Power or ALC in XMT mode - there should be separate meter modes in Tune.

PTT lockout switch (mechanical) would be nice. Concern about computer crash setting transmitter on. Is there a failsafe timeout timer?

PTT safety - what happens if a crash leaves the PTT on? What protection is there?? Timeout timer???

Noise Rejection (NR) works with little distortion, with default settings, and makes a mild improvement in the noise level. Not impressive.

The ANF (auto notch) works, but adds more distortion than the one on the ProIII. It also lowers the gain a lot. NOT VERY USEFUL. Again, this with default settings.

ProIII Spectrum Scope memory mode nice, compare to SDR Histogram. SDR Histogram with AVG mode on is better.

SDR Panadapter resolution nice

Speakers - no amp, low power out, etc.

Sub-band markers - not always visible, depending on resolution

Second VFO markers - not shown

Narrow Filter Options are selectable even when they do not actually work

Changed computers to a desktop type 2.66 Ghz P4. CPU load went down to 25% at 96k sample rate, changed to 192k sample rate, cpu load at 40% or so.

After several days, I am getting the hang of the receiver tuning. You can view a signal on the Panadapter display (which covers 192 khz on my setup now), judge where the carrier should be, and click on that spot. The receiver is instantly tuned to that frequency, and you can copy the signal. It may not be precisely tuned in, but it is close. Due to the large energy in the low frequencies of most SSB signals, and the abrupt rollof near the carrier frequency it is pretty easy to hit the carrier frequency closely.

RFI on transmit is a major problem for me at this point. The receiver is working nicely, and is very clean and 'hi fidelity' sounding. The transmitter does not work very well due to RF feedback issues. I am working through different combinations of grounding, snap on ferrites, and so forth. Using an inverted L fed with balun/coax and tuner (took the SGC 239 out from the base of the antenna after it started acting up). There may be some problem with the antenna, however the SDR with all the cables and boxes is quite susceptible to RF.

RF Feedback Fixed
To summarize a lot of experiments, the advice on the Flex website about connecting the mic directly to the Edirol appears to be poorly advised. Despite many configurations and tests of this configuration and variances thereof, SSB was unusable until I recabled the mic to run it through the SDR1K box. Now we have a working rig.

RF Transmit Metering
It is difficult to tell when the levels are right for SSB. I think I will get a really excellent external wattmeter as otherwise one cannot really tell if things are set up right. The LP-100 looks interesting...

When in the tune mode the SWR readout works. Great. When in transmit it does not. Huh? The ProIII does not have this problem... It would be reasonable for the transmit meter to revert to SWR when in the tune mode, and something else in the operating mode - it should be separately selectable in metering function between modes Transmit vs Tune. Or have more metering visible, as many of the modern rigs have. There is no reason to not have SWR visible all the time. Plus power and ALC.

Just received the LP-100 wattmeter. It shows 1.5 watts when the tuning is set for 1 watt. Can easily tune the Palstar AT1KM tuner with this 1.5 watts!

At full power on 80m the continuous carrier power output is 113 watts. This in tune mode set at 100. Individual dits are somewhat lower at 98 W.

ONE PROBLEM - when in CW mode, touching the mic PTT causes carrier to be sent. NOT GOOD. The mic PTT should be disabled or merely shift to XMT mode w/o carrier.





-- AlanB

---

CategoryHamRadio