Saturday, May 22, 2010

Back from Dayton

My feet and knees hurt, I had no voice left and had to endure flying home in thunderstorms, but boy was Dayton really fun. I really enjoy meeting in person the folks that I correspond with on the reflectors and get to spend some quality time with the Austin crew (I live in North Carolina). It has taken a few days to recover and having to work a couple 20 hour days for my "day job" the week after I arrived home didn't help much either. I now feel more normal again after a few good nights worth of sleep.

By now you have probably figured out that there hasn't been a lot of new things for the FLEX-1500 testers to be testing due to the lack of activity on the ye ol' blog. That is because for the past several weeks, FlexRadio has been completely revamping the audio/control interface for the FLEX-1500. The version I have been testing utilizes the native Windows USB HID audio drivers. This approach would have allowed the FLEX-1500 not to have used a custom device driver which would have made it easier to install and support the FLEX-1500. Testing however showed that the native Windows USB HID driver just didn't have the stability, performance and timing required to deliver a really high quality product. So the developers and engineers started building their own streaming kernel mode USB driver for the FLEX-1500 and initial lab testing has been really promising in the performance arena. The RX->TX->RX switching time are really fast. I can't say how fast yet, because it all has to be packaged up with PowerSDR, but the numbers are really small (small is good). The CW folks are going to like this radio.

On another note, one of the lead developers, Eric is now experiencing a maternity non-maskable interrupt. His XYL gave birth to their fourth child on Friday May 21. Samuel Ray and mother are doing fine, but the timing of Sammy's arrival will put a slight delay the driver integration into PowerSDR progress. He will be working from home and coding in-between feedings, diper changes an taking care of the other kids while his wife is recovering. So a bit more patience will be in order. I know the whole FlexRadio engineering & Development team is working long and hard hours to get PowerSDR v2.0.0 ready for the public beta and subsequent final release. This is another "baby" that they really want to get birthed too.

Wednesday, May 12, 2010

Off to Dayton

I am off to the Dayton Hamvention with my FLEX-1500 in tow. Having received one of the first pre-production beta units, my FLEX-1500 needs a few hardware updates to bring it in line with the other production units. There was a couple of minor codec filter changes to reduce the DC noise at the "0 Hz IF" and a change to expand the capabilities of the RX preamp to give it more stages or levels to choose from depending on band and band conditions.

I'll be part of "Team FLEX" working the booth and doing hands on workshops at the Flex Banquet on Saturday night. Make sure you stop by the FlexRadio Systems booth at location SA315. See ya there!

Saturday, May 1, 2010

The FLEX-1500 goes ATOMic

I dug out my Intel ATOM 330 (D945GCLF2) XP SP3 machine from the bottom of the closet (it was stored there when it wouldn't play well with the FLEX-300) and after spending an hour or so doing BIOS, driver and other updates, finally got it running with the FLEX-1500 using the latest version of PowerSDR v1.19.3 from SVN. The results are better that when running with the FLEX-3000 (since it basically didn't).

On receive with no RX DSP special processing, it runs at about 37% utilization.

With NR/ANF/NB1/NB2/BIN enabled on SSB, it caps out at ~55% CPU with no degradation of the display or audio.

Now load up VAC 4.09, Com0Com and Fldigi for operating digital modes and the CPU utilization jumps up to ~48%. If you add on a logger or DDUTIL, I suspect I would be at the ragged edge of what this little PC can do.

This is just how it played on my ATOM desktop machine so there are no promises with your ATOM based laptop. Your mileage will vary.

I also finished my Rubidium (Rb) disciplined oscillator project. I put a LPRO-101 Rb stabilized reference oscillator in a forced air, ventilated enclosure with a DEMI 1x4 10 MHz distribution amplifier/filter so that I can slave multiple radios off of one Rb frequency source. It's 10 MHz stabilized signal is driving the DDS for both the FLEX-5000 and the FLEX-1500. I hope to have a few pictures and a new frequency stability run using it with the FLEX-1500 in the coming days. So, its true, the FLEX-1500 is really "gone atomic"!

Friday, April 30, 2010

It is Shipping!

Yeah Baby, that is right. The first "off the production line", no more beta units, a "real McCoy" FLEX-1500 shipped on April 30,2010 to an undisclosed recipient who is part of the extended beta testers group. This is a BIG milestone as the production hardware boards are finalized and production ramp up can begin as soon as all of the materials have been received to start the production runs. That will happen in a couple more weeks. For the hundreds of you with FLEX-1500s on order, please be patient for just a bit longer. They are coming soon. FlexRadio will notify you when they start shipping.

To prove one really did ship, there is photographic evidence of the unit being picked up by UPS


A big thanks goes to Graham (KE9H), pictured in the center who was the primary engineer on this project who took it to the finish line. That is Julie the Shipping and Receiving Manager on the right.

Now PowerSDR v2.0.0 isn't complete yet, but it is getting there. More coding was done on the BITE routines and that work is getting close to being done. Additionally, a slew of bug fixes and enhancements were added to the pre PowerSDR v2.0.0 alpha code this week, fixing a few long over due issues like a wider and user definable deviation for FM, the last changes for the software to be fully Vista/Win7 compatible so that it will operate in user mode and not need any of UACs "tricks" to make it work, new NR and ANF default settings, ANF was fixed so that it works in AM & SAM modes and a new crash recovery option that will allow you to reset the database if it is so hosed up that PowerSDR will not start are just a few of the things being added. There is a big push to get PowerSDR v2.0.0 done "real soon" so expect a lot of development activity in the coming days. The beta testers are going to be really busy.

Tuesday, April 27, 2010

I know, it has been awhile.

I know it has been awhile since my last blog entry. There really hasn't been much to blog about. Most of the BITE functionality has been put into the FLEX-1500, but there are still a few test that need tweaking. The input mixer functionality was added and that was about the biggest change recently.

One thing I have been doing is interfacing a Rubidium (Rb) disciplined10 MHz clock source to the FLEX-1500. I am using a LPRO-101 that I got from WB6RQN and I have been putting that together in a nice enclosure with a 1x4 10 MHz distribution amplifier so I can drive the FLEX-5000 and FLEX-1500 from the same clock source. There were a few software changes needed to be able to transmit and receive using the external clock source with the FLEX-1500, but it is working fine now.

The main work being done at this time is optimizing the audio subsystem for response and quality. There have been a few things uncovered using the Window's HID driver so the developers are digging into it. More on this when there is something to report.

Oh yeah. One thing I did forget to say. I established a new FLEX-1500 distance record a few days ago. On April 21, I had a nice "rag chew" PSK31 QSO on 14.070481 MHz with Harald, DL2OCE in JO41XW on the FLEX-1500. This was a distance of 4300 miles. Antenna was a 20m dipole @ 45' and the power out was about 3.5 watts as reported by the wavenode. I received a nice 579 signal report and an IMD report of -28 dB.

Propagation was good that morning and what was most important was the IMD numbers, indicating that the transmitter was operating very cleanly. This is going to be a cool little radio.

Tuesday, April 13, 2010

Still Testing...

Development of the BITE (built in test equipment) is still progressing. There isn't a whole lot to report at this point. There were some updates to the preamp configuration late last week. Rather than having a three state preamp (on/off/attenuate), it has been changed to provide +30, +20, +10, 0 and -10 dB of pre-amplification and attenuation. This will help with sensitivity on the upper bands and the two level attenuator will help with blocking dynamic range on the low bands.

Other than that, I have been listening to a local beacon on 6m since I got my 6m Moxon back in the air this weekend. It had a little accident a few weeeks ago when I was taking it down for it annual pre-eSkip maintenance. Regardless what the farmers say, planting an aluminum in the ground will not make it sprout small antennas; no matter how much you water them. Par Electronics was very prompt in sending me element replacements and for that I am very grateful. I fixed the antenna mast mount by engineering a simple tilt-over mechanism and put some shear pins in the mast sections to prevent the sudden and unexpected rotation that caused the damage. So hopefully there will not be any more accidental antenna "plantings" in my future.

Tuesday, April 6, 2010

Studying for exams and new feature alert!

One of the most intriguing features of FlexRadio Systems' SDR is their ability to calibrate themselves using BITE (built-in test equipment). There are a series of calibrations tests that the radio can run on itself to optimize various operating parameters or check the operation of the hardware. Some examples of these tests include checking the integrity the RX/TX filter banks, calibrating the S-meter so that -73 dBm is S9 on all bands and suppressing the opposite side band for single sideband modes.

Adding the BITE functionality into PowerSDR for a new radio is not a trivial task, but it is a foundational aspect of hardware integration that must be done right first off. These are the activities that have been going on for the past several days and why the blog entries have not been as numerous. The FLEX-1500 is studying hard to pass his tests. It is also a fairly slow time for the beta testers. The developers work all day to add a new hardware calibration test and the testers wring it out in 15 minutes. It isn't very exciting for the testers, but it is a critical path process that will lead to getting other features working like CW and optimizing phone modes.

There is a new feature to talk about that has never been publicly stated. You will be able to connect a receive only antenna to the FLEX-1500. That is right. You can select the transverter XVRX port or the transverter XVTX/C as a receive only port with the PA (5 watt output) port in addition to being able to the the PA port as a transceiver port. This is controlled by an antenna form similar to the one for the FLEX-5000. Like the FLEX-5000, there is an expert antenna mode where you can define the RX and TX antenna on a per band basis. It is a really nifty feature, especially for a low-end QRP rig.