Space radio NR, ANL, and other updates

Here is an article I saw on the Flex reflector, which speaks about using SDR in space

http://gcn.com/articles/2009/03/09/nasa-software-radios.aspx?s=gcndaily_160309

one of the stalwarts on the reflector designs space radios for JPL. His radios do -157dBm MDS

Now that's what I call a radio!!

The band stunk this morning so nothing to report on that front.

Bob N4HY has migrated from his custom beta thread to the main beta thread on the SVN, the code he has been working on regarding the RX image rejection is automatically adapted to the best performance on the frequency you are operating in real time. He has also redone the Noise Reduction and Auto Notch. For the first time NR really does something. I have owned half a dozen radios that claimed NR and half a dozen external DSP filters that claim NR. Usually they just add a lot of beeps and boops to the noise. This one has no boops but there is a clear change in the amount of noise in the pass band. The ANF is great. I was listening to a S-3 SSB signal last night when a s9+10 carrier came on. The ANL took out the carrier, then another S9 carrier came on a couple hundred hz higher and both were eliminated. In the mean time the AGC was left unaffected since the ANL comes before the AGC loop in Bob's scheme. Slicker n blobs of mucopolysaccharides (aka snot)!!

TX adaptation is soon to follow

73

SO2R part deux

In the Full duplex article below I broached the topic of SO2R. I wanted to elaborate a little further on this mode since there may be some interest. Its a very cool feature.

The way the radio is set up you have 3 separate processes going on all the time RX1 RX2 and TX. The antennas can be independently assigned as well as amp keying

Here is a shot of the antenna selector in expert mode.



In this case I have RX1 and TX assigned to antenna 3 when the radio is switched to 80M. I also have amp keying set to output 3. What this means is when I have 80M selected TX is directed to output to ant 3, the RX is connected to ant 3, RX2 is connected to ant 1, and if I transmit the amplifier that is in ant 3's line will transmit. So in this screen I control the RX and TX including the amp for my 80M part of the SO2R experience.




This is my 160M screen. In this screen I have ant 1 connected to RX 2. TX is also connected to ant 1. RX1 is still connected to ant 3 and now I can control the amp that is in the ant 1 signal path. If I transmit my signal goes through the ant 1 amp to ant 1. There are 2 totally independent signal paths from the radio to each amp and each antenna.

Here is a shot of the radio with 80M chosen


You can see the red square on VFO -a and the antenna selector out to the side.

Here is how it looks with 160 chosen


Note the red square again. To switch to 160 I simply click that TX spot with my mouse. I can tune 160 while doing a run on 80M Here is a shot of me transmitting on 80 1500W on 3504 while listening on 160. Note the RX2 S meter reading in the above shots and also in this shot.


I chose 80 and 160 in the middle of the day as the examples so you can see there are no spurs or any kind of signals present on the 160M panadapter while I have 1500W pumping out the wire. The antennas are separated by about 100ft. I get this same kind of response between any 2 bands set up like this.

The way the audio is set up I can pan RX 2 across the stereo space, in a single ear or in both ears. RX 1 is in both ears unless I turn on the multiRX receiver. In that case I can also pan the audio from those 2 receivers across the stereo space. By controlling the various volume controls I can also set up the mix of loudness from any one source.

What the heck is a watch receiver? (note the multiRX button is now active)


A MultiRX or watch receiver is a receiver done in software. In the bandwidth of the panadapter, the way the software is presently written you can have theoretically 16 different software receivers. Only one is turned on, beside the main receiver. It would be impossible for someone to listen to 16 receivers and control 16 receivers without some kind of control circuit that is surgically implanted, or that is controlled by a neural network or some form of artificial intelligence. My origional idea for a skimmer like device was to have the 16 receivers scan little 5 khz slices of band width and identify each signal in its bandwidth and then create a cue of signals that were needed by the operator. Alex's approach to skimmer I think is a better concept than my addle minded approach, but still the AI thing has merrit.

You will note that a blue line has been added to panadapter 1. This line represents the watch receiver. I now have 3 active receivers in this radio. RX1, RX2 and Watch. If you look at VFO A you will note there is now a frequency associated with TX. That is the frequency where the watch receiver resides I can use this to work a split pile up while I am doing SO2R. If the DX is transmitting on 3504 and listening on 3507.354, if I hit the key, I will be calling where he is listening. I can also listen to what is on both of those frequencies by adjusting the various audio controls next to the multiRX switch.

In the setup menu I can also control muting.


The first choice mutes RX2 on TX and the second mutes RX1 when switched to work a station using RX2. If you are confused by the cacophony of all the signals in your headset, you can use these muting options to help you sort out things out as you go from one band to the other.

The MultiRX feature is available whether you are SO2R-ing or not. Here is how you tune this beast.

The tuning cursor is changed by right clicking in the panadapter, It toggles between no cursor, yellow cursor and red cursor. No cursor allows you to drag the panadapter with the mouse


In this mode I just left click the screen and the little hand grabs the screen and moves it and that tunes the radio. If I right click again I get the yellow cursor:


If I left click now, wherever the yellow cross hairs are centered, is where my VFO A will go. If I right click again I get the red cursor:



If I click now I will move the watch RX and also the TX freq up and down within the panadapter without changing the RX1 freq.

If I move the yellow cursor down on panadapter 2 while in yellow cursor mode I will tune RX2



It sounds way more complicated that it is. In fact after a short time its all second nature and you don't think about it. If you tried to do this with knobs you would hurt yourself!!!

So that's pretty much the deal. SO2R in a box, right out of the box. All I did was hook up 2 amps, 2 antennas, a mic and a key and some headphones. Stick that in your Sennheiser and smoke it!

See ya in the contests!!! (Actually I gave up contests for Lent)

73

VP8KF

This morning I am listening to VP8KF down in the Fauklands. I started listening before his sunrise, and listened while the terminator crossed his QTH. He is 166 degrees south from my QTH. I never do very well on the north-south path from my QTH. He is working a lot of JA's, a few W1's and not much else. The terminator is crossing the W1's about now. I get into the frey halfheartedly for a couple of calls as the terminator passes over his head. He peaks briefly and then heads down into the mud. My band noise is up a little this morning in the -111 dB range. I watch the JA's coming back to him on skimmer they look weak but they are perfectly Q5 on the screen. I can see all the dits and dahs, and I can make out the reports ..... -. -. or ..... ..... -. - ..- Skimmer uses intensity as a cue for signal strength so by studying the look of the signal over time you can tell if the station has key clicks or chirp and how far out of the noise the station is. You can tell if he is using a bug or a strait key as well. The JA's are a solid 5 dBm out of the noise or better. The noise looks amourphous and random in the background.

Every year some of my buddies enter a kind of low band competition. They try and work as many DX stations as they can from the beginning of Sept to the end of March. They usually get up around 170. One of my friends is pushing 180 this season on 80M. I tend to be more of a listener than a competitor so I don't enter, but it is interesting how many DX stations you can work on 80 in a season. You don't really think of 80 as a DX band. All you read about is how the sunspots have killed DX, when are the sunspots coming back. In November of 2004 I workded 100 countries on 80 using a Butternut HF-2V in just 3 weeks. Guess I didn't know the conventional wisdom.

I switch over to JA7NI on 160. He is much better copy. A lot of big signals are calling him but I guess he isn't hearing much because he is CQing right through them with an occasional QRZ? It would be so much fun to be able to tune an RX halfway in between W1 and JA and hear what that sounds like. Looks like he QRT's and I have to do the same

Yesterday I was looking at the crossfire probe antenna RX-4 square on the DXE site. I'm sure this would help. Its an interesting box

73

Full Duplex

There is a recent thread on EHAM regarding the Glories of QSK. The discussion revolved around what has come to be known as QSK in the ham community. Real QSK has to do with full duplex. It was from the days of separate receivers and transmitters. In the days of coastal stations and ships at sea, the coastal station had a transmitter plant, and a separate receiver site. The operator had access to the key, which controlled the transmitter, and the receiver. Since the RX and TX were active all the time you could hear between "dits" if you will.

Transceivers today are not full duplex. They share circuits like filters and the receiver shares the antenna with the transmitter. Thus the transmitter is switched on when a signal in the transceiver usually something called T+ is signaled to turn on by some external event like clicking the PTT or hitting the paddle. This turns off the receiver or at least parts of the receiver, turns on the transmitter and switches the antenna relay over to the transmitter. The receiver can not be used for receiving stations off the air when in transmit. When T+ turns off things go back to receive. I used to pick up this voltage in every so called "QSK" radio I ever owned to use to switch my amp since this was the earliest voltage to go positive. Ten Tec radios were designed like this and they were really the first to hawk this "feature" of so called QSK. It is not real QSK but simulated QSK since transceivers are half duplex. Like roofing filters you can advertise a concept into the general ham psyche and today we call this fast relay half duplex QSK. It's not QSK its just advertised as QSK. Like roofing filters we have been sold a load. That load makes us look like the smartest guy in the room so we continue to propagate the myth.

The Flex 5000 is different. It really can be made to do full duplex QSK like days of yesteryear if you have the second receiver installed. For true QSK you need 2 antennas. The Flex has a seperate RX2 antenna input. It also has a separate RX1 input and the ability to break the signal path of RX1 and insert something like a band pass filter or an amplifier.

Here are some shots of the various choices you can make in choosing antennas for 2 receivers and the transmitter


RX1 tap will tap RX2 to what ever antenna is connected to RX1. You can see the check mark box to open a the loop in the RX 1 signal path. The reason you have all this flexibility is you have 3 simultaneous processes going on in the flex at once i.e, full triplex RX1 RX2 and TX. If you choose you can set up the radio for full duplex operation by hitting ALT-D on the keyboard.

Here is my radio set up for full duplex


If you look at the VFO area you see both receivers and the transmitter tuned to 3501. RX1 and TX are set to Ant 1, and RX2 is set to RX2 ant input. On Ant 1 I have a full size 66ft 80M vertical. RX2 has my shorty 55ft self supporting 80M vertical that is on another part of the property. If I switch on the MOX I am in full duplex mode.


RX 2 is still active on 3501 and My TX is active on 3501. If I hit the key I transmit. No relays will click RF will exit antenna 1 and I will hear the RF on RX2's antenna

What you see on the panadapter is my transmitted signal received on RX2 If you look at the upper right Wattmeter you can see I am transmitting 97 watts. My transmitted signal reads +1 dBm on the RX 2 meter and you can see my CW spurs are down about -60dBm. Bob N4HY is working on an algorithm now that will improve this spurious response because the algorithm will automatically tune the signal for best spurious response in an ongoing fashion based on the frequency you are operating. It takes about 50ms to bring the transmitter into best alignment no matter where you are in the spectrum. The way the algorithm works now is the radio is aligned at once frequency in the center of the band. Spurious response on transmit is in the -100 dB range at that center frequency.

If I let up on the key I hear whatever is in RX2 NO RELAYS CLICK. The signal is always present.

Here is a shot of CO8LY transmitting on 3506.975 he is the guy on the green line which represents my 100hz filter on the panadapter


and here is a shot of me transmitting on 3501 while listening to CO8LY call CQ 6 khz up the band from my transmitter. I can copy him just fine even though I am transmitting 100W from a vertical that is 100 ft from a resonant vertical on the same band. Why I feel just like a coastal radio station. Hey wait a minute I live about 5 miles from the Atlantic!!! I am a coastal radio station!

CO8LY is perfectly Q5 and I copy his CQ and his call and I even hear a K9 calling him, through my string of dahs.

If you look here you will see a pile up on VP8KF on 3503 between me and CO8LY, the MOX is still on and I am still in full duplex QSK mode.


If I switch on the amp, I have 1.5 KW full QSK available. With the MOX turned on I run idle current on my amp and when I hit the key there are NO RELAYS that switch. 1500W flys out antenna 1 and I can hear it on antenna 2. I let up the key and still no relays, I hear on antenna 2. This is true QSK.

Here is a shot of the rig running SO2R


In this case RX1 and TX are on 80M. RX2 is on 40M I can tune RX 2 (in the left ear) while I am doing a run on 80M (in the right ear). If I hear something I want on 40 I click the TX button at the bottom of VFO B


and I am ready to transmit on 40M for the multiplier. Then I click back to 80 and continue the run. On 40 there is absolutely no indication you are transmitting on 80. No spurs or anything. The same is true of you are listening on 80 while transmitting on 40 or 160. I will include a picture of me transmitting 1.5kW on 80, while listening to 40 on a dual band vertical cut for 80 and 40.

So there you have it. I usually don't run full duplex since it requires 2 antennas, and my big vertical is a multiband antenna so I can flit around with ease working DX. I have a couple of rapid band change amps (Acom 2000 and ALpha 78 and LK-500NTC) that allow either for automatic band switching or one click switching. I also have the antennas set up so they can follow the band switch automatically. Pretty much point and shoot. Personally I run about 50ms of delay in my radio so I hear between words insteand of between dits, and my amps don't sound like machine guns. The AGC in this radio does not pump.

This is real QSK, the way QSK was done in the old days, which are now once again the new days. It is not the "advertised" QSK simulation that you see in QST adds. Adds with crap like "OOOO so smooth", while the relays in your amp sounds like a machine gun, and the AGC is pumping like someone pumping a hand water pump. Move over boys there is a new sheriff in town, and he's software defined.

73

Just my luck

I heard JD1BMM CQing his butt off on 160 this morning. No takers and I HAVE to leave BOOOO. I was exploring a group of blips on the panadapter around 1823 thinking to myself "looks like summer time conditions are here". My band noise was -116dBm so it wasn't real summery just yet, but I haven't heard much all weekend. There was a drift beacon and some weak warrbly watery carrier. As I got to the top of the blips I heard him plain as day. I went to spot him on the cluster but a JA had beat me to it

You never know when one of those blips might be very interesting

I was looking at oscilloscopes done in software. (The inevitable SDR tie in) Here is one with an arbitrary wave form generator


Who needs an OGEE curve anymore when you have one of these? I could probably use this sucker to make patterns for my router (as in wood, not as in data)

73

Discovery

I was out in the back yard surveying if I can get a NE-SW lobe on my array. I step inside and see the shuttle is ready to leave for orbit in 2 minutes. 3...2...1... I step out my back door and watched it go off. It's a perfect launch.

The southeast tree line lights up pink and orange when the candles are light, and in a few seconds I see the space ship start its climb over the tree line. A perfect white cloud exiting the rocket motors. As the shuttle lifts I see the gas cloud turn orange. I think to myself OH SHIT. The economy where I live is heavily dependent on space and if one of these goes down we all choke... for years. A lot of my friends are actual rocket scientists. I have friends in the control building right now. While I was in the operating room all day working on their family members, they have been out at the space port getting ready for just this moment. I have the best connection when I am at work. Husbands text their wives who are my nurses and I get the inside buzz. What I am seeing is the sun reflecting off the contrail and not some disaster in the making. It's dusk and the terminator is flying over my head towards Tampa.

The shuttle is probably 10 or 15 miles up by now. At exactly 63 seconds I start to hear the the roar of the rockets. It rattles the windows on my sunporch. (I'm 32 feet above sea level so you do the math.) That sound starts toward me from the moment the rockets are lit, and not from the shuttle overhead. I will continue to hear the noise from the shuttle overhead for about 6 more minutes. Pretty soon I see the solid boosters flame out and jettison. They sway away from the main path of flight at about 5 degrees one to the left and one to the right, and start their descent. Usually I loose them pretty quickly but tonight the sky over the Atlantic is clear probably for 100 miles and the contrast of the dusk sky to the east is perfect and I can see the white hot dots chute their way down to be picked up by their Navy tenders. I watch them fall nearly all the way to the tree line before I loose them

The shuttle continues. The solid rockets are orange and the fire tail is very long. The shuttle motors on the other hand are blue white and look like a star rising over the horizon and then the star starts to descend to the tree line as the shuttle gets out over the curve of the earth. I watch it until it disappears into the tree line, and walk back into the house at 6 minutes 40 seconds into the mission to hear the announcer say the shuttle is now 66 miles high and running 16,000 miles per hour. The shuttle is flying toward Spain, so it is a perfect marker for me to follow exactly where Spain is on my horizon. The shuttle's path is right in line with one of the possibilities for my array.

It never ceases to amaze me. The whole endeavor never ceases to amaze. Last week I watched a Delta II with the extra boosters night launch off toward Africa with the Kepler satellite on board.

Sandra Magnus is blogging from space. Me? I'm blogging from about 19.736 km from ground zero. The shuttle pads are to the east of the blue marker. Cape Canaveral Air Station is to the south of there. I think Barbara Eden still lives there.

If you look to the NW you will see the township of Turnbull. I live there. If you look real close at street view to the east side of the house you can see my 80M slef supporting vertical rising above the trees. It's the smudge that looks like a vertical. (Its actually on the back part of the property)

73

Total Daydream

I spent last night making a spread sheet to see how to optimize my vertical array. It is going to cover 160 80 and 40 and pinch hit on 30. The patterns are very predictable figure 8 and cartioid on the lower 2 bands and multi-lobed with gain on 40. I wanted to see how the spacing between elements affected the gain and pattern of the various bands. As you switch between different phasings, it turns out for example the gain goes up for 0 degrees as you move elements apart but the gain goes down at 90 degrees as you move elements apart, and the variations change for each band, so its an interesting exercise in optimization. My verticals will have multiple elements one for each band, and each element will be tuned to resonance at the base. I will chose elements using a coax switch to switch between the output of each vertical. As I played around with the phasing and patterns and mutual coupling between elements, I started thinking about W8JI's probe antenna arrays. I think its very interesting that a guy who has a lot of land in a very quiet area, and who had a bunch of the so called gold standard Beveredges basically has abandoned that antenna for an array of little bitty verticals that sits on a plot of land 70 x 330 ft It means there is hope for us mere mortals with a couple acres. I think Tom's research on this topic is outstanding.

I was reviewing W8JI's website regarding small vertical receiving arrays The feature of this array is the small probe style vertical antenna that is heavily loaded so there is basically no mutual coupling between elements. This allows for nice precise patterns to be acheived.

It occurs to me instead of a FET amplifier at the base of the antenna, and a phasing arrangement and miles and miles of CTV coax, you could put an SDR at the base of each antenna, and bring the signal from the base of each antenna all the way to a digital I/Q signal. If you had an array of 4 antennas therefore you would have 4 streams of sampled digital I/Q data. If you then sent this data back to your computer, in the computer you could build a virtual array and steer the array by adjusting the phase and amplitude of the various probe antenna signals in the computer, and then go ahead and process that signal through the regular SDR software. You could even have the computer optimize the RX and if you had 2 arrays well seperated you could diversity receive as well.

As I was reading Tom's article, he also speaks about his transmitting arrays sending re-transmitted noise into his receiving array. He says if you de-tune the transmitting array that helps to cure that problem. It occurs to me you could sample the output of the transmitting array into an SDR receiver and send that data back to the computer to cancel the noise.

The main problem with this is how do you time signature each bit of data. If you want to manipulate the data you would need to be able to manipulate all the arrays of data from the correct time frame. Maybe you could do something like TCP the data for reassembly back at the ranch (computer).

Like I said its just a day dream (yep I'm spending my Sunday in the OR) but it may be a way to make some really interesting high performance arrays. Some of the A/D converters are beginning to achieve a dynamic range and a frequency range where you could just input strait into the A/D For example:

something like this might be just the ticket. See this article

Yea I know, I need someone to unscrew my head from my neck pipe and check my wiring.

73