December 2012 - PSDR - mailman.hamwan.org

[PSDR] Rubidium Frequency Standard
by Bart Kus 29 Dec '12

29 Dec '12

Hello, I've finally become successful in my long quest to have a working frequency standard. Here's the all-integrated unit I just finished putting together: http://imgur.com/a/IzROb Left LED = power Mid LED = phase lock indicator Right LED = unused. Was supposed to be PPS blinker, but this unit doesn't seem to output PPS :/ It puts out 10MHz @ about 10dBm. Uses about 18W when running, and gets quite hot! Since I couldn't find a decent high current 15V regulator @ Fry's, I just used a high gain (hfe min = 500) 80W/4A NPN and a 16V zener reference with a 20mA quiescent current. The regulated ~15.5V feeds the secondary 5V regulator, which is a normal 7805. The lock LED was a little tricky. The lock output from the unit is active-low (GND), but in the unlocked state it hovers around 2.5V output. I had to use a 22k pull-up (to +5V) resistor to eliminate glow from the indicator LED, while not taking too much current out of that weak pin when it's active (low). You can do fancier things with a pair of transistors, but given this is all rats nest wiring, that's too complicated. A single resistor works just fine! The mounting screws may seem overkill, but they serve the purpose of heat-sinking the rubidium unit to the aluminum chassis. The +15V supply output during the high-current warm-up period is not pretty. Quite the ripple is visible, but it does not seem to affect the unit at all. Once lock kicks in the +15V bus current draw drops off quite a bit. 60Hz harmonics are not a problem in the output, measuring about -73dBc. It's not the best, but it's good enough for my purposes! Using this I can now calibrate all the signal sources and spectrum analyzers for frequency accuracy! Yaaaaay! :) --Bart

1 0

[PSDR] We now have a properly characterized reference antenna!
by Bart Kus 12 Dec '12

12 Dec '12

Hello, I was able to finish the software and experimental work required to fully characterize the errors of the reference antenna! Full details, including code, new graphs and exact reference antenna gain characteristics here: https://www.hamwan.org/t/tiki-index.php?page=Antenna+Analysis&structure=Ham… What this means is we have a properly calibrated field strength emission system and any antennas dropped into it and measured will result in ONLY their characteristics showing up in the readings! I've settled on the use of a high directivity antenna to minimize and ignore the impact of multipath effects. Now the next step is to drop the candidate antennas into the test setup and compare their behavior with respect to frequency. After that, I need to get busy on building the rotor rig to be able to measure radiation patterns, not just frequency response. Ever onward, --Bart

1 0

[PSDR] Reference antenna problems
by Bart Kus 11 Dec '12

11 Dec '12

Hello, So I tried to measure the reference antennas (simple dipoles). It's kind of a disaster. With the antennas only 2ft apart: The received signal is pretty weak: The marker is down 20dB from the previous measurement: I guess that pair of sectors I was using were really helping things out. I'm also concerned about the uneven response, but that could be multipath given the really low signal levels. Does anyone have any experience with antenna measurement like this? I'm thinking my next step is using the 21dBi parabolic dishes as the reference antennas instead. Having thought about it, the high directional gain, although it would need precise alignment to make accurate measurements, would help with ignoring/overcoming multipath, no? Some of those peaks and valleys might disappear. A 20dB range over the spectrum is pretty ridiculous. --Bart

1 1

[PSDR] New test equipment has arrived!
by Bart Kus 11 Dec '12

11 Dec '12

The device on top is the HP 8672A and it seems to be fully functional. It can generate signals from 2GHz - 18.5GHz, with power levels from -120dBm to +10dBm, and with AM or FM applied. It's fully GPIB controllable. Here you can see it sending a 0dBm signal @ 18GHz into the top spectrum analyzer. Anyone up for a 10GHz QSO? ;) Speaking of the top specan, the face is now reconstructed and working well! The main dial knob is not working yet, needs a #683 light bulb replaced. A pack of 10 is in the mail. The display section of the top specan is also buggered up. Any RES BW < 3kHz has severe attenuation. I'll have to inspect the filters in there and figure it out. The bottom specan's only remaining problem is a dim/fuzzy display. There's apparently a fix for this. Last night I also thought the high range (2-22GHz) of the bottom specan was screwed up since it was showing signal levels 25dB down from what the calibrated power sensors were reporting. Turns out nothing was broken, just the YIG tuned mixer was misaligned and needed to be corrected (there's a digital correction table inside the instrument). When I pulled the memory board out of the bottom unit to help troubleshoot the top unit I cleared all that calibration data. All of these need to be calibrated/performance tuned still. I have a replacement atomic frequency reference on the way since the one that arrived last week is broken. Took 15 minutes of heating time before it achieved signal lock. And then it only locked for about 60 seconds, never to return. The other parts missing for calibration are extension boards (custom bus extender PCBs for these model specans) and an active probe. I think I have everything else to perform a good cal. Ever onward, --Bart

1 0

[PSDR] Antenna gain vs. frequency measurement
by Bart Kus 10 Dec '12

10 Dec '12

Hello, After a week of fixing equipment (still ongoing) I've finally been able to make some solid antenna measurements! It's still very much a work in progress and will likely take all of this week as well, but it's being documented on this brand new wiki page: https://www.hamwan.org/t/tiki-index.php?page=Antenna+Analysis&structure=Ham… Here's a delicious graph from that page: Image If you'd like to know how to interpret that, you'll have to click the link. :) Here are a couple pictures of the incredibly messy test setup in action. The ARC Wireless TX antenna: The frequency tripler (top) and amplifier (bottom): The 8660C sweeping signal source: The Laird RX antenna (bottom left) and the 8566A taking the measurement: The 2 specans that needed fixing (1 in progress still): The facial reconstructive surgery work for the top specan: Ever onward, --Bart

1 0

[PSDR] November HamWAN update
by Bart Kus 03 Dec '12

03 Dec '12

Hello, A quick & dirty update for everyone. 1. HamWAN Labs has undergone a major upgrade in terms of microwave equipment capability. New equipment added to support testing: 1. HP 8566A 100Hz-22GHz spectrum analyzer with 10Hz RBW and -130dBm DANL! This unit has a broken 1st stage converter though, which drops its output level by 50dB when the converter gets warm. I'm working on replacing this converter, but for the time being I've externalized it so it's kept cooler than the inside of the analyzer. This keeps it working nicely. Pictures of externalization, and specan bottom: https://lh4.googleusercontent.com/-xz21Of-xjJw/UL1A_kUNztI/AAAAAAAAAdk/gW_f… 2. HP 8672A 2-18GHz signal generator. Still waiting on shipping, but this will provide very steady test signals so narrow band testing can be done. The present YIG oscillator setup has a 4MHz bandwidth and I cannot stabilize it without throwing it into a PLL. Plenty of other benefits to the HP generator here, such as GPIB control, built in attenuator, and AM/FM modulation capability. Picture of signal generator: http://www.surplus-electronics.co.uk/urchin/assets/images/_mod/20111109_152… 3. A 4-axis CNC milling machine! With adjacent channel interference being an issue, and the cost of microwave cavity filters astronomical, I believe this is the best/cheapest way to do channel filtering, by home-milling microwave cavity filters. It's also useful for PCB fabrication. Pictured below. 4th axis not shown. 4. Lots of various mill/drill bits for the CNC mill, as well as a light tachometer. 5. Lots of new SMA interconnects/adapters to make working @ microwave easier. 6. Boonton 2500 DC Calibrator. Finally found this authentic unit for a reasonable price. This puts an end to the HamWAN DC Calibrator project, but I'm leaving the pages up since they contain lots of good tolerance analysis work. May be useful later. 7. Boonton 25A 1MHz RF calibrator. It was really tricky calibrating the thermocouple sensor (4200-8E) using this since it needs to be cal'd @ 50MHz. This took a few days. But the RF calibrator is a very precise unit, and it calibrates diode sensors (such as the 4200-4E) beautifully. I was able to use this RF calibrator as a standard and then reflect the power levels up to a couple different RF sources @ 50MHz to do the actual thermocouple calibration. The +20dBm power level range calibration is only approximate as a result of my process, but it's good enough and the best I can do. Who would have guessed you have to account for LMR400 losses @ 50MHz over 4ft? Well, it's true! There's a measurable loss. Here's the unit: 8. 10MHz Rubidium Frequency Standard. Sadly this unit is not working. Trying to get a replacement mailed out. It will allow for very precise frequency measurements and can be used to align microwave gear. Here's a picture of the stock unit, and my attempt at getting it to work by cooling it: 9. A linear 1.3-10GHz 0.6-10dB 1W attenuator. 2. GPIB control has been established with all 3 important instruments: the power meter, the frequency counter and the constant current source. Faced some challenges with getting the Boonton 4200 and the EIP 451 to do GPIB, but they were overcome. Always remember to configure your terminating characters properly! With all 3 instruments being controllable now, the next step is automating sweeps. The YIG oscillator is not very leveled across its spectrum so I'll have to subtract an error curve. The spectrum analyzer supports this operation internally, but I have not yet figured out GPIB control of that unit so that I could extract the data/graphs from it. 3. Cougar site acquisition has stalled, and I'm not pushing it. I want to be damn sure that the sector antennas can co-exist without interfering before throwing anything up there. Deployment can be expensive. Here are the guts of the most promising antenna: 4. I'm on a HamWAN vacation all week to be able to finish up these tasks. Couldn't ask for a better time. :) 5. All HamWAN Labs equipment is purchased out of my own pocket and is not considered HamWAN assets. 6. HamWAN balance is $350 with only one recurring donor. Technically, HamWAN can't throw up even one site should Cougar become available. While I don't mind spending money on HamWAN Labs stuff, I'm gonna be pretty strict about HamWAN network gear itself being funded out of not-my-pocket. It doesn't scale. You can paypal money to donate(a)hamwan.org or use the website. I've also got a "PayPal Here" credit card scanner on the way so the next time I present on HamWAN subjects, I should be able to take donations right on the spot via my phone + the card reader. 7. Managed to hack a little bit on the rotor for antenna pattern testing. It's a synchronous motor in there, and I got it to run @ 10Hz without saturating the core. :) With this slow rotation, angle measurement can be done based on time instead of more trickly angle encoders. 8. HamWAN almost has $2mil site liability insurance. Benjamin is working on the details with the ARRL. This should cover all the sites we deploy onto under 1 policy. Ever onward! --Bart

1 0