Thursday, May 30, 2013

Don't play Ingress at Radio Observatories

A few weeks ago, the IPG got some curious email from some ABQ-ians asking if they could play Ingress at the VLA to capture some GPS-based portals. If you've never heard of Ingress, think of it as geocaching with a Virtual Reality spin. Check out their website here.

Ingress is played on smart devices, which require data connections to operate. These data connections are fine and dandy unless you're at the world's greatest radio observatory; here they aren't so dandy.

RF-EMS

Below is a screencap of our RF-EMS (Radio Frequency-Environmental Monitoring System) which captured two WiFi access points (the darker blotches) from an RV containing a Verizon 4G hotspot and another router for something else.
Your VLA on WIFI

In the last blog I described the 10' dish for pinpointing RFI. We also have a (usually) 24/7 monitor that uses some pretty nifty antennas and preamps on a 50' tower, sending it to an HP 70000 Spectrum Analyzer in a RF-shielded room from which we can record and upload plots like the one above, every day for the past 5+ years. 

RF-EMS Tower and Bunker

The biggest downfall is adequate locating of interfering transmitters. Currently, I'm designing a method which will allow the IPG to quickly and accurately pinpoint people with any kind of transmitter, be it a cell phone, hotspot, or vehicle keyfob (if we wanted to locate such things). My idea is based on multilateriation, which uses multiple receivers around the site which compare arrival times to calculate a four dimensional location. Keeping the bill of materials as low as possible, simplicity, ease-of-use and network integration (without causing RFI itself) a prime focus.

It may be overkill, but it gives me something to do in the free time.

Other Doin's: Testing out and Debugging the 74 MHz System

When I'm not having free time, this is what I'm doing. A new feature of the Expanded-VLA is observations on the 4 meter band. The current system in place uses these simple crossed dipoles hoisted a few meters below the sub-reflector.






















Installing the 74 MHz Crossed Dipoles

The cross dipoles connect to our receiver, which hooks up to the rack that magically digitizes the signal and turns it into pulses of light which the correlator feeds upon.

One of the problems we face are things broken that don't have to do with our antennas and receivers. For example, the first test we do to examine the receivers performance is a band pass plot. Often times, we see something like this:
A bad bandpass plot caused by a faulty relay in the T301.
This is ugly! What we want to see is this:

A beautiful bandpass! You can see 4 band on the left, and P-band in the middle with RFI spikes all over.



First we go digging in the LO-IF and FE racks for a place to stick a spectrum analyzer to...



Eric the BAMF next to the LOIF and FE rack. Our culprit is on the left, in the middle of the top rack of modules
And from that we figure its' this T301 which does the first IF up-conversion from 0-1GHz to 1-2GHz.













The faulty module in question

We get a new one, stick it in, turn it on and voila, it's alive!




    Tuesday, May 28, 2013

    (A)Typical LBR installation

    Like I mentioned, today Paul Harden, da boss, and I installed two more low-band receivers today, but with only a 50% success rate.

    The receiver is  dual banded for 74MHz/350MHz and is housed in a big and heavy 8"x12" metal box. for thermal stability. We call them LBRs or 4/P receivers -- 4 band is 58-80MHz (eg 4 meter wavelength) and P-band is 325-420 MHz or so. P comes from it's old military designation, like L, S, C, X, Ka, K, Ku, radar bands, etc.

    She's mounted on a sturdy mount inside the "barrel" also known as the Focus Rotation Module, or simply the apex. It's also my second home. Inside the barrel is a central square steel through which the 4-band antenna and receiver output cables go through. The receiver mounts near the bottom of the apex, and plugs in to 8 different cables - two pairs of Heliax for the 4/p band crossed dipoles, one Heliax pair for the output, a power cable, and a calibration signal cable.

    Each antenna input requires two pair because of the dipole design -- they're orthogonal crossed dipoles that are fed into the receiver as two linear polarizations. Some magic happens in the LO-IF rack that either sends them as a linear polarization to the correlator (mainly for ionospheric observations) or combines them in quadrature to provide left- and right-circular polarization for observations outside our own solar system.
    The first installation went smoothly -- we removed a malfunctioning receiver from antenna 18 and replaced it with a fresh one -- it's pretty much plug and play. Somewhere through, we lost a two-way radio. No idea where it went. I think it always existed in some kind of quantum state, and now it's hiding.

    The second was troublesome. Each output, left and right polarization (or X and Y) should show clearly the band passes of each band through a spectrum analyzer. What we saw was spikes caused by the 74MHz *something* oscillating like a runaway Hartley. (that was a pretty poor play on words)

    So we hit it a few times, jiggled the cables, and poof, it was back to normal! We came back down to the vertex room to find it again went nuts. So we went back up, wiggled some more, and took the SA with us. Now the X side was gone! We took the cover off and poked around, but it was dead. Methinks the oscillation was strong enough to saturate an LNA, bust a cap or diode or two, and kill X.

    Paul's gonna play with it tomorrow, and we'll see if we can make our pseudo-deadline of getting everything working by Tuesday night. Lots of pressure from the higher ups.

    Tonight, I'mma gon' write a blurb about the SDR demonstration at the ARRL Youth Lounge at Dayton for Ward Silver, N0AX, upvote and argue things on reddit, and buy some wire and ladder line for field day. Need to make a 9:1 balun by then too.


    View of South baldy and the Magdalena Ridge from the bus ride home

    What I'm up to at the VLA

    Back last November I was offered this internship to work for up to 8 months at the Very Large Array, as you see pictured behind this post. The VLA is the worlds most prolific radio observatory, having the most citations in of any radio observatory in all of science. I could not turn this offer down/

    The VLA is a well oiled machine of 27 cassegrain feed parabolic reflectors each 25 meters, or 82 feet wide, that send concentrated RF to massive feedhorns at its vertex. One VLA dish covers all frequencies from 1-50 GHz, and has two extra bands at 74 and 350 MHz using extra antennas. Together, they can make a dish that theoretically measures over 20 miles in diameter!

    Apex of a Dish Showing the Cassegrain Subreflector, 350MHz cross-dipole at the center of it, and the new strut-straddling sleeve-dipoles for 74MHz around the edges being installed
    My primary job at the VLA is to learn. Everything RF exists here, so I'm soaking up as much as I can. I've learned more about microwave RF design, Radar, synthetic aperture synthesis, correlation, radio astronomy, antenna design, transmission lines, test equipment, RFI, and much more that I have ever (and perhaps will have ever) at college.

    Learning is a sort of meta-job. What I really do is two-fold.

    The first fold is Interference Protection. The IPG specializes in the detection, location, analysis, and mitigation of radio frequency interference that has the potential of ruining and/or corrupting observations of the radio sky. The VLA is located in a lake bed, 20 miles away from any town, 90 miles from Albuquerque, surrounded by 360° of mountains that buffer the observatory from radar, wifi, cellular, aircraft, and other terrestrial sources of RFI. Satellites are also a source of RFI, so they must be documented and their transmissions well understood so that the VLA Correlator  can learn how to discern orbiting transmitters from galactic transmitters.

    On IPG, we've done a few RFI Site Surveys at places like the Magdalena Research Observatory on South Baldy and at the Pie Town VLBA site. Unfortunately, a Verizon 4G LTE cell tower exists on a hill only 5 miles from the center of the array. The signal it produces isn't bad -- it's the amount of visitors who show up with a full signal, assuming it's okay to use their smartphones to upload photos, videos, sprout WiFi APs, and cause all kinds of problems. That's when we break out the CELL PHONE DESTROYER 6000.

    THE CELL PHONE DESTROYER 6000 is nothing more than a 1.8-2.4 GHz feedhorn attached to a spectrum analyzer, which is directly integrated into the neural cortex of this RFI Seeking Unit known only as Mert.
     For quieter RFI, like satellites and terrestrial transmitters, we use a big-ugly 10-foot dish with a wideband conical-log spiral feedhorn to determine bearings to RFI: 




    Along with intentional radiators, unintentional  transmitters like microprocessors, screens, TVs, and other electronics produce noise that can be detrimental to the RF environment. Such electronics need to be tested in the Reverberation Chamber, and shielded if necessary.

    The second fold is Front End -- Front end of receivers that is. On the FE group, I've been tasked of removing old low-band receivers (74 and 350MHz) and installing new consolidated ones along with about 200 feet of heliax and control cables for them. I've also been building and improving antenna designs for the LBRs, including the new strut straddling dipoles pictured above. I've come across a cool 74MHz widebanded antenna design that may or may not be patentable, so we'll see from it's creator if I can get acknowledged in a paper or something :-)

    The LBRs are located in the apex, also pictured above. It's fun becoming a grease monkey while getting a view of the VLA few have seen.


    Such a view

    So how did I get this job? Ham radio of course! My boss actually e-mailed me after an unsuccessful run through career fairs and internship searches for this summer. He noted my experience with the Missouri S&T ARC, W0EEE, as a big kicker on my resume. At S&T I've done a lot of work mitigating RFI, fixing repeaters, learning about radio, and having fun with it. He himself is a ham, albeit inactive, but still understands the value a license can have at such a job.

    Ham radio gets jobs! (Girls? No.)

    Today Paul Harden (NA5N) and I will be installing a few new LBRs in antennas 18 and 20. The weather looks beautiful for it too! Check out Paul's personal website, chock full of receiver data, photos, and the history of the area.

    That's about the gist of it. 73 for now!

    A Zero in Five Land - Welcome

    With midnight inspiration comes morning desperation. For sleep. Heh heh.

    Welcome to the new blog of N0SSC -- ARRL Youth Editor, EE degree seeker, Very Large Array intern, and among other things, a radio amateur. Ideas have been flowing from my mind into thin air, so I have done up this blog to start to capture them.

    I'm a zero in five land. My call sign - N0SSC, shows that I am from zero-land: the ham radio call district that makes up most of the midwest. Due to my internship, i've been displaced to New Mexico, which is in the fifth call district - 5 land.

    So, this blog will focus on my ham radio activities while in New Mexico, as well as my internship at the Very Large Array. I have a lot of fun stories to share!

    Stay tuned for more! For now, follow me on your favorite social networking site -- just google N0SSC and you'll find me.