Introduction
It shouldn't come as a surprise that being interested in radio, electronics and astronomy that I should also be interested in radio astronomy. In fact, I have been interested for a very long time. But other projects got in the way, and RA got shelved. I am now working on my newest optical telescope, the Stargate-Next Generation (SG-NG, for short), and I want it to be more than just a telescope; it will be a scientific instrument. Two things that I was considering were a 2800 MHz detector to measure solar flux and a VLF receiver to monitor for storm activity (This is an all-metal telescope, and I'm particularly sensitive to the thought of lightning. Shazam!). I have no experience with electronics in the microwave range, so I focused on the VLF receiver. Using Joseph Carr's book, Secrets of RF Circuit Design, as a reference, I realized that I could also do RA work. Well, as someone prone to Creaping Feature Syndrome (CFS), I wanted more! I went to the Internet looking for ideas, starting with the Society of Amateur Radio Astronomers (SARA). I dusted off two books that I got from Radio Astronomy Supplies (RAS) years ago, and while dating back to the 90's, they had plenty to start with. I was rather disappointed that both are no longer offered by RAS, that RAS now mainly caters to deep-pocketed government institutions and universities. Just like then, the small experimenter is on their own and dependent on their own skill; only the technology has changed. I found another RA web site by some French amateurs, which offered some great ideas to start with. After reading the site several times, I realized that a 1420 MHz hydrogen line receiver was possibly within my grasp. Now. Right now, I'm just looking for a simple approach to start with, then build upon.Quick and Dirty
The French web site started with a satellite finder- essentially a microwave-band crystal radio. All the parts mentioned were doable by my shaky hands. I'd found an excellent and seemingly simple-to-use LNA MMIC by Mini-Circuits. Great! The excitement wore off when I saw that it was packed in a SOT-343 SMT package. VERY TINY. For now, I may buy an LNA. But my later designs will require that I build my own. With this setback, I came up with the QuAD approach- quick and dirty. I didn't even want to try building a superhet receiver at 1420 MHz. Not yet. So, the TRF approach looks so appealing. My goal is to integrate a QuAD receiver and a high-gain helical or Quagi antenna with the SG-NG. A basic design starts with a string of MAR-6 MMICs, a cavity filter made for 1420 MHz (that I can build for a tiny part of the cost of one from RAS), and a microwave diode detector. Maybe a diplexer tuned to deal with broadcast radio interference. A thermoelectrically cooled LNA. A tuneable receiver (i.e., a superhet). First, I need to make the SG-NG a working telescope. May 19, 2013
I have been researching ideas, mainly to implement a 1420 MHz QuAD. A list of microwave components is being compiled, so I will have a stockpile to experiment with when ready. One interesting point to note: the Pendicton solar flux receiver (at the Dominion Radio Astronomy Observatory in Canada) is a TRF design, and not a superhet (See the February 2013 issue of QST, pages 39-45.).September 21, 2013
Whilst on a hunt for a 23 cm band transverter, I discovered that Mini-Circuits makes a conveniently packaged LNA for 1420 MHz offering 10 dB gain and 0.5-0.6 dB noise figure (ZX60-P103LN+). They also make a detector module, but for now, I will stay with a simple diode detector. While not as sensitive, the diode detector doesn't come with a 100 mA current requirement. Ouch!September 29, 2013
I went back to Mini-Circuits, looking for a LNA for another project, and found the ZX60-P162LN+. While it has a slightly higher noise figure (NF), it also offers a gain of nearly 20 dB. Plus, it is both cheaper and uses less power. Also, it has a much more restricted bandwidth (700~1600 MHz) vs the 50~3000 MHz of the ZX60-P103LN+. But in this application, performance comes first. I plugged numbers into the Friis formula, and the higher gain swamped out the higher NF; the total NF of both LNAs plus a chain of MAR-6+ MMICs was virtually the same. Sold! While the MAR-6+ is convenient, I'm less than thrilled with the 3dB NF (Given only at 500 MHz. I'm guessing the NF at 1420 MHz is closer to 4 dB). I could not find a lower NF MMIC, other than the RA-6+. And that one costs four times of what the MAR-6+ does, so, for now, I'll stick with the MAR-6+.October 1, 2013
I came up with an alternative to the tuned cavity filter- a quarter-wavelength stub, cut to 1420 MHz.January 24, 2015
I have obtained enough parts to build a barebones prototype: a 15-element Quagi, ZX60-P162LN+ LNA, VHF-1320 HP filter and a ZX47-60+ power detector. I added the HP filter to get rid of cellular, broadcast and commercial radio noise. I decided on the power detector, while being a battery hog, it gives me more sensitivity and will probably need less gain somewhere else to offset the 100 mA cost. Now when I finish this and test it with the sun as a source and it works suitably, I will build the MAR-6 stages and tuned cavity filter. Also, I have made a stub antenna for the 40 kHZ VLF receiver. Winding nearly 500 feet of #30 enamel wire onto a length of half-inch PVC pipe is an exercise in patience.June 5, 2015
I have redesigned my hydrogen line cavity filter. Both would be machined from solid copper bar stock, but the second has both imput/output SMA connectors on the end, rather than the sides (awkward). Also, a parts list had been compiled, and a PC board and brass box has been designed and modelled in AutoCAD. Soon, I will port the design into Mastercam so that it can be made on a CNC machine.March 27, 2016
To say the least, a lot has happened since my last entry. First, I switched from a Quagi antenna, to a helical design. The helical being both more forgiving of varying incoming wave polarization and mechanical tolerance. It was simple to build, but time-consuming. A 16-turn helical required 31 teflon spacers, mounted on a piece of 3/4" PVC pipe, and held in place with nylon screws. Despite all of the holes in the pipe, it seems relatively stiff. Each spacer had an angled hole in it to accomodate 1/8" copper tubing. All that remains is to obtain and mount the female SMA connector.The brass box for the MAR-6+ MMICs has been redesigned, as was the cavity filter. I have done all of this in SolidWorks, this time, and most of the MasterCAM work has been done. The LNA booster box and cavity have been 3D-printed, and look good. The approval process to get to cut these on my school's CNCs is slow. I still need to do a test cut with plastic first.And another speed bump has been lobbed my way. I found that my understanding of the resonator length was wrong. I thought that the length was determined by 1/4 the wavelength times an indeterminant Vf for copper. Vf should be determined by the dielectric constant of the cavity, which is air. At the same time, a new debate was ignited over fringe capacitance between the resonator and end cap. I now have three different capacitance models, but none account for the wildcard in that the two effective plates of the capacitor are asymmetrical. Before, I had intended for the filter to be no-tune. But now I can't predict how much capacitance is added at the end. So, I have to cut it perhaps 5% short, and add a tuning screw and disk.
