Tuesday, 18 October 2016

AF amplifier plus line out -- Jupiter Modular Receiver

SECTION 6  —  AF amplifier plus line out

Click for the Master Index to this project

It's fun and useful to listen to your signals from Jovian reception adventures. Along with electromagnetic Jovian bursts, you may also hear bad noise like RFI or hum which might not glare apparent on your monitor.

I designed and built an AF preamp with a buffer, splitter, active AF gain control, a tone control and finally, a discrete component PA for speaker drive. The enclosure also contains 4 extra DC ports to provide DC voltage to other modules via short, shielded RCA cables.
Josh M pointed out that  I really have to act carefully to avoid mixing up my AC and DC ports!

Above — Input to tone control output.  This PA goes in a metal box since the last thing we want is RF from the AM BCB band getting detected and amplified. The input 39R and 0.22 µF capacitor low-pass filters all signals above 18.6 KHz including offensive RF.

To provide capability to drive a sound card with an isolation transformer, I split the input with 1 path going to the AF preamp and the other to an output RCA port.

1/2 of a TLO82 op-amp generates the virtual ground and it's ground referenced to a single point along with each of the 2 AF boards. Yay! No hum was detected in this module.

Many of the ideas for this module came from Douglas Self.  I refer to his book Small Signal Audio Design: 2nd edition whenever I fancy making some AF circuitry.

The active volume control gives very low noise at lower volume setting levels. R1 and R2 set the maximal gain and can be manipulated with the standard formula: Vout /R2 = -Vin/R1.

Speakers, cabinets and rooms color sound. I feel a tone control is necessary to allow pleasing sound in whatever speaker you ply in your listening room. This is essentially a 1 control Baxandall circuit developed by Douglas Self and modified for single supply.

Above — PA stage with its own virtual ground to use up the other half of the 5532 op-amp. Giving the PA board its own virtual ground made bench testing this board easy.

I've written about this circuit before and developed it for Regen #5. Click and scroll to read about it.

In order to get a power amp output signal swinging as close to the DC rails as possible ( between 0 and 12.3 VDC ), you must depart from a complimentary pair of emitter followers. In these, the collector-to-emitter saturation voltage plus VBE keep you from getting close to the rails. Thus, I've adopted the topology shown -- a common-emitter configuration for the output pair.

Above — a 12-bit  FFT of my PA into a 8 Ω resistor. Drive is here set to give 877 mW output power and each vertical division = 10 dB.  877 mW sounds loud and all harmonics lie under 60 dB down.


Click for Module 7, the logarithmic amplifier 
Click to return to the main index for this project

Local Oscillator -- Jupiter Modular Receiver

SECTION 4  —  Local Oscillator

Click for the Master Index to this project

This year, I built and designed 3 local oscillators, however, this VXO leads in simplicity. Click for my synthesizer for Jupiter receivers.

Above — Entire schematic for the VXO.

2 standard value crystals in "super VXO" format provide a delta F of 135 KHz; just enough to steer around any interference on or around 20.1 MHz. Stability tests showed average frequency variation at room temp is 1/10 of 1 hertz. No need to ovenize crystal oscillators for Jupiter reception.

You may wish to increase the delta F by increasing the inductor size and/or employing a variable capacitor with a wider delta C. It's important to test the output in a 'scope and counter to ensure reasonable frequency stability and that it remains oscillating across the swing of the variable capacitor.

For the main oscillator, I employed the medium power BC337 transistor -- a device with low flicker noise. Mine are original Fairchild parts [ now ON Semiconductor ] as terrible bootleg copies are widely sold now.

Vojtěch Janásek keeps an interesting thread on low noise AF transistors including the BC337. Click for his pdf page link.

Following the Colpitts oscillator, an emitter follower's output port provides a good input match for a common base RF amp. The CB amp provides strong reverse isolation between the oscillator and its load. Again, another emitter follower transforms the high output Z of the common base amp to a low Z which gets low pass filtered and buffered with a 4 dB pad.

 Above — Output of the VXO in a DSO.

Above — Simulation of the low-pass output filter in GPLA from EMRFD.

This VXO looks pretty unremarkable -- but adopts careful RF bypass of DC power lines in every transistor amp and runs just enough current to minimize distortion as the signal moves through each RF amp.

The enclosure looks ugly -- it's a re-purposed box from some old project. I threaded in a bolt to block the hole seen on the right side in the reverse view photograph.

Click for Section 5
Click for the Master Index for this project

Monday, 17 October 2016

Input filter and preamplifier -- Jupiter Modular Receiver

SECTION 3  —  Input filter and preamplifier

Click for the Master Index to this project

Above — Build of the input and RF preamplifier stage. I put in a 0 or 6 dB attenuator right at the input. That's double-sided Cu+ board with copper vias joining the 2 surfaces intermittently and also at crucial RF ground points.

 Above —Schematic with some data.

Above — SA + TG sweep of just the input filter board [ low-pass FL, plus triple-tuned band-pass FL ].  I ran powdered iron toroids  [ T44-6 & T50-6 ] wound with 21 gauge wire. The trimmer caps were SMT parts with a measured Qul of 1500 @ 1 MHz.

Standard band-pass filters filter steeper and deeper in the high-pass skirt, so I added the input low-pass filter to bolster the low-pass transfer function. It worked.

Above — After breadboarding the preamp, I put a connector on the input and output and measured some parameters.

I ran a RFMD [ now Quorvo ] SiGe HBT MMIC amplifier; a 5 GHz part with a decent OIP3 and that well behaves with respect to parasitic oscillations. Still, then, with such a high Ft, it's suicide unless you adopt a good layout, bypass into UHF, and run 2-sided copper clad board with via wires near the RF ground terminals etc.

With the 1 nF series input and output caps, I measured gain from 1.459 to 50 MHz in this sweep. The 1 nF caps decrease the gain at 20.1 MHz and also at the top end of the broadcast AM band. The gain at 1459 KHz = only 7.59 dB. Some of these high Ft MMICs exhibit massive gain down low and this can trash IMD from AM BCB interference. With the 1 nF coupling caps and the front-end filters, I don't have to worry about the 10 KW AM transmitter located only a few Km from here.

Above — Sweep of the MMIC from 64 - 640 MHz

Above — Sweep of the MMIC from 0.65 - 3 GHz. I could find no unwanted parasitics with bench testing. Satisfied, I hooked up the preamp to the filter and enclosure mounted it.

Above — Sweep of the whole RF filter amplifier from port to port after re-tuning. Happily, the Hammond enclosure didn't add any cavity effects to trigger unwanted oscillations.
I added the 3 dB output pad to further throw away some gain and ensure a wide band 50 Ω impedance output into the mixer of the main receiver. Final S21 = 9.56 dB. The output 3 dB pad = 294.1 Ω 1% tolerance size 0805 resistors from a reel sent to me free by a kind reader + an 18 Ω 5% tolerance resistor I hand picked after measuring 17.7 Ω. The input switch attenuates the signal by 6.2 dB measured.

My goal was 10 dB gain with a stable, wide band match on the input and output. I did not go with the common gate JFET as it's difficult to get a decent output return loss ( > 20 dB ) across a wide bandwidth into 50 Ω; plus I wanted a better OIP3. Still, too, the JFET proves a good low current choice for a Jupiter receiver preamp and its output Z is well suited for easily matching a NE612 Gilbert cell mixer input Z.  I also purposely avoided a choke on the MMIC to keep the gain under ~10 dB.

Above — Return loss sweeps of the output port of the RF preamp/filter. Outstanding S22 over a wide band.

Above — Another photo of the entire stage. Built with Ugly Construction, plus hand carved pads of the correct width [ 50 Ω impedance] for the MMIC.

Click for Section 4
Click for the Master Index for this project

Antenna -- Jupiter Modular Receiver

SECTION 2  —  Antenna

Click for the Master Index to this project

Like for amateur radio and SWL DXing, the antenna ranks as somewhere between important and crucial in amateur radio astronomy.

The current most recommended antenna for monitoring Jupiter and perhaps the galactic background = a pair of dipoles with a power combiner plus a delayed feed line to correctly phase the antennas. The phasing line length establishes the proper impedance match plus obtains the desired radiation pattern. This gives ~~ 3 dB directional gain over a plain old dipole at the same height.

Using a registered version of Radio Jupiter Pro, a program written by Jim Sky  I learned that for 2017; at my latitude, a pair of east-west dipoles up 6 meters with a 135 degree phasing line in the south dipole's feed will give the greatest gain to receive Jovian storms. Jupiter’s peak elevation is ~ 35 degrees above the horizon and this set-up maximizes directional gain to match Jupiter's path.

Big problem. My lot goes North to South and poses some challenges to this ideal antenna. On August 14, 2016 I erected a single, east-west oriented half-wave dipole up ~4.9 meters high.  To change the beam pattern, my only option is to raise the antenna up or down on my 2 anchor poles.

I tuned it to a center frequency of 20.150 MHz with a tracking generator/spectrum analyzer plus a return-loss bridge. When I connected it to my RF filter/preamp, main receiver, and AF amplifier with speaker output, the galactic background was strong, although local noise was much decreased compared to the 40M band 1/4 wave vertical antenna that was in the exact spot.

I built 2 wooden supports from 2X4 lumber and some plywood. They rise 1.22m ( 4 feet ) and hold a 4.37m (14 feet) hemlock pole inserted 0.6m ( 1.5 feet ) down inside each base. Thus the top of my poles lies at 5 meters height ( 16.5 feet ). Wire sag drops the dipole distance from ground down to about 4.9 meters. I can move these 2 supports to change the dipole direction and currently it runs exactly North to South.  I painted them "radio observatory white".

Under the grass below my dipole lies 43 buried radials, some of which are 18 meters long.

Above — Across each of the three 2X4 legs of my support lies a chain with some nylon tent poles driven into the turf. My poles cannot fall down. After installation, I irrigated the lawn and the grass looks lush and green now.

Above —a skyward view of 1 antenna support rope. 1The actual dipole was constructed  from 16 gauge copper wire.

Above —a skyward look at the feed point when I initially erected my antenna. I can lift the wire up or down for maintenance by tipping the poles and can tighten the slack by separating the poles. I've got 6 FT43-43 ferrites over the coax at the feed point.

Radio Jupiter Pro shows, that at my location, Jupiter will move above the horizon at night-time in ~February 2017. My local RFI levels fall off nicely and I'm able to monitor signals only at night. I'll keep this antenna up and test it in 2017.  I'm hopeful to gain some dipole reference signals to compare to any other antenna schemes I might employ.

Click for Section 3
Click for the Master Index for this project

Introduction and block diagram of all the modules -- Jupiter Modular Receiver

SECTION 1  —  Introduction and block diagram of all the modules

Click for the Master Index to this project

You'll find many good homebrew receivers offered for Jupiter reception and I add 1 more design to the fray. Click for the NASA list of Receivers for Radio Jove. A treasure trove of information lies at the NASA Radio Jove project website.

My hobby: Designing and making scratch receivers evolved skyward once I learned that amateurs around the globe are making radios to receive signals from space.  I share Radionova 1 as fodder for your own experiments and to promote this growing, science-based, fun, hobby.

Many authors have written how to start into radio astronomy and I'll provide a couple of links:

[1] From SARA -- The Society of Amateur Radio Astronomers. Click for their getting started web page.

[2]  Yahoo Groups for the Radio Jove enthusiast. This will connect you to Dr. Eng. Victor Herrero-Arrieta and his links, email posts and further; access to qualified, helpful people for your questions and general information. Victor's radio astronomy blog is linked from my blog.

Block Diagram

[1] From the block diagram above, the other sections should make sense

[2] I built Radionova 1 in discrete modules to allow stage isolation, future changes, module A/B testing - and re-purposing these modules in other receiver systems.

[3] I follow the NASA Jove project of choosing a receive frequency near 20.1 MHz. Full decametric reception ( 20-40 MHz ) requires an alternate local oscillator plus modification to 1 low-pass filter on the main receiver -- and also the front-end filter.

Above — Early photo of 4 of the modules. I'll update this page with more photos when all the modules are built.

Click for Section 2
Click for the Master Index for this project

Monday, 10 October 2016

Citizen Scientist & Build Season 18 begins in 1 week


To help convert to amateur radio astronomy, I needed to brush up on many science topics. I spent much of the summer reading and asking questions about topics that ranged from astrochemistry, parabolic antenna design, general astronomy and pulsars, to Jovian planet RF research from both space probes and earth labs.

Soaked with science, amateur radio astronomy seems the ultimate hobby for those passionate about working with radio waves who also love science. Imagine listening to DX that's actually Gaussian noise lying 60 dB below your receiver noise floor !

I've built up my library with many astronomy books and also periodicals. If you're lucky and your university subscribes to a large catalog of journals, you'll enjoy access to scientific papers that you can locate with popular search engines such as OVID.

Further, I ordered many of the parts needed for my 18th build season over the summer.  I can't wait to start back on the bench in the next week or so.

Above — My 4 favorite books from this summers reading break. Of anything, pulsars interest me the greatest.  I  learned that amateur neutron star enthusiasts experiment within the amateur radio astronomy hobby. Click for the website of Giorgio Dell'Immagine. The Neutron Star Group keeps an anthology of information on this web site

Above — A reading sample. I also spent many nights looking skyward with binoculars and my 150 mm X 750 mm Newtonian light bucket --- and the Sky and Telescope Pocket Atlas proved helpful. Although, I favor binoculars, on the telescope, my main eyepiece = a TeleVue  32 mm Click

Above — Access to scientific papers about radio astronomy + many online scientists proves very exciting for me.

Above — Some of my Jupiter receiver modules which I'll present in a 7-part segment later this Fall and Winter.

Above — My log amp prototype schematic which I'll build and edit this Fall. The radio receiver thus becomes a calibrated measurement receiver with a large dynamic range. This schematic is really just a well-filtered, low noise, calibrated S-meter. I'll drive an ADC with it-- so the output goes to a computer.

After the Jupiter receiver system gets completed, I'm moving to UHF and plan a superhet receiver for 400- 470 MHz with a first IF of 140 MHz.