Budget Radio Telescope for DIY

8/10/2019 Budget Radio Telescope for DIY

1/14

A 21cm Radio Telescope for the Cost-ConsciousMarcus Leech, Science Radio Laboratories, Inc

Abstract

We show two slightly-different designs for a simple, small,effective, radio telescope capable of observing the Sun, andthe galactic plane in both continuum and spectral modes,easily able to show the hydrogen line in various parts ofthe galactic plane.

Introduction

The emergence of new, relatively-inexpensive tools for the RF tinkerer has allowed a re-examination ofthe bottom end of small-scale and amateur radio astronomy !ntil "uite recently, anyone on a verysmall budget, or with outdoor space constraints would likely be stuck with a simple radio telescopesuch as theI!" #hile such an instrument is an excellent device for portable classroomdemonstrations of the very-basic principles of radio astronomy, this author wished to explore that spacefurther $pecifically, the possibility of an instrument with a roughly %m reflector, capable of observingthe hydrogen line, and the continuum emissions from the galactic plane, in addition to the $un $incethe &'T is only really capable of observing the $un, it was felt that a richer educational experiencemight be reali(ed by having a slightly-more-capable instrument within a similar budgetary envelope

Software Defined Radio

&n the past, having a radio astronomy receiver that was capable of providing spectral information wasout of budgetary reach of the small-classroom environment Fine instruments such as the Spectra#ybermade by Radio )stronomy $upplies, for example, are easily affordable by most higher-eductioninstitutions, but usually not at the high-school level

&n the past few years, Software $efined Radio*$+R has emerged as an exciting new technology that isincreasingly being leveraged by amateur radio astronomers to expand their experimental hori(onsrelatively cheaply #ith an $+R, the RF signals are typically converted to so-called baseband form,and then digiti(ed at one or two megasamplessecond, and then forwarded to a ./ for furtherprocessing &n this way, the signals are presented in their rawest form, so that they may be processedin many different ways to extract different types of experimental data #ith $+R, there is no hardwarethat is specifically dedicated to continuum observations, or spectral observations, or other types of

speciali(ed modes )ll the processing happens in software, rather than analog hardware$oftware frameworks, such as 0nu Radio have significantly simplified the ad-hoc construction ofsoftware-based signal-processing chains, allow many different types of experiments to be conductedon real-time radio telescope data, using $+R receivers as input

% $ee, for example1http1wwwgbnraoeduepoibtshtml
http://www.gb.nrao.edu/epo/ibt.shtmlhttp://www.gb.nrao.edu/epo/ibt.shtmlhttp://www.gb.nrao.edu/epo/ibt.shtml


2/14

Receiver Hardware: RTLSDR

&n early 23%2, it was discovered that certain +4'-T dongles *!$' devices intended to allow you towatch over-the-air +4'-T broadcast television had a special mode that allow them to be used ascrude $+R receivers &n this mode, the digital base-band samples bypass the +4'-Tdemodulatordecoder in the RT52672! chip, and are sent over !$' 8ormally, these devices sendpartially-decoded 9.:0 transport frames over the !$', but in this $+R mode, they send raw &;base-band samples instead

These dongles are typically shipped with one of two or three different tuner chips, and the mostpopular ones, the R623T and :


3/14

$hielded enclosure

The author has used :ttus hardware in radio astronomy proDects since 233> and finds that they are anexceptional value

The radios are priced as you might expect for a niche market item intended for technology

developers, however, in that market they are extremely competitively priced. For a school wishing toget noticeably-better results, the :ttus '%33?#'B9 in diameter or somewhat larger #hilesuch a dish affords excellent portability, the aperture is somewhat too small for sensitive observationsof the galactic plane at 2%cm

The author has achieved excellent results using an offset-fed satellite dish, measuring E7cm x 6>cm,and it is suggested that experimenters try to find a dish of roughly this si(e $uch small, offset-feddishes are available online>from a number of online satellite retailers as well as e'ay &n many largecities, there are still brick and mortar satellite e"uipment retailers where such dishes may bepurchased reasonably cheaply

ccasionally, you may find people discarding these dishes, having migrated back to cable televisionaway from satellite television

&tAs important that any dish ac"uired include the feed-arm hardware There are also many packagedeals that will sell you a dish and a Gu-band 58' The Gu-band 58' may be used in otherexperiments, but primarily it helps to give a good notion of how to position the 2%cm feed-hornassembly that is described later in this article

The feed-arm hardware will have to be modified to accommodate the much-larger 2%cm feed-horn

assembly

Alternative hardware: ~1m prime-focus dishes

There are also small prime-focus dishes available from some places, roughly %7m in diameter Thesehave the advantage that theyAre easier to point, but are generally significantly more expensive, andharder to find, than the offset dishes of roughly the same diameter 9anufacturers such as Honsa andForetec $tarImake such dishes, but it can often be hard to find local distributors

The feed assembly for these dishes is less complicated to make as well, since they tend to be deeper*376 to 3lb of coffee

< $ee1https1wwwettuscomproductdetails!$R.J23&nstantJ23$+RJ23'%33J23'undleJ23-J23'%33J23?J23#'BJ23'undleJ23J26>3-22339C(J2EJ23?J235ive!$'

> $ee1http1wwwsadouncom$atrder+ishesFree-to-air-satellite-dishhtmI $ee1 http1angelelectronicscashopsatellitedishesfortecstar-fc3I
https://www.ettus.com/product/details/USRP%20Instant%20SDR%20B100%20Bundle%20-%20B100%20+%20WBX%20Bundle%20(50-2200MHz)%20+%20LiveUSBhttps://www.ettus.com/product/details/USRP%20Instant%20SDR%20B100%20Bundle%20-%20B100%20+%20WBX%20Bundle%20(50-2200MHz)%20+%20LiveUSBhttps://www.ettus.com/product/details/USRP%20Instant%20SDR%20B100%20Bundle%20-%20B100%20+%20WBX%20Bundle%20(50-2200MHz)%20+%20LiveUSBhttp://www.sadoun.com/Sat/Order/Dishes/Free-to-air-satellite-dish.htmhttp://www.sadoun.com/Sat/Order/Dishes/Free-to-air-satellite-dish.htmhttps://www.ettus.com/product/details/USRP%20Instant%20SDR%20B100%20Bundle%20-%20B100%20+%20WBX%20Bundle%20(50-2200MHz)%20+%20LiveUSBhttps://www.ettus.com/product/details/USRP%20Instant%20SDR%20B100%20Bundle%20-%20B100%20+%20WBX%20Bundle%20(50-2200MHz)%20+%20LiveUSBhttp://www.sadoun.com/Sat/Order/Dishes/Free-to-air-satellite-dish.htm


4/14

KouAll need to empty the can, either through gradual consumption of the contents, or something a bitmore draconian

The cans are usually sealed with a thin aluminum membrane as well as the plastic lid &f making a feedfor a prime-focusdeep dish, retain the plastic lid, it will make a useful weather cover

nce youAve removed the sealing membrane, there will be a little ring of aluminum still attached to thecan !se a can-opener to remove it

#ash and dry the can using a bit of dish soap

+rill a hole through the side of the can for your chosen connector Type L8A is popular, but this authorused a tin-plated type LFA since all his plumbing is type-F compatible The hole should be 7 fromthe bottom of the coffee can, as shown below

&f youAre using a tin-plated type LFA bulkhead or ./' edge-mount connector, you can solder theconnector directly to the coffee can KouAll need to use a soldering-gun to do that to provide the extraheat that will be necessary

&f you use a type L8A or other connector, youAll have to drill holes as appropriate /hoose a connectorthat is compatible with whatever 58) youAll be using, with minimum adaptors re"uired, to reduce

front-end losses

ne thing to note is that drilling thin steel sheet re"uires considerable care, and itAs usually necessary tostep your drill si(es up gradually, so as not to try to remove too much material at a time

KouAll need to solder a probe to the center conductor of the connector, approximately %6 long, andmade from %3 or %2-gauge copper wire 'rass tubing also works well

&f using a prime-focus dish with a F+ of 376 to 3, the coffee-can feed alone is sufficient to produceacceptable results .ut the plastic lid back on, and cut a small cross with a sharp knife in the middleof the lidMthis allows pressure e"uali(ation so you donAt get moisture pumping into the can interior'efore you do this, you can also give the can exterior and interior a couple of light coats of clearlac"uer spray-paint N this will slow down the inevitable corrosion of the steel +ont forget to tape-off


5/14

the connector when youAre doing thisO

&f using a shallow offset-fed dish, youAll need to augment the coffee-can feed with a pair of taperedadapters that increase the gain of the feed, to properly illuminate the dish

0o to your nearest C4)/ e"uipmentducting supplier and get a I to 6 round galvani(ed taperedadapter, and another 6 to %3 adapter &f you can, get the type that are crimped on one side, so thatthey mesh together easily therwise, youAll need to use tin-snips to create a few slotsMone set toallow the I to 6 adaptor to fit the coffee can, and the other to allow the I to %3 adaptor to fit intothe I to 6 adaptor nce you have them all fit together it should look, in profile, roughly like this1

+rill some small holes, and use self-tapping 76 long, PI sheet-metal screws to hold all piecestogether !se them liberally to improve mechanical rigidity, and electrical conductivity between all thepieces

)gain, apply a couple of light coats of clear lac"uer to the assembly *although tape the connector, sinceyou donAt want to make it non-conductive

) sheet of clear polyethylene over the end will keep birds from wanting to nest in the feed 9ake surethereAs a small vent hole somewhere


6/14

Hardware: Feed Arm Modifications

&f using an offset-fed dish, modifications to the feed arm assembly will be necessary to make sure thatthe focus is in the middle of the %3 throat of the feed $ome amount of inventiveness will be re"uired

here, since the feed-arm arrangements are different on different dishes

n the authors E7cm offset dish, thereAs a single support arm that comes from behind the dish aroundto the front Removal of roughly %2cm of feed arm allowed for the %3 horn throat to be at roughly thecorrect location at the focus

The feed should be aimed so that itAs pointing at Dust below the center of the dish )gain, a certainamount of inventiveness will be re"uired, and itAs likely that post-construction signal measurementswill need to be made to optimi(e the feed location and pointing

&f all the feed hardware was supplied for the dish, along with a cheap 58', you can use that as a guideas to where the focal point is, and adDust accordingly for the much-larger feed-horn for 2%cm

The author uses an 5 bracket attached to the front of the feed, and bent so that the feed points at thecorrect location on the dish, with the other side of the 5 bracket attached with band clamps to the feedarm

Hardware: Low Noise Amplifier LNA!

&t is vital that the first stage of amplification in a radio telescope have as little self-noise as practicalThis is known as the%oise &igureof the amplifier $ince the signals weAre looking for are very faint,having a low noise figure improves our chances of detecting them &n selecting a low-noise amplifier itshould be1

Q %d' noise figure *8F

23d' gain

&deally, it should be filtered to allow only the 2%cm band to pass

ptionally, it can be powered via the output connection, so that +/ can be fed up the coax

$uch 58)s are available from a few suppliers on the &nternet, includingRadio 'stronomy Supplies,and$own (ast Microwave

Hardware: Line amplifiers

KouAll very likely need two more stages of amplification before the signal is presented to your receiver

The most convenient method is to use /)T4$)T-T4 line amplifiers that offer between %3 and23d' of gain, but have poor noise figures These units are usually available for about =>33 each, andare powered over the coaxial connection on their output

&f you choose to power the amplifier-chain via the coax, it is also necessary to obtain a +/ powerinDectorMavailable from the same suppliers as the line amplifiers


7/14

Hardware: "oa#ial cablin$

The author has a distinct preference for @>-ohm, R0-I; cabling, because is of sufficient "uality, andlow-enough cost that it can be used in longer cable runs The /)T4$)T-T4 ecosystem thatproduces R0-I cabling and assorted connector technologies has a lot of good bits of plumbing thatare both inexpensive and high-"uality /ompression connectors, for example, have completelychanged the @>-ohm small-cable world from rather cheesy to first-rate #hich was a necessity whenthose cabling systems started to be used for $)T-T4 systems

R0-I and R0-I; *"uad-shielded R0-I cable is available at most hardware stores, including Come+epot, 5owes, etc They also usually have the specialty compression connector tooling and connectorsavailable, although on-line dealers of such e"uipment generally offer much better prices

Hardware: A %" computer

KouAll need a computer to process the data coming from the receiver, and do useful things with it#hile you donAt need the latest, high end gaming machine, you should avoid pulling something outof the trash bin for your $+R-based radio astronomy experiments

) dual-core system of ca 23%3 to 23%2 vintage, running at 260C( or better, with at least 20' of ramis recommended to support the 0nu Radio software that this author uses

The author runs the observing software on his )9+ .henom && B< system, which has 70' of ram, andruns at 720C( &t consumes about %7rdof that system in terms of /.! resources, at %>9sps samplerate

For both :ttus hardware and the RT5$+R, youAll need a spare !$'-23 port available on the system

Hardware: %uttin$ it all to$ether

nce you have assembled all the individual pieces, including mounting the dish on an appropriate

platform, itAs time to this collection of parts into a cooperating observing system

CereAs a diagram showing the overall layout1


8/14

Kou may be able to eliminate one of the line amplifiers if you have a shorter coaxial cable run to your

observing station )nd youAll need RF connector adaptorsMexactly which ones depends on yourchoices of cabling systems, etc, and which $+R you chose

&oftware: Linu#: 'he system of choice for science the world over

The author is an unapologetic 5inux (ealot The suggestion would be to install a recent !buntu *%23


9/14

experiments to be conducted "uite easily

nce you have a 5inux $ installed *!buntu and Fedora are likely the least-painful choices, you canuse the build-gnuradio script to install the latest 0nu Radio and hardware-driver plugins from thesourcecode repositories

!se the script at1 http1wwwsbracorgfilesbuild-gnuradio

This script will take "uite some time to run $o have a coffee *which you should have plenty of at thispoint, or maybe two or three Kou might even want to consider a "uick vacation to #est 4irginia orsomething while itAs running also, but thatAs up to you

Kou may also need to install pyephem via, pip-python1

pip-python install pyephem

&f your system doesnAt have pip-python, then youAll need to install it first, using your package-managerThe package name is usually python-pip

&oftware: 'he simple_ra(nu )adio applicationRecall that we referred to 0nu Radio as a signal-processing framework &t isnAt in and of itself anapplication, but rather is used to buildapplications that do ./-based signal processing

The simpleSra application is maintained at thehttps1wwwcgranorgsite which has many applicationsthat use the 0nu Radio framework

&nstalling simple)rais fairly easy First1

svn co https*++www.cgran.org+svn+proects+simple)ra

This will create a directory called simpleSra Kou then Dust need to makeinstall it1

cd simple)ra+trunmae

mae install

There is a R:)+9: file in the trunk directory that you are $TR805K encouraged to read

%re&inar' resu&ts

First, a little bit of sky at declination ?


10/14

Then a similar bit of sky, but this one, roughly %6 degrees higher in declination )gain, the center isroughly in the region of /assiopeia1


11/14

'oth of those plots show so-called total power or continuum plots as seen by the radio telescope&n some sense, a radio telescope is like a single-pixel light-meter, showing you the sky brightness as

parts of the sky move through the beam 'oth of these continuum plots are showing the effective skybrightness of the 0alactic .lane as it moves through the beam of the instrument

'oth of the above plots were produced with the processSsimpleStpat tool that comes with simple)raThe processing tools have the ability to automatically overlay multiple-days data, aligned on 59$T, toproduce composite sky plots This allows the ability to produce, over time, high-$8R sky maps

$ince the instrument we describe here is a $oftware +efined Radio, it is able to do other things as well,namely, produce spectral data at the same time as it collects continuumdata

Cere is a spectrum of the hydrogen line, taken near the peak of the transit of the galactic plane near#ygnus


12/14

8ote how the plot is calibrated invelocity, rather than, for exampleM/ The script that producesthese plots,process)simple)specdat can automatically calculate the doppler shift from the data, sincethe data are internally annotated with respect to center fre"uency of the observation

&n the plot above, we can clearly see a triple humped spectral structure, indicating that at the time ofthe observation, there were three distinct hydrogen clouds within the field-of-view of the radiotelescope

CereAs a plot of a bit of hydrogen towards )uriga


13/14

8ote the excellent $8R of these plotsMthe simple)ra receiver application performs a fair amount ofintegration on the individual specta, to improve the $8R, and theprocess)simple)specdat script furtherimproves the $8R by averaging all the log records within the given time window, with simple)ranormally logging high-resolution *%>> meterssec in the above plot spectral estimates every 73seconds

"onclusion

)dvancements in technology have allowed back-yard radio astronomers to perform observation featsthat in years past might have them branded as witches &n the early %E>3s, when :wen and .urcell firstconfirmed the existence of the hydrogen line@, they had a budget of =>3333, operating out of Carvard

!niversity &n todayAs dollars, thatAs a roughly =33 expenditure $till, an outstandingR& for ascientific observation that was of such extreme importance

Geep in mind that :wen and .urcell werenAt able to determine the spectrum of neutral hydrogen, Dustconfirm that it must exist along most lines of sight through the galaxy, both confirming the theory of4an +e Culst, and at the same time laying-to-waste 4an +e CulstAs statement that weAd likely never beable to observe the line, due to its very-diffuse *and thus, weak nature

:wen and .urcell had an instrument with an aperture of roughly %2m N roughly %I times the

@ $ee1 http1wwwnraoeduwhatisrahistSewenpurcellshtml
http://www.nrao.edu/whatisra/hist_ewenpurcell.shtmlhttp://www.nrao.edu/whatisra/hist_ewenpurcell.shtml


14/14

collecting area of the dish described here Ket with an investment of less than =23333 of todayAsdollars *roughly =2>33 in %E>%O, an amateur can vastly exceed the kinds of observational "uality that:wen and .urcell were able to make from the roof of a building at Carvard, back in %E>%

) proDect of this type is a worthwhile enterprise both for the individual, and schools both at thesecondary, and post-secondary levels

Budget Radio Telescope for DIY

Documents

Transcript of Budget Radio Telescope for DIY