Small horn 21cm hydrogen line survey finished.

[jan lustrup LA3EQ]

I finally finished the 21cm hydrogen line survey with my small homemade horn antenna.

 

For those interested, may download this big Word dokument with all 241 spectrogram images and equipment information from my survey.

 

This is a 125 page and 5,7MB large file so it may take some time to download for those with a slow internet connection.

 

http://www.qsl.net/la3eq/radioastronomy/profiles/profiling.doc

 

 

For just the rundown info and no profile images then load this one:(only 297kB)

 

http://www.qsl.net/la3eq/radioastronomy/profiles/info.doc

 

 

 

Now, I must build a second horn antenna and try out 21cm interferometry!  Nice spring project. ;)

 

Jan Lustrup, LA3EQ

Norway

 

[Al Aburto]

Excellent work Jan!

--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en

[Bruce Rout]

Very cool. Very nice work.

[jan lustrup LA3EQ]

Thanks for kind words all my friends on my 21cm survey.....

 

 

It was a lot of work, but also alot of fun and educational.

Next time I would use a larger dynamic range( maybe 10 dB or at least the difference of quite sky and ground noise) on  Radio Skypipe chart and never touch the gain or offsett settings until all imagingen was finished. This way, each and every graf could be compaired with each other, both signal wise as well as noise wise. The background ground noise will show up naturally as a rise in signal level+noise level. The vertical scale should  be calibrated in dB.

 

Karl-Heinz...here is a link to the horn parts size in centimeters you asked for. http://www.qsl.net/la3eq/radioastronomy/horn/horn dimisjon.jpg

Important: Be aware that the length size for side panels and top/bottom panels are a little different! This is correct.  

I added 3cm to the 3 big sides on each plate, but not to the the little back waveguid side. These extra cm's are bent 45 degrees to the same outward side. This way you do not need to use 90 degrees angle stock inside to hold every thing in place. Just bore holes on these extra "lips" and poprivet them together on the outside. I addition I use metal tape (not ducktape) to all long inside corners so to be safe that no ground noise leaks into the horn! Bend the tape 90 degrees lengthwise before inserting. The hole for the probe is made to fit your coax conector(SMA or N type...do NOT use BNC type). Close the back with a small piece of metal. The probe (less that a 1/4 wave long and 7 cm from back) can be made of the outer shielding part of UT-141 semi rigid coax. Screw in a small brass screw and brass locking nut at the top for tuning return loss (SWR).

tip....make a tiny scalled  trial horn out of cupboard paper first to make shure all the parts fit correctly. Now you will see why two of the sides need to be a bit shorter! :)

 

Good luck Karl-Heinz and let us know how it works for you. And good luck to all the others doing this project too.

 

 

 

Jan Lustrup  LA3EQ

Norway

[Karl-Heinz DL6EBS]

Hallo Jan,

thank you for your informationen! I hope i get my horn ready this
summer, in the moment i test a little bit with the commercial 17dBi
Kathrein Horn.
regards,
Karl-Heinz
DL6EBS

[Bruce Rout]

Hi,

I'm getting a horn made following the SETI design. I will have a
female N type coax connector sticking out the bottom of the thing.
What do I attach to it? An LNA? I take it I run a feed somehow into a
laptop with appropriate software. Do I amp it before it gets to the
laptop or can I just throw some Marcus magic at it?

-Bruce

[Marcus D. Leech]

> Hi,
>
> I'm getting a horn made following the SETI design. I will have a
> female N type coax connector sticking out the bottom of the thing.
> What do I attach to it? An LNA? I take it I run a feed somehow into a
> laptop with appropriate software. Do I amp it before it gets to the
> laptop or can I just throw some Marcus magic at it?
>

You'll definitely need an LNA right at the feed horn. Something on the
order of 20dB of gain, and < 1dB noise figure.

After that, you need a receiver of some sort. There are a huge plethora
of choices here, and it depends on what your
science goals are, and your budget.

--
Marcus Leech
Principal Investigator
Shirleys Bay Radio Astronomy Consortium
http://www.sbrac.org

[Bruce Rout]

That's what I thought. I believe you have an appropriate LNA and I can
use your software for a receiver. How much are your LNAs?

-Bruce

[Marcus D. Leech]

On 07/04/2011 10:10 PM, Bruce Rout wrote:
> That's what I thought. I believe you have an appropriate LNA and I can
> use your software for a receiver. How much are your LNAs?
>
> -Bruce
>
>

$110.00 apiece

But you still need receiver hardware, Bruce. There's no such thing as a
purely software-defined
receiver--something needs arrange to provide (the veritable torrent)
of samples for the computer
to munch on.

My IRA software lives on top of Gnu Radio, which in turn relies, for
this type of science, on one of
the fine products made by Ettus Research, LLC.

--

[Bruce Rout]

So I get onto Ettus for a receiver which plugs into the laptop. This
is downstream from the LNA which I get from you. You got paypal?
Furthermore, can I arrange an array of these things?

-Bruce

[Marcus D. Leech]

On 07/04/2011 11:04 PM, Bruce Rout wrote:
> So I get onto Ettus for a receiver which plugs into the laptop. This
> is downstream from the LNA which I get from you. You got paypal?
> Furthermore, can I arrange an array of these things?
>
> -Bruce

Again, I'll ask. What are your science goals? I think you imagine
that it's more plug-n-play
than it actually is.

In SDR receivers, the computer is an integral part of the receiving
chain, which means it may be
called upon to process many millions of (complex) samples/second, and
"do stuff" with those
samples. Right now, down-stairs, I'm running my IRA software on a
6-core AMD system running
at 3.2GHz, processing 12.5e6 samples/second, and computing a
high-resolution (~10Hz) FFT,
total power, pulsar folding, and transient detection. That consumes
roughly 60% of the CPU
resources on the system. For just doing HI spectral studies, you need
to be doing roughly 2.0e6
samples/second, to cover the necessary doppler velocities of the
shifted HI line.

Yes, you can "array" these things, but each receiver requires a
dedicated computer to go with it,
and if you want to do "phased array" type stuff, you'll need a
high-quality reference clock for
the receivers, usually a GPSDO, providing 10Mhz and 1PPS reference
clocks. Plus you'll have a
pile of software to write, since my software doesn't handle
phased-arrays of receivers.

[RFSPACE]

Marcus,

Can you share any results of your pulsar observations. I would like to
see some of the pulsar measurements done with the USRP.

Thanks,

Pieter

[Bruce Rout]

To simplify: I want to point the thing up and record 21 cm response as
the sky passes overhead. I would like to record the response in files
as time vs signal strength. That is goal number one.

Goal two is to record across a band around 21 cm as the sky passes
overhead. But I would first like to achieve goal number one.

I would like to start simple, then get more complex step by step. As I
learn, I will expand my objectives.

-Bruce

[E.S. McCauley]

Yes, can I second that,
    it would be terrific to see some of the analysis on this data, ..

• .. 6-core AMD system running
  
• at 3.2GHz,
  
• processing 12.5e6 samples/second,  FFT, ..
  

• pulsar folding, and
  
• transient detection.

• HI spectral studies,
  

• doppler velocities of the shifted HI line.

• even though 12.5e6 s/sec is a bit slow .. :-)
  
• and
• it will likely be a while before ..
• Bruce's data is available.

I am especially interested in your folding method ..
Steve
ACkRO

[Marcus D. Leech]

> Marcus,
>
> Can you share any results of your pulsar observations. I would like to
> see some of the pulsar measurements done with the USRP.
>
> Thanks,
>
> Pieter

I don't have any at this point. The pulsar mode has been waiting for a
dish to go with it for a couple
of years now, which will hopefully be sometime this summer (the SBRAC
dish). My older code,
was used at PARI several years ago to capture pulsar profiles, but I
wasn't directly involved. I don't
know where that data is now.

[Marcus D. Leech]

Yes, can I second that,
    it would be terrific to see some of the analysis on this data, ..

• .. 6-core AMD system running
  
• at 3.2GHz,
  
• processing 12.5e6 samples/second,  FFT, ..
  
• pulsar folding, and
  
• transient detection.
• HI spectral studies,
  
• doppler velocities of the shifted HI line.
• even though 12.5e6 s/sec is a bit slow .. :-)
  

12.5e6 s/sec *is* slow compared to what the major observatories do.  They very often sample at
  100s of megasamples/sec.  But they aren't processing the raw samples directly with a computer,
  but rather they use arrays of FPGAs to peform various FFT and cross-correlation functions, and
  the computers are simply logging integrate-and-dump summary data.

In an SDR design, the samples are streamed continuously to the host computer, so you're limited
  both by the bandwidth of the sampler-to-host connection, and the ability of the host computer to
  "keep up" while doing staggering amounts of float-point arithmetic on the incoming samples.

• and
• it will likely be a while before ..
• Bruce's data is available.

I am especially interested in your folding method ..
Steve
ACkRO

The post-detector data is presented to the pulsar algorithm at a high rate--100ksps is typical, but it
  depends on pre-detector bandwidth, since the post-detector sample rate must be (for efficiency reasons)
  a proper divisor of the pre-detector bandwidth.  A "mission elapsed timer" is maintained, which moves
  forward by 1.0/pulsar-sample-rate for every sample.  Samples are binned according to the pulsar
  cadence and the "mission elapsed timer".  Bins are filtered with a single-pole IIR filter.

Furthermore, the pre-detector bandwidth is coherently de-dispersed using an FFT filter based on the
  work of Hankins and Rickett.

[E.S. McCauley]

In Aus,
    quite often little car clubs develop on Friday nights. The cars are
marvelous, polished to a tee and with all the latest extractors and fat
chromed exhausts. Young lads with slick hair and flash clothes stalk
about in front of breathy young ladies. And then the rubber burns as the
backs of hacked Monaros (and others) slide across the road. The
smoke and stench raises the hackles. The girls swoon.
    But in the end I always wonder where the data is. Do they ever race?
And if they do - do they win? The girls perhaps - but what might they
have learned of cars and Racing - if they really put their car to the test ..
right under the pump. On the race track.
    I would Really like to see some data from your work Marcus.

Steve
ACkRO

--

[jan lustrup LA3EQ JO28XJ]

Hi Bruce,
The simple approch is to detect the sky passing by (maybe for a 24 hour
period). Start by getting a band pass filter for the 21cm band (If the pre
amp has one built inn, good, then you won't another on) and then you need
som more amplification,,,,,( I used som cheap inline satelite amps found on
E-bay for a few dollars apice), then a diode detector. Now you have a dc
signal to work with.....build a simple intergrator (1uF cap and 2M Ohm
resistor) and feed this to an op amp (ln-741 or better), this will get you
10 to 50 times more gain.......and then feed this to an A/D converter...(Jim
Sky sells a cheap kit for this) witch connects to your PC printer port or
if you want to use the USB port then buy a "Labjack" 12 bit A/D converter.
Now get "Radio skypipe II" installed in your PC. Jim sky sell this
too....(download a "free version" to get your feet wet).

Now you are set up for "Total Power" observations of the sky. Point your
Horn antenna due south and upward a bit and let the heavens past by for a
day (called a Meridan transit) and watch your recordings on Radioskypipe. If
you are online 24/7 you can shear your recordings live with others, or look
in realtime to other peoples recordings from around the world.
Try pointing the Horn in differnet elevations and see the change over time
on your recording,!!
"Radio eyes" program will let you see what to expect and give you antenna
elevation for different things of interest.

Maybe your next step will be to do a scan of the hydrogen-line and observe
the H-line profile....
To do this, you keep your Horn, your LNA , detector, intergrator and op amp,
and your radio skypipe program. You will need a receiver with a VFO that
will scan, There are some cheap second hand amateur HF, VHF, or UHF
receivers on E-bay that will scan and to use them you will need a down
converter from 1420MHz to your receiver frequency. You can build one
yourself or buy one ready made. Let the down converter be as close as
possilble to your LNA / antenna as to minimize cable loss at 1420 MHz.(Cable
loss at 28MHz 144MHz og 432MHzis much less)
You let the reciver tune a 1 MHz slice of the hydrogen band (1420.5MHz in
center). Turn off the AGC control and feed the audio output fro the receiver
to yor detector diode which in turn is conected to the inrergrator-opamp-a/d
converter-PC as before.

If you want to play around with the Radio skypipe, then I could send you a
.WAV fil sample of the hydrogen line profile you can play back and watch it
on radioskypipe, do changes in the program to see how it works. Save it to
tape, play it back and you can try out your diode detector and
intergrations setup, and your A/D too...lots to try out and to learn...

Good Luck,
Jan Lustrup LA3EQ
Norway

----- Original Message -----
From: "Bruce Rout" <bbr...@rasl.ca>
To: <sara...@googlegroups.com>
Sent: Tuesday, July 05, 2011 5:04 AM
Subject: Re: [SARA] Small horn 21cm hydrogen line survey finished.

[Marcus D. Leech]

On 05/07/2011 7:55 AM, E.S. McCauley wrote:

In Aus,
    quite often little car clubs develop on Friday nights. The cars are
marvelous, polished to a tee and with all the latest extractors and fat
chromed exhausts. Young lads with slick hair and flash clothes stalk
about in front of breathy young ladies. And then the rubber burns as the
backs of hacked Monaros (and others) slide across the road. The
smoke and stench raises the hackles. The girls swoon.
    But in the end I always wonder where the data is. Do they ever race?
And if they do - do they win? The girls perhaps - but what might they
have learned of cars and Racing - if they really put their car to the test ..
right under the pump. On the race track.
    I would Really like to see some data from your work Marcus.

I had data several years ago--scads of it, some of which was published in my first SDR paper at the SARA 2006 conference.
  I lived on a 37-acre farm, where I had installed a 12ft dish.  It was in a reasonably radio-quiet location.  I was taking
  data on a daily basis, doing scans of Sgr A, Cass A, etc.  Getting good spectral profiles of the hydrogen line, etc.
  But that all came to a grinding halt when my (now ex) wife decided that she didn't want to live out in the country
  any more.  I used to publish interesting graphs on my Propulsion Polymers website.  But that business doesn't exist anymore,
  and due to a disk crash there in the months before I shut down the business, the data from those days is just gone.

In the intervening time, I moved to a location where I can't do radio astronomy in my back yard (except for the riometer
  and VLF receiver).  The riometer was getting reasonable "first light" data until the city installed new induction lighting, and
  then 8-12 hours a day of data became utterly useless.  So, I have to move the riometer to a quiet location, but I have
  constraints on my time.  Full-time job, etc.  Can't just up and take my gear to the hinterlands for a week :-(

I'm still waiting for my new dish antenna (http://www.sbrac.org) to reach a state where I can use it for data.  At such time
  as that happens, there will be on-going public access to all the data.

So, yes, I have bright shiny toys, and haven't been able to apply them to the very-important work of data-taking for  a few
  years now since I lost my 12ft dish.  So, in the mean-time, I work on improving the software, and testing it as best I can
  without a dish.  None of the handful of customers of IRA are currently using it for pulsar work, but I'm sure they all have
  scads of spectral and total-power data they can share.

I publish my VLF data on a daily basis at the Stanford "SID" site.  While those data aren't based on USRP samples, the
  Gnu Radio framework that is used to produce it is the same framework as I use for IRA and the Keo Riometer.  The VLF
  receiver is the only thing that works adequately at my current location, so I contribute lumps to the dataverse as best
  I can.

[Marcus D. Leech]

On 05/07/2011 12:26 AM, Bruce Rout wrote:
> To simplify: I want to point the thing up and record 21 cm response as
> the sky passes overhead. I would like to record the response in files
> as time vs signal strength. That is goal number one.

Simple total-power observations are relatively straightforward, you'll
need an LNA, filter,
plenty of gain and a detector of some sort. That *could* be
SDR-based, or something
else.

Bruce, I'm reluctant to separate you from your hard-earned money by
suggesting that you rush out and
buy a bunch of SDR equipment, because it's a significant investment,
and I think that, from an engineering
perspective, you may be in the "learning to walk before you can fly"
stage.

My own LNAs are quite good, but there are others out there as well. You
could order one from me, but I should
warn you that delivery times will be long, since I moved out of my
commercial space a couple of months ago.
You might get shorter lead-time ordering from Down East Microwave, or
Radio Astronomy Supplies.

[E.S. McCauley]

It is my guess that
    god hates you. Either he or possibly satan is really dead set against you. But
it is equally clear to me that you know more and can do more than three quarters of
the biomass of SARA. You can't let all this rubbish get at you! If you are a radio
astronomer you are drawn to the stars and the subtle sparse information they send
our way.
    I intend to build a truck setup and go mobile. Marvelous stuff and you get to go camping
too. My first targets are the Big Desert and the Sunset Country (NW Victoria), fabulous
birdlife and anything you want from the Murrayville General Store.   You lot have got a car
on Mars, .. surely a couple of dishes and a bucket of computers in the mobile home is
not too hard ..?

Steve
ACkRO
oh .. and a hard drive for the data ..

[Bruce Rout]

Marcus, as always, you are spot on. I am trying to walk first one step
at a time. I'll write you on email and there is no rush since I
haven't gotten the horn yet (and I also need to mount it). Jan, your
suggestions are extremely helpful. Thank you so much and I will
definitely get back to you as I progress.

-Bruce

[Rodney Howe]

In response to Steve's data capture request? 

 

DSES has years (2004 -2009) of spectral drift scan data captured with the 18 meter dish.  http://www.deep-space.org/images/smp/interim_report_v200606.pdf 

 

And, specifically these data are looking at the Cygnus clouds. Oddly enough these data cover the same patch sky the Kepler Satellite has been covering for the last year or so.  http://kepler.nasa.gov/Science/about/targetFieldOfView/  So, it might be an interesting project to crop these data to the Kepler FOV and SEE IF THERE ARE ANY SIGNALS !! from out there ;-) 

 

This would take some analysis and programming skills I think.

 

R.

 

 

 

[Marcus D. Leech]

On 05/07/2011 9:54 AM, E.S. McCauley wrote:

It is my guess that
    god hates you. Either he or possibly satan is really dead set against you. But
it is equally clear to me that you know more and can do more than three quarters of
the biomass of SARA. You can't let all this rubbish get at you! If you are a radio
astronomer you are drawn to the stars and the subtle sparse information they send
our way.

Life certainly hasn't been overly kind to me in the last few years, to be sure.  Not ready to ascribe that to the
  actions (or inactions) of anyone's favourite deity/anti-deity.

    I intend to build a truck setup and go mobile. Marvelous stuff and you get to go camping
too. My first targets are the Big Desert and the Sunset Country (NW Victoria), fabulous
birdlife and anything you want from the Murrayville General Store.   You lot have got a car
on Mars, .. surely a couple of dishes and a bucket of computers in the mobile home is
not too hard ..?

If I were wealthy enough for that, it's certainly something I'd consider.  But see above, re: life kindness.

[E.S. McCauley]

Got that Rodney,
    I'll have a look. While it's certainly true that a good analysis takes ages, there are often a number
of interesting "quick grab" insights for the thoughtful.
S.

[James H Van Prooyen]

Hi All Marcus, it was interesting to read about you approach to pulsar detection work, thanks for the description. All, good e-mail thread, I enjoyed reading each of your views... Thanks Jim Van Prooyen

[Marcus D. Leech]

The proof of the pudding is in the eating, as they say.

So everybody send good vibes towards the SBRAC project, and perhaps there'll be some good pulsar data in the coming months.  We have
  both a 21cm and 400-600Mhz feed ready to go in place.  There's just this small matter of getting up to the feed point, given that we
  have no money to pay for a crane, and the existing servicing gantry is in such poor shape that it might kill whoever attempts to use
  it :-(

[E.S. McCauley]

Oh bloody hell Marcus,
    get off the floor boards. We don't believe you.

S.
did you find that hard drive .. ?

[Marcus D. Leech]

Oh bloody hell Marcus,
    get off the floor boards. We don't believe you.

You are, at times, utterly unparseable.

S.
did you find that hard drive .. ?

-- 
Marcus Leech
Principal Investigator
Shirleys Bay Radio Astronomy Consortium
http://www.sbrac.org

[E.S. McCauley]

Hi Marcus,
    the truth of the matter is that I am a bit of a pr*ck when I get going. I have all that practice
as a school teacher - seeing the problems and nailing the bloody things (so I think). I am
far too fast at times with the mouth/word perhaps. My apologies if I have gone beyond the
bounds of what is fair to say.

regards
Steve
War can protect, it cannot create - WHITEHEAD

--

[Marcus D. Leech]

Hi Marcus,
    the truth of the matter is that I am a bit of a pr*ck when I get going. I have all that practice
as a school teacher - seeing the problems and nailing the bloody things (so I think). I am
far too fast at times with the mouth/word perhaps. My apologies if I have gone beyond the
bounds of what is fair to say.

regards
Steve
War can protect, it cannot create - WHITEHEAD

Didn't really know what to make of it.

You might do well to remember that we're all mostly, at least notionally, grown-ups here, and snarking at us like we're a room
  full of unruly, deceitful school kids might not be as productive as you think.  Just friendly advice.

[E.S. McCauley]

.. yes, - definitely made a blunder, ..
    never had a wife you see .. these girls keep a man in one piece ..
well who knows - .. plenty of fish in the sea.
S.
sorry ..

--

[Dale Hooper]

Speaking of precise time references, I received some info about the following atomic clock on a chip a few days ago. They have single unit prices of about $1500 and a development board available for about $800. It provides 10 MHz OUT and 1 PPS OUT.

http://www.symmetricom.com/products/frequency-references/chip-scale-atomic-clock-csac/SA.45s-CSAC/

Clear skies,
Dale.

> -----Original Message-----
> From: sara...@googlegroups.com [mailto:sara...@googlegroups.com] On
> Behalf Of Marcus D. Leech
> Sent: Monday, July 04, 2011 9:22 PM
> To: sara...@googlegroups.com
> Subject: Re: [SARA] Small horn 21cm hydrogen line survey finished.
>

[Marcus D. Leech]

On 06/07/2011 11:27 AM, Dale Hooper wrote:
> Speaking of precise time references, I received some info about the following atomic clock on a chip a few days ago. They have single unit prices of about $1500 and a development board available for about $800. It provides 10 MHz OUT and 1 PPS OUT.
>
> http://www.symmetricom.com/products/frequency-references/chip-scale-atomic-clock-csac/SA.45s-CSAC/
>
> Clear skies,
> Dale.

Yup, saw that. Also, Jackson Labs has a full GPSDO product based on
the symmetricom chip as the "backing reference".