Anyone doing hydrogen line observing?

[Carl Reade]

Hi Coto yes drift scan. I use a program called Sellarium to show when and where the Milky way is. I then use the alt/az to set the antenna using a compass and inclinometer.

The beamwidth is 3db 15 degrees. Frequency 1.40 to 1.44Ghz.

To produce the data I need to shut down Sharpsdr processes so it's running in the background then use another program to gather spectrum FFT data every 5 mins. This data can then be graphed in Excel.

A quick way is to take a 5 mins file about and hour before the galaxy then one mid pass then devide the two in Excel on a graph and you will see the result.

For processing here is a link to a site which is fantastic. With amature gear you are trying to listen to a conversation at a concert. This is one way of doing it.

https://www.google.co.uk/url?sa=t&source=web&rct=j&url=http://parac.eu/projectmk3.htm&ved=2ahUKEwj6ot3d5KPdAhXDWsAKHdvQCS8QFjALegQIBRAB&usg=AOvVaw3JqQ9Z-YTu83lhKfmFUHoW

[Coto]

6 hours ago, robin_astro said:

This shows the spectrum of neutral hydrogen at 21cm, Doppler shifted by different amounts in different spiral arms. You cannot measure these Doppler shifts in continuum spectra like synchrotron or thermal emission.

I don't understand what you're saying. By spectral observation I mean observing a wide range of frequencies simultaneously (depending on the antenna bandwidth (not frequency hopping)). I wasn't referring to the continuum emission of the target. His graph* shows the spectrum of neutral hydrogen at 21cm? Looks like an Intensity vs Time graph to me.

(Yes, I knew you can't detect Doppler shifts with synchrotorn/thermal radiation observations.)

*The .docx file download link seems down, so there might have been something more in there that I missed?

[Carl Reade]

32 minutes ago, Coto said:

I don't understand what you're saying. By spectral observation I mean observing a wide range of frequencies simultaneously (depending on the antenna bandwidth (not frequency hopping)). I wasn't referring to the continuum emission of the target. His graph* shows the spectrum of neutral hydrogen at 21cm? Looks like an Intensity vs Time graph to me.

(Yes, I knew you can't detect Doppler shifts with synchrotorn/thermal radiation observations.)

*The .docx file download link seems down, so there might have been something more in there that I missed?

Hi the chart is a 2.4 MHz wide spectrum chunk centred on the hydrogen line around 1420.410. A live snap shot. It's in an Excel chart to define and enlarge. If that helps.

[robin_astro]

4 hours ago, Coto said:

I don't understand what you're saying. By spectral observation I mean observing a wide range of frequencies simultaneously (depending on the antenna bandwidth (not frequency hopping)). I wasn't referring to the continuum emission of the target. His graph* shows the spectrum of neutral hydrogen at 21cm? Looks like an Intensity vs Time graph to me.

The clue is in the thread title This explains the measurement. (From the Jodrell Bank radio astronomy course) 

http://www.jb.man.ac.uk/distance/radio/course/sourcesII/sourcesII4.html

EDIT: also here from University of York.

http://www-users.york.ac.uk/~phys38/galhydexp.htm

A classic practical experiment for any radio astronomy student

Robin

Edited September 5, 2018 by robin_astro

[Coto]

@Carl Reade Please reupload the .docx file so I can get a better understanding. Why did you pick Cygnus A as your target instead of any random point in the sky (outside the Milky Way line) and do the same if you want to observe the Doppler shift in the Milky Way?

[Carl Reade]

Hi Coto forget the docs it was just a bigger photo of the one posted.

Forget Doppler at the moment.

As the Hydrogen line is a very weak source to see in the noise (unless you have a very large dish) the best way to see if your scope is working is to point away from the miklyway and record a spectrum for 5 mins.

This should be flat. Then point at the miklyway take another 5 mins recording.

Place the two readings in a graph in Excel then divide the two and you get the difference between them. Which should be the hydrogen line.

To get a good Doppler reading you then take a few hours of 5 mins recordings then place them all in a graph and you should see the hydrogen line hump move through the graph at an angle though time.

The link I posted has examples in post processing under projects on the page.

Carl

[Carl Reade]

The latest upgrade hopefully bit more gain.  A 1.2m offset dish with a biquad feed. Have to say the feed was a tricky one to build adapting thing to fit was a nightmare but got there.

The biquad was sized to work on 1.420Ghz via an online calculator.

No doubt testing and tweeking to follow.

[Carl Reade]

It works!!

Set the dish up due South and roughly 70 deg elevation last night for a drift scan in the Cygnus area. was not sure if there would be anything as its a rough focal point and my engineering skills are rusty.

I made an animation of 8 screenshots from 20:30 to 01:00 ish. Basically from start of pass to end with peak inbetween.

This is live spectrum on screen with no processing. Its the highest resolution i can get on SharpSDR.

The dish is a winner over Yagi which is no surprise. not so much signal strength but I think there is a lot more resolution.

Setup is Dish with biquad feed 30db 0.5 LNA, 2 x inline sat amps 20db each, bias tee, Filter then SDR dongle.

Hopefully the GIF works.

Carl

 

 

[robin_astro]

A very nice signal !  

I guess the dish will have a better front/back ratio too which should improve SNR.  I would be interested in the details of the feed design

Cheers

Robin

[Carl Reade]

54 minutes ago, robin_astro said:

A very nice signal !  

I guess the dish will have a better front/back ratio too which should improve SNR.  I would be interested in the details of the feed design

Cheers

Robin

Hi Robin thanks chuffed it's working. Yes to get the Yagi working I added a a reflector tin biscuit lid which helped big time and eliminated some ground noise.

The feed is a biquad normally used for WiFi long range and works well with dishes. I used the following calculator. Use the second calculator the first is for a four-way.

https://www.google.co.uk/url?sa=t&source=web&rct=j&url=https://buildyourownantenna.blogspot.com/2014/07/double-biquad-antenna-calculator.html&ved=2ahUKEwiNxNab8K3dAhWMLsAKHQVYB0cQFjABegQIBRAB&usg=AOvVaw1JCwSFCn-kzwmHTIRK8xWT

There are various ways of making it. I used 1.5mm electrical solid core to make the diamonds. Would recommend using a copper reflector, I only had a piece of sheet steel and had to use a bit of copper circuit board to join the n connector to the element and screw together.

I had to use elbow connectors at the back so the LNA was away from the dish arms in order to adjust, whatever works.

Blow torch soldering is handy as well.

A plastic tube hot glued is then the connection to the LNB holder long enough to adjust focus (which I need to do yet!)

Mine is a bit messy but first effort.

Carl

Edited September 9, 2018 by Carl Reade

[Coto]

@Carl Reade Can you check the SDR's operating temperature and try to reduce the temperature down a bit and see if noise is affected? I'd be curious to see if lowering the temperature a bit can really present reasonable improvements to SNR.

[Carl Reade]

1 hour ago, Coto said:

@Carl Reade Can you check the SDR's operating temperature and try to reduce the temperature down a bit and see if noise is affected? I'd be curious to see if lowering the temperature a bit can really present reasonable improvements to SNR.

Hi Coto unfortunately the dongle is now attached to a heatsink so can't really do a comparison. The chips were running too hot to touch before now they are just above room temp. I have seen them measured at 80deg.

Heat is an enemy with radio so it will help with noise and stability. How much I cannot measure.

I did notice most if not all interference spikes disappeared.

The noise figure at the antenna/LNA sets the benchmark of the system and SNR is more important. The key to these setups are good antennas and the best LNA noise figure you can get hence they are cooled at the antenna in professional radio telescopes.

Carl

[Carl Reade]

Hi all here is an animation made from Excel data from last night from 6 pm to around 12am. its a bit crude as the graph auto fits as the signal gets larger. Its around 60x 5min chunks of data. Will need a better way of displaying but work in progress.

The system I think is working beyond what I expected for a 1.2 m dish.

 

[Coto]

What’s your integration time again? And is this the Milky Way? Have you been able to observe anything beyond the Milky Way with the dish?

Also, is 1.2m the long or the short diameter of the dish? Since it’s an offset, the short diameter is the one that should concern you and should be used to calculate antenna Beamwidth etc. (looking at the dish from your target looks like a circle, with the diameter being the shorter diameter of the ellipse, so the beam should be as symmetrical as a Prime Focus dish).

[Carl Reade]

5 hours ago, Coto said:

What’s your integration time again? And is this the Milky Way? Have you been able to observe anything beyond the Milky Way with the dish?

Also, is 1.2m the long or the short diameter of the dish? Since it’s an offset, the short diameter is the one that should concern you and should be used to calculate antenna Beamwidth etc. (looking at the dish from your target looks like a circle, with the diameter being the shorter diameter of the ellipse, so the beam should be as symmetrical as a Prime Focus dish).

Hi I am averaging in FFT this link should help

http://parac.eu/projectmk4.htm

Yes it's the Milky Way. A drift scan of the arms.

It's a 1.2m dish therefore 1.2m in diameter.

Beamwidth 10-12 deg.

I am currently still fine tuning the system and have not tried to receive anything beyond the galaxy at this time. All the stronger sources outside the galaxy are in the southern hemisphere. If you mean other galaxies I would doubt a dish of this size would be capable, possibly pulsar observations could be achieved. You also have to take into account that radio sources are brighter at other frequencies than 1.4Ghz.

Carl

 

 

[Carl Reade]

Hi all turned the scope towards Cassiopia for a drift scan not sure if Cas a is in there somewhere but was certainly within the beam. Below is an animation and a still in Cassiopia.

Carl

[robin_astro]

1 hour ago, Carl Reade said:

 not sure if Cas a is in there somewhere but was certainly within the beam.

 

Yes I guess it will be tough to pull this out with this beamwidth.  (How about a second dish to do some interferometry ?)  If you are an early riser you could perhaps try for Tau A which is better isolated.  I found this useful chart of comparative signal strengths a while back somewhere on line.  

Robin

[Carl Reade]

1 hour ago, robin_astro said:

Yes I guess it will be tough to pull this out with this beamwidth.  (How about a second dish to do some interferometry ?)  If you are an early riser you could perhaps try for Tau A which is better isolated.  I found this useful chart of comparative signal strengths a while back somewhere on line.  

Robin

Thanks for that Robin. I am pushing my luck with a single dish this size in the garden ?

Yes Taurus is another to look at. I can leave the scope recording 24hrs with the FFT program. Currently I need to get it on a mount so it's easier to set Alt/Az. But its producing good results and lots of data.

Carl

[Carl Reade]

Finally the scope is mounted. I knew the concrete pier would be multiple use. Now I can get some proper settings done.

LNA now boxed and single 40mm pipe used for the feed.

 

[Carl Reade]

Hi all let the scope run from around 17:30 through to 09:30 this morning. This would pick up two sweeps of the galaxy. 

Excel seems to struggle with 3d graphs but managed to put one together. There are two interference bands and in-between them the milkyway peeks. The rear peek is the strongest I have seen so I will check where it passed later.

The gragh is 2.4 MHz wide spectrum over the time period.

 

[Carl Reade]

More playing with data and contour graphs the green areas and all within is hydrogen. Even between the two arm sweeps hydrogen is present.

The centre Freq should read 1.420410Ghz.

Carl

[robin_astro]

On 09/09/2018 at 10:36, Carl Reade said:

 

Setup is Dish with biquad feed 30db 0.5 LNA, 2 x inline sat amps 20db each, bias tee, Filter then SDR dongle.

 

Hi Carl,

What filtering are you doing before the SDR?  Is it a narrow passband filter specifically tuned to 21cm or is it just something to keep the strong broadcast signals out ?

Thanks

Robin

[Carl Reade]

58 minutes ago, robin_astro said:

Hi Carl,

What filtering are you doing before the SDR?  Is it a narrow passband filter specifically tuned to 21cm or is it just something to keep the strong broadcast signals out ?

Thanks

Robin

Hi Robin, it's a specific one made by Radio Astronomy Supplies. I think it was +/- 15Mhz. The label has came off over time. (Photo)

I should also mention the LNA is the SBA1300-1700 from RF Design (G8FEK).

The guy Adam who sells LNA4ALL also now makes H line filters at 20 Euro I have ordered before with him and good to deal with.

http://adsbfilter.blogspot.com/2015/06/hydrogen-line-1420-mhz-filter.html.

Carl

 

Edited October 22, 2018 by Carl Reade

[robin_astro]

Thanks Carl, 

For a filter at that price it hardly seems worth the metal bashing to build one.

(Wow that callsign is almost as old as mine, though I have not been active for many years)

Cheers

Robin

[Carl Reade]

1 hour ago, robin_astro said:

Thanks Carl, 

For a filter at that price it hardly seems worth the metal bashing to build one.

(Wow that callsign is almost as old as mine, though I have not been active for many years)

Cheers

Robin

Yea I know I don't have the test gear to make them even. Haven't been on the air myself in a while. I'm guessing there is a build in the pipeline?

Carl