Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

dspira-lessons/forum/antennas/ #3

Open
utterances-bot opened this issue Nov 19, 2020 · 48 comments
Open

dspira-lessons/forum/antennas/ #3

utterances-bot opened this issue Nov 19, 2020 · 48 comments

Comments

@utterances-bot
Copy link

utterances-bot commented Nov 19, 2020 •
edited by PranavSanghavi

Antennas Forum – Digital Signal Processing in Radio Astronomy - Lessons Portal

Lots of lessons

https://wvurail.org//dspira-lessons/forum/antennas/
@nmaron utterances
Copy link

nmaron commented Aug 16, 2021

The instructions for the horn antenna can say to use 5-1/4 inches for dimensions. Shouldn’t the instructions say 5.25 cm (=2-1/16”)?

@jmakous utterances
Copy link
Collaborator

jmakous commented Aug 16, 2021

Yes, of course! Thanks for pointing that out. It's been fixed.

@Tynetenna utterances
Copy link

UK based teacher here finding it hard to find an F style 1 gallon can. Any suggestions where to get one? Thanks

@nmaron
Copy link

nmaron commented Sep 8, 2021 via email

@Tynetenna
Copy link

Thank you. It’ll need soldering presumably? How precise do the dimensions need to be?

@nmaron
Copy link

nmaron commented Sep 8, 2021 via email

@nmaron
Copy link

nmaron commented Sep 8, 2021 via email

@WVURAIL WVURAIL deleted a comment from jmakous Sep 8, 2021
@WVURAIL WVURAIL deleted a comment from GlenLangston Sep 8, 2021
@glangsto
Copy link
Collaborator

glangsto commented Sep 8, 2021 via email

@Tynetenna utterances
Copy link

Thanks Glen and Neil. Glen, did you say there was a table attached to your message, I’m afraid I can’t see it. Thank you very much, Guy

@glangsto
Copy link
Collaborator

glangsto commented Sep 9, 2021 via email

@GlenLangston
Copy link
Collaborator

GlenLangston commented Sep 9, 2021 via email

@PranavSanghavi
Copy link
Member

Hi Glen this service is based of github issues specifically for this thread #3 Within email it seems to strip all attachments and formatting and sends a plain text. any attachments I believe can added from the issue itself via that link.

I shall look into fixing this seems like a github quirk.

best,
Pranav

@ZL4DK utterances
Copy link

ZL4DK commented Sep 23, 2021

I wonder if the horn can be constructed from fine galv "chicken" wire mesh say 15mm? It would obviously need support ribs in the corners. The foam board doesn't seem to be common in my country and mesh would have an advantage in wind.

@nmaron
Copy link

nmaron commented Sep 23, 2021 via email

@ZL4DK utterances
Copy link

ZL4DK commented Sep 23, 2021

Maybe, that would certainly solve the problem. I have a 1 metre diameter solid aluminium dish I might try first. It has a feed that is supposedly OK for 1.4 to 1.8 GHz

@nmaron
Copy link

nmaron commented Sep 23, 2021 via email

@ZL4DK utterances
Copy link

ZL4DK commented Sep 23, 2021

I'll comment on my electronics on the Electronics forum

@ghost utterances
Copy link

ghost commented May 18, 2022

I have built a small radio telescope using the DSPIRA GNU Radio-based software running on a Raspberry Pi, a Nooelec SMART SDR module, a Nooelec Sawbird LNA for the 21 cm hydrogen line, and a satellite antenna, not the DSPIRA horn antenna. The software ran correctly but I was unable to see the 21cm hydrogen bump in the spectrum plot. I am thinking that my problem is either the antenna and/or my location. I am in a densly populated city near Boston and my backyard is closely surrounded by other houses although I have a clear view of the sky. Any ideas on what could be the problem with my system would be appreciated.

Here is a link to the antenna specs. https://www.amazon.com/gp/product/B005M8KU3W/ref=ppx_yo_dt_b_asin_title_o01_s00?ie=UTF8&psc=1

This antenna was used successfully by amateur radio astronomers, see this link:
https://www.rtl-sdr.com/cheap-and-easy-hydrogen-line-radio-astronomy-with-a-rtl-sdr-wifi-parabolic-grid-dish-lna-and-sdrsharp/

Thanks!

Tom Consi

@GlenLangston
Copy link
Collaborator

GlenLangston commented May 19, 2022 via email

@ZL4DK utterances
Copy link

ZL4DK commented May 19, 2022

Hi Tom,
as Glen says get a 50 Ohm termination and test your whole system without the antenna. Once you confirm that is working then you only have the antenna to sort out.
I think I have an antenna very similar to that one a hiding in a shed. I might drag it out in the weekend and give it a try and let you know how it works. However at first glance I see some issues with it. Firstly it is designed for 2.4 Ghz. The dish surface will be OK at 1.4GHz (just a little on the small side) but the feed antenna will be inefficient at collecting energy at 1.4Ghz. The feed antenna will also not have the correct pattern (when used at 1.4GHz) to receive energy just from the dish. You also have a tail of coax (I'm not sure what quality) but coax losses before the LNA will add unwanted noise and reduce your signal. I assume you have the LNA mounted as close to the antenna as possible any haven't introduced any extra coax.
I will let you know how I get on with my antenna with its 2.4GHz feed and if I have time will look at soldering some wire to a sma connector to form a real 1.4Ghz feed and see if I can get this to work better.

@ghost
Copy link

ghost commented May 19, 2022

Glen and ZL4DK,

Thanks for the advice and the speedy respones to my post.
I will try the 50 Ohm termination test to see if my electronics are working correctly.
If the problem is my antenna then I will dissect the feed and try to modify it to be more efficient at 1.4GHz.

Another question: what are the specifications of the DSPIRA horn antenna such as the gain beam pattern? I can't find that information on the DSPIRA website. Such information would be useful to compare the homemade horn to other types of antennas, in particular, dish antennas.

Cheers!

Tom Consi

@glangsto
Copy link
Collaborator

glangsto commented May 19, 2022 via email

@ZL4DK utterances
Copy link

ZL4DK commented May 22, 2022

Hi Tom,
Today in between showers of rain, I did some testing of the antenna I believe is similar to yours. The one I have is very old and was recovered from a 2.4 GHz link used many years ago. It is a cast aluminum dish made of bars spaced about 20mm apart. It is 1 metre in diameter but has the sides cut narrow so is only 600mm in width.
With the original 2.4 GHz feed I could not receive a thing. In fact it was so bad I might as well have had a termination resistor connected instead. I got almost no change in signal no matter where I pointed it. It is possible that water has got into the feed antenna and completely wrecked it.
Anyway next I tried replacing the original feed antenna with a PCB Log Periodic Yagi similar to this one
https://www.wa5vjb.com/pcb-pdfs/LP8565A2.pdf
Mainly because I already had it and thought it might be worth a try. This feed antenna has a sma connector soldered to it so I was able to mount my LNA directly to this (without any coax) and put them both at the focus point of the dish.
This then started to work. I could see some signal level differences between pointing at the ground and pointing at cold sky. However the end result was marginal. possibly due to the fact that I wasn't sure exactly where to put the antenna in order to get it at the focus point and was working on trial and error. Anyway I did see a slight bump at 1420MHz when pointing at the Milky Way during some of these tests but it certainly wasn't conclusive. There are probably better antennas to try at the feed. Maybe a simple 2 or 3 element yagi made of copper wire would be better.

Regards
David

@ghost utterances
Copy link

ghost commented May 24, 2022

Glen & David,
I ran NsfIntegrate24.py as per the instructions in Light Work Memo 28 (rev. 7). The window only had one gain box that was set at a default of 49.6. Keeping that gain I ran a test and got a cold noise of about 140K, the hot noise was 730K but the plot was a straight line with no variation. The median was centered around the cold noise line, which showed variation. I am testing a Nooelec Smartee SDR with a permanent bias T attached to a Nooelec SawBird H1 LNA. The SDR gets very warm, could that be the cause of the high hot noise?
Is there documentation on NsfIntegrate? If so please point me to it.

Thanks!

Tom

@GlenLangston
Copy link
Collaborator

GlenLangston commented May 24, 2022 via email

@ZL4DK utterances
Copy link

ZL4DK commented May 25, 2022

Hi Tom,

most SDR dongles run fairly warm so I think yours will be fine. A high "Hot" temperature reading is a good thing. I am not completely familiar with the NsfIntegrate software but I think those readings suggest everything is fine. Glen will be able to answer this one better.
The next interesting test is to connect your antenna see how much of a difference you get in readings between pointing at the "hot" ground, compared to pointing at "cold" sky. When pointing at the ground avoid any metal objects. When pointing at the sky try and find the coldest (lowest reading) direction.
I had a closer look at the link to your antenna and your feed seems to be slightly different to the one on mine so hopefully it will behave better than what I got.

@ghost utterances
Copy link

ghost commented May 27, 2022

Glen & David,
Success! I think I finally am seeing the 21cm line. I changed to an AirSpy R2 SDR and added two wide-band amplifiers sandwiching a 1.420 GHz Mini Circuits band-pass filter, much like was done in LightWork Memo 29. I used spectrometer_w_cal.grc and did a calibration. The calibration was not so good.
So now, where to go from here. I think my antenna stinks, do you think a Dspira-type horn antenna would improve my signal? I'd ultimately like to map some large radio sources.
Thanks for all your advice and help!
Tom

@ZL4DK utterances
Copy link

ZL4DK commented May 28, 2022

Hi Tom,
Great stuff that you've got it working. I think there are some options to improve your dish by changing the feed antenna. The feed antenna needs to have a wide enough beamwidth to look at all the dish but not too wide that it looks past the edge of the dish. Unfortunately the beamwidth is never sharply defined and all feed antennas are a compromise. For radio astronomy it is best if we compromise on a narrower beamwidth that may not see all of the dish well, but reduces picking up signals from outside of the dish edge as usually these will be noise from the hot ground behind the dish.
Your dish is slightly unusual in that it is not circular. This means we need a feed antenna that has more gain (narrower beamwidth) in one plane than the other.
I am thinking something like the "more simple collinear ant for 1090" about halfway down this page
https://www.qsl.net/g4hbt/1090%20mhz%20ant.htm
This antenna is designed for 1090MHz so the wire lengths need to be scaled back by about 30% for 1420MHz. I would suggest making this out of a single piece of wire and instead of a connector block simply solder a SMA connector at the feedpoint.
You also need to mount a reflector in front of this feed antenna so that it looks towards the dish. A rectangular piece of metal about 300mm by 100mm should be fine. This could be anything metal, but thin aluminium sheet would be best. Even aluminium foil glued to a sheet of cardboard would work. It needs to be spaced about 50mm in front of the collinear feed antenna. The whole lot then needs to be mounted at the focus point of the dish instead of the feed you have at the moment.
I am keen to try this myself and will let you know how it works out.

Regards
David

@glangsto
Copy link
Collaborator

glangsto commented May 28, 2022 via email

@dave290
Copy link

dave290 commented May 31, 2022

Hi All,

I was interested to discover if a very small horn can still get useful signal.  A small version would be easy to weatherproof and possibly automate for rooftop operation.  Without doing any math, I made a truncated horn (15 inch diagonal, 5 inch depth) and attached it to the standard F-gallon can (see attached pictures.)  

I ran this using the RTL-SDR combined with the NOOELEC Sawbird and got signal levels which were quite good, even in comparison with a much larger horn that I typically use.  The attached data file shows signal at galactic longitude=30 deg, galactic latitude=0 deg, averaged for 600 seconds.  The background temperature is higher than when I run the larger horn (300 K compared to 180 K) but the overall peak intensity is similar (~35 K) with both horns.

For an automated version, I'm thinking about automating the elevation angle only, and keeping the azimuth fixed along the north-south meridian. I'm also going to experiment with a light plastic covering over the opening to keep out rain, but pass through signal (?)

Dave

600_second_average_withbkgnd
IMG_1820
IMG_1822

@glangsto
Copy link
Collaborator

glangsto commented May 31, 2022 via email

@ghost utterances
Copy link

ghost commented Jun 5, 2022

David (ZL4DK),

I built a new feed for my parabolic grid antenna. It was simply a wire cut to 5.5cm (1/2 lambda) backed by an aluminum plate 300mm x 100mm placed 50mm from the feed. The wire was mounted parallel to the long axis of the plate. My LNA was attached to the back of the plate and connected to the feed via a 6" SMA cable. A 3' SMA cable connected the signal to a wideband amplifier, a 1.420 GHZ bandpass filter and an AirSpy SDA.

I saw a persistent bump at 1421.3 GHz, not sure if that was the H1 band. Unfortunately, the quality of the signal seemed to be about the same as when I had the original 2.4 GHz feed attached. I did notice, however, that the signal seemed to be stronger when my new feed was oriented parallel to the long axis of the antenna (it is a partial parabola). I mounted the new feed with 1/2" pvc pipe and included a slide coupling so I could vary the distance of the feed to the dish. I think the original feed was at 33cm. Varying the new feed +/- 5cm from that point did not do anything to the signal.

I'd appreciate any comments or suggestion to improve the feed.

Thanks!

Tom

@GlenLangston
Copy link
Collaborator

GlenLangston commented Jun 5, 2022 via email

@ZL4DK utterances
Copy link

ZL4DK commented Jun 24, 2022

Hi Tom,
Today was a new public holiday in my country to celebrate the return of the Matariki (Pleiades) constellation, so a good day to do astronomy. I tried what was effectively a full wave dipole feed. All I can say was it didn't work. I was using the spectrometer_w_cal.grc software. I found doing a Hot-Ground/Cold-sky calibration very difficult as there was little difference in signal between the two. I possibly also had some interference issues that didn't help, however the system temperature from this calibration was well over 500 degrees. I never found any signal from the Milky way. I did get a noise increase when pointing at the sun. You should see noise from the sun fairly easily and could use this to adjust your slider to get the feed at the focus.
I ended the day by rechecking my equipment on my 1.2 metre solid dish. It took me a while to get this working as water had gotten into its feed. However once I drained this I got clean readings from the Milky Way again.
I have seen elsewhere others use a loop feed with your type of dish. However it is probably better to spend the effort building a horn antenna. I am thinking of building a smallish horn myself so that I have a more portable antenna option. We don't have US gallon paint thinner containers in this country but I'm sure I can fabricate something that will work.

Regards
David

@ZL4DK utterances
Copy link

ZL4DK commented Jun 27, 2022

Tom,

You may be interested in the loop feed about halfway down this page;

https://www.qrz.com/db/KK4VB

@ghost utterances
Copy link

ghost commented Jul 7, 2022

Glenn & David,
I have built an RF Ham Design 1m mesh antenna with their "Radio Astronomy" helix feed (1400 - 1427MHz). The antenna is directly connected to a GPIO LNA filtered for the hydrogen line, that feeds an AirSpy SDR. I power the LNA with an external supply, I do not use the AirSpy bias Tee. I also have a DC block between the SDR and the LNA. I am running spectrometer_w_cal.grc on a Raspberry Pi with 8GB ram and the Raspbian operating system. I pointed the antenna at zenith at a time of day (11:50am EDT) when a wisp of the Milky Way was crossing zenith. I used 400ms for short averaging and 10s for long averaging. I did not calibrate the system. Here is what I observed:

The zenith-looking plot was a slanted line running from about 500 (signal strength) at the low freq. end to about 600 at the high freq. end. There was no obvious H1 bump, such as seen in DSPIRA videos, using both short and long integration times.

I pointed the antenna directly at the ground and the overall signal strength went DOWN (!) to about half the zenith level. There were also large bumps in the plot throughout the frequency range. The plot went back to its original shape when I pointed the antenna at zenith.

With the antenna pointed at zenith, if I rotated the feed away from the center of the antenna the overall signal strength went down and then back up to the original level, like an AGC.

Looking over all my past results I cannot say for sure if I have ever observed the H1 line with any version of my set-up. Needless to say this is frustrating because the DSPIRA videos and posts from others readily show the H1 line.

I repeatedly used nsfintegrate100.grc to get a noise figure of my system but I found the program very inconsistent. It sometimes produced nothing after the procedure, the outputs that I have seen certainly did not resemble what is shown in the Memo 28.

I am pretty much at my wits end with this. It seems that every element of my system is suspect. Please give me some suggestions as to how I can untangle this mess.

I can email you plots of the results mentioned above if you give me an address.

Thanks,

Tom

@ZL4DK utterances
Copy link

ZL4DK commented Jul 9, 2022

Hi Tom,
It's time to do the resistor test again. This time do it with the spectrometer_w_cal.gr program. Tell it that the resistor connected to the LNA is the Hot calibration and nothing connected to the LNA is cold calibration. Then look at the system Temp/Gain screen. Hopefully you should see something less than 100K here for just the resistor test. If that checks out OK then it could be you have a strong local interference problem. Put the antenna back on and on the spectrum display you can have a look at the "Unfiltered spectrum with no calibration" screen. Don't be too concerned if you see some rubbish here but it may be worth sharing this screen and seeing how bad this is. The third thing to try is can you find the sun? Your 1m dish should detect this as a signal increase.

my email address is Z***K(at)yahoo.com

Regards
David

@ghost utterances
Copy link

ghost commented Jul 11, 2022

Is there a way to do the "Hot" calibration without having to point the antenna directly at the ground. I find it extremely difficult to point my antenna below 0 elevation.
Can we just use a 50 ohm resistor to get the hot signal?

Thanks,
Tom

@GlenLangston
Copy link
Collaborator

GlenLangston commented Jul 11, 2022 via email

@ghost utterances
Copy link

ghost commented Jul 13, 2022

Glen

I have had a lot more success using SDR#+the IF WIndow Average plugin, compared to spectrometer_w_cal.grc running under Raspbian on a Raspberry Pi 4. I think it is because I cannot sufficiently point the antenna to ground to get a Hot calibration. Even a one meter antenna with its feed struts is a very bulky item and difficult to manipulate. SDR#+plugin uses the 50 ohm dummy load for its background calibration and does not need ground-pointing.

What are the major difference between SDR#+plugin compared to spectrometer_w_cal.grc? From a signal processing point of view are they doing the same thing?

Another contributor to my success is the addition of amplifiers to my signal train:
1 m antenna helical feed > Sawbird LNA > DC block > 2m coax cable > Uputronics broad-band amplifier > Uputronics brosd-band amplifier > Minicircuits H1 bandpass filter > AirSpy 2 SDR. The amplifiers were externally powered and the SDR bias T was off. This configuration was from Lightwork Memo #29.

I think the long-term way to go is to use a GNU Radio-based program running on a Raspberry Pi for its customizability and the potential for remote operation of the antenna. However, for simply seeing the H1 signal, and knowing that I can see it from my observing site, I find SDR#+plugin very useful.

Cheers!

Tom

@whrzg utterances
Copy link

whrzg commented Aug 22, 2022

I have completed the building of the can, horn, LNA, and most of the cradle. I would like to test the LNA before I put the cover on. I read that the best coax cable to use for radio astronomy is 9913; however, this seems very expensive. Is 9913 what I should use or is there another type that is acceptable? Thank you.

@GlenLangston
Copy link
Collaborator

GlenLangston commented Aug 23, 2022 via email

@whrzg
Copy link

whrzg commented Aug 23, 2022 via email

@GlenLangston
Copy link
Collaborator

GlenLangston commented Oct 11, 2022 via email

@ghost
Copy link

ghost commented Oct 11, 2022 via email

@ZL4DK utterances
Copy link

ZL4DK commented Oct 11, 2022

Hi Tom,
you should easily be able to detect the sun with your setup. During the day point your antenna towards the sun and the whole spectrum should rise significantly. You should be able to compare this with the signal level you receive when pointing away from the sun.

@glangsto
Copy link
Collaborator

glangsto commented Oct 11, 2022 via email

@JRichardT utterances
Copy link

Good Afternoon
The DSPIRA site seems to be missing the "Interferometry" lessons. Are they being updated or are they no longer available?
Respectfully,
JRichardT

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests