First Horn Telescope - Problem Solving

[Nathaniel Butts]

Howdy folks.

I've been working for a few weeks on my first horn antenna for H1 emissions measurements.  It's a slightly modified version of the ones from WVU and https://physicsopenlab.org/ but instead of a can for the waveguide I built mine to spec with wood and aluminum foil.  I do, however, have near zero resistance from the aluminum covered Styrofoam pyramid back to the SMA connector, so I reckon it should work.

After a few weeks trying to get software, radio, and computers to cooperate, I have some thing that at least functions.  Now I'm in the trouble shooting phase.

My setup:

* Raspberry PI 4b 8GB

* NOOELEC Smart V5 RTL-SDR

* SAWbird + H1 LNA

* Ted Cline's ezRA (latest from GitHub as of today)

* Waveguide 110mm x 89mm x 250mm

* Horn: Will have to measure again but about 1.0m^2

* RG136 cable

* Location:  36.9N 86.4W

I thought I was seeing some "ok" data, but I'm doubting after my latest scan.  For this one I was measuring zenith and letting the galaxy arms drift by.  I was expecting to see some bumps along the galactic lines on the sigprog image, but I'm far off.  However when I put those lines on the ant RA Avg image, I do see what I think might be a corresponding negative bump around 16:10UTC.

Overall, the data seems messy so I suspect something is wrong.  I'm hoping the collective wisdom might be able to point out the obvious to me.  I did put some undue stress on the rg cable last night, but I don't think it did any damage.  We are about to get a few days of rain here so I'm going to pull it back in the garage and start examining those again.  We do live in a suburb, with LED street lights and lots of houses.  I was hoping the zenith scan would limit any interference, but that's always possible.

Always appreciate any feedback!

Thanks,

Nathan Butts

Wanna-be Astronomer

Bowling Green, KY  USA

[tedcl...@gmail.com]

Hi Nathan,

Thanks for trying ezRA !

I like your antenna.

I think the aluminum foil should work.

Your    ezCon261antRAAvg-marked.png     shows it took about 20 minutes for the receiver to warm up.

As the electronic circuits warm, their gain (amplification) is reduced, and the signal power gets smaller.

The signal power increased as the night grew cold, and then about 1200 UTC (8am local) the Sun appeared and warmed up the air, and the signal power decreased.

Yup, the USB to the SDR is working.

I see you have a "SAWbird + H1 LNA",

      https://www.nooelec.com/store/sdr/sdr-addons/sawbird/sawbird-h1-barebones.html

Is there power to that LNA ?

Is the LNA's LED lit always ?

I think your

      "NOOELEC Smart V5 RTL-SDR"

      https://www.nooelec.com/store/sdr/sdr-receivers/smart/nesdr-smart-sdr.html

would   NOT   power the LNA directly.

But a "SMArTee" version

      Nooelec NESDR SMArTee v2 SDR ... Bias Tee

      https://www.nooelec.com/store/sdr/sdr-receivers/smart/nesdr-smartee-sdr.html

would power the LNA directly.

Swap to a "SMArTee" SDR ?

or

Add power to the LNA power connector ?

    But then, is a DC-Block in the coax needed ?

          https://www.amazon.com/male-female-connector-block-50ohm/dp/B07PZ4K1G4/ref=sr_1_3

    A Bias Tee (also can add that power),

          https://www.minicircuits.com/WebStore/dashboard.html?model=ZFBT-4R2G-FT%2B

          or the much cheaper

          https://www.amazon.com/10MHz-6GHz-Broadband-Frequency-Microwave-Amplifier/dp/B07Q7HKWZZ

can be used as a DC-Block

I am curious what your    ezCon017antMax.png    heatmap looks like.

Post just that one plot ?

Pointing straight up from your Kentucky location should work fine, if given enough time for the Milky Way to pass overheard.

The signal is weakest at the antenna feed, so a short cable (or just single coax adapter) to the LNA is best.

Let us start with the debugging there.

Thanks !

---

Ted Cline
TedCl...@gmail.com

[Nathaniel Butts]

Thanks for the feedback Ted!  I've watched a lot of your videos on SARA youtube.  I like ezRA, I do a fair bit of python programming so it's perfect for me.  I will have questions on the software at some point!  I'm working on a micropython SID detector now after my failed attempt at building a pre-amp and using a python based SuperSID software.

I grabbed the antmax image and have attached it.  Also grabbed some pics of the LNA setup.  I had it setup with a 120mm lead line to the LNA due to space constraints, but a minute ago I swapped it out with a M-M adapter and restarted the file.  It has power from a 2A usb adapter but I will be replacing it with a smartee in the near future due to aforementioned space concerns.

I should see the galaxy sweep across in about 6 hours.  Rain should hold off until tomorrow so maybe I can drag a dry antenna back inside in the morning.  I'm not beyond adding some sort of passive or active cooling to the LNA and/or SDR to help performance....that just sounds fun!

Thanks,

Nathan Butts

Wanna-be Astronomer

Bowling Green, KY  USA

[tedcl...@gmail.com]

Hi Nathan,

Wow, your ezCon017 plot is really boring.

Just some vertical "curtains" on the right (caused by temperature changes ?).

Are you sure the LNA LED lights up ?

Or maybe you are using a "Premium RTL-SDR" with an "E4000 tuner IC" ?

        https://www.nooelec.com/store/sdr/sdr-receivers/nesdr/nesdr-smart-xtr-sdr.html

        https://www.nooelec.com/store/sdr/sdr-receivers/nesdr/nesdr-smartee-xtr-sdr.html
I never had one to test, but I have recently learned ezCol is unable to increase that "E4000 tuner IC" SDR's gain from 0.0 !

That could explain such a boring ezCon017 plot.

Start ezCol, after a minute, what gain is reported in the pretty graphic window, in the text near the top of the 3 stripcharts ?

exCol tries to set the SDR gain to 9999, but in my world, I see that the SDR refuses, and reports the new gain is only 49.6, which works.

Maybe you see "Gain   0.0".

You said you have touched Python.

Maybe you can help fix that "E4000 tuner IC" gain stuck at 0.0 problem ?

A lower library problem ?

We can chat by email.

---

Ted Cline
TedCl...@gmail.com

On Wednesday, July 12, 2023 at 6:53:12 PM UTC-6 Nathaniel Butts wrote:

Thanks for the feedback Ted!  I've watched a lot of your videos on SARA youtube.  I like ezRA, I do a fair bit of python programming so it's perfect for me.  I will have questions on the software at some point!  I'm working on a micropython SID detector now after my failed attempt at building a pre-amp and using a python based SuperSID software.

I grabbed the antmax image and have attached it.  Also grabbed some pics of the LNA setup.  I had it setup with a 120mm lead line to the LNA due to space constraints, but a minute ago I swapped it out with a M-M adapter and restarted the file.  It has power from a 2A usb adapter but I will be replacing it with a smartee in the near future due to aforementioned space concerns.

I should see the galaxy sweep across in about 6 hours.  Rain should hold off until tomorrow so maybe I can drag a dry antenna back inside in the morning.  I'm not beyond adding some sort of passive or active cooling to the LNA and/or SDR to help performance....that just sounds fun!

Thanks,

Nathan Butts

Wanna-be Astronomer

Bowling Green, KY  USA

...

 

[Nathaniel Butts]

Nope, I've got the non-premium.  When I check the gain on ezCol it shows 49.6.

https://www.amazon.com/gp/product/B01HA642SW/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1

I was changing stuff around today and unplugged the LNA while leaving ezCol running.  Took a peak at the data a minute ago and the charts go straight to the floor when it was unplugged.  I'm feeling pretty confident it's getting and keeping power.

However, it does definitely seem to be temperature dependent.  I've had it pointed at Zenith since I got home, waiting for a galactic crossing around 1am CDT.  You can see a definite shift in the data when the sun goes down at 01:00UCT, but I think this is showing as 02:00UCT on the charts.  Is there a compensation for DST in the settings somewhere?

Reviewing some of my cameras, I think the street lights may be coming on at the 02:00 point......this could be a source.  I'll verify on the next non-cloudy day.  There's also a noticeable shift in the spectrum at the same time.  May have to find a quieter place to do these measurements.

Thanks,

Nathan Butts

Wanna-be Astronomer

Bowling Green, KY  USA

[fasleitung3]

Hi Nathan,
My suggestion for a basic check would be to take a single spectrum at a
time when the antenna is pointing roughly towards the galactic plane.
Essentially this is just one of the measurements used for the antMax
diagram. I am not very familiar with ezRA but Ted will certainly be
able to explain how to do it.
With your setup, you should clearly see the hydrogen emission line(s)
in your spectrum. If this does not show up, then there must be
something basic wrong. One question: From the photos it seems that the
probe with the SMA connector is not centred on the rectancular
waveguide. This may be just a matter of the perspective, but it would
be good to know that this is in the right place.
Best regards,
Wolfgang

[Nathaniel Butts]

Thanks for the idea, Wolfgang.  I will be trying to figure out how to look at individual spectrums soon.  If we don't get flooded out I will also just hookup SDR# and check out some waterfalls.  As for the SMA connector, it's right in the middle of the wider side of the waveguide.  I actually drilled the hole large enough that I could center the feed between the walls and also to the rear of the waveguide to the quarter wavelength measurement.  it's screwed together so I'm going to disassemble at some point so I can double check.

I did run some additional testing last night.  Letting it run for a couple of hours to stabilize temps, then running a fresh file to see the effect.  Detection doesn't seem to have been helped, though.  Going to have to start doing some additional studies.

I did put each component on my nanoVNA, included LNA, and got pretty reasonable results.  I haven't tried putting the entire antenna on it, though.  Not sure if that would be useful information or not.

Thanks,

Nathan Butts

Wanna-be Astronomer

Bowling Green, KY  USA

[Alex P]

Hello Nathan,

I will make two suggestions.. echoing Wolfgang,

  1) save individual spectra and plot.

  2) Take longer averages .. 4 minutes of time = a sky drift of  only 1 degree .. with the beam-width of small aperture systems being 20 degrees or so,

'nothing happens'  faster than that and a longer avg will yield smoother, higher S/N plots.  try 4 mins

( I use SDR# w/ IF_Ave plugin   it creates individual *.txt files making easy plotting of single spectra via Excel )

Alex KK4VB

[b alex pettit jr]

Example from a  1m photo umbrella antenna..

[Nathaniel Butts]

Correct me if I’m wrong, but the horn should be conductive, right?  🥴

--
--
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
---
You received this message because you are subscribed to a topic in the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/sara-list/qDRdbDfY1b4/unsubscribe.
To unsubscribe from this group and all its topics, send an email to sara-list+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/687010062.2116998.1689274724377%40mail.yahoo.com.

[Alex P]

Yes, I believe that to be a prime requirement !

Stick the probes into the aluminized film to be sure its just not a surface covering which is non conductive .

Alex

[fasleitung3]

Looking again at the photo and reading the description it seems to me that the waveguide is contructed from wook with the aluminum foil outside. If I am interpreting this corretly then you would have a wooden layer inside the waveguide. Wood is absorbing microwaves at least to some extent, depending on water content. This would be highly counterproductive and destroy antenna performance.

Wolfgang

--
--
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
---

You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/14f1e652-f4eb-4153-bd86-565142c1fa13n%40googlegroups.com.

[Nathaniel Butts]

Looks like I got bamboozled by Lowes. Thought this was aluminized but it appears to be a reflective Mylar coating.  No conductivity even when penetrating the outer layer.  I'm going to run some tests by coating it in cheap aluminum foil to confirm, and also test the affect of the wood in my waveguide.  I'll be doing at least a partial rebuild soon.

Thanks for the pointers guys!  This is definitely being a learning experience.

Thanks,

Nathan Butts

Wanna-be Astronomer

Bowling Green, KY  USA

[b alex pettit jr]

Also

Do not rely on 2" wide aluminum tape .. although within itself it is near 0 ohms.

I found No Continuity between adjacent ( & overlapped ) segments ..

Alex

[Nathaniel Butts]

Thanks for all the suggestions! 

I covered the inside of the horn with aluminum foil and now have a conductive surface from horn to waveguide to sma connectors.  Ran it last night and got what appear to be much better results.  At least more consistent.  I don't see a clear hydrogen signature, yet.  

ezRA uses 31,000 samples/integration which takes my poor little pi 10 seconds to process.  I'm going to bump that up to 744,000 samples/integration tonight to get closer to a 4 minute or 1 degree integration time per Alex P's recommendation, assuming a linear response in time versus samples taken.  If that doesn't provide any results I'll start rebuilding the waveguide sans the wood.

[b alex pettit jr]

Nathan,

Can you convert the ODP to another format .. I use MS Office.

Thanks.

With a  Lenovo notebook I run at 3000 Frames/Sec  ... so 31 K / 10 sec is correct .. but beyond quick look that is WAY too short a time period.

And you need to use software which will capture a spectrum and post test be converted to dB and normalized to give a background of " 0 dB " .

Its the only way data can be compared between systems and between configs.

Alex

--
-

[b alex pettit jr]

When you Do get the system working, define a routine ...

This time of year, Dec +20 dg  RA 06:15 hrs is a good hot target to use.

Stellarium will help you locate it . Currently that area is ~ 1 1/2 Hrs West of the Sun ...

Learn to find it and use this as a reference for changes/mods to the system

Alex

[Nathaniel Butts]

Thank you for the advice Mr. Petit.  I upped the sample rate to 744000/integration which took about 3:55 minutes to process.  Don't know if that means it saw 1 degree of sky or not, but Mr. Cline seems to think so....and he wrote the software (and also been a huge help through this).

I do see a good sized spike at 07:00UTC on the ezCon110 image, which is the close to the time it should have transited, so I may have caught it.  Even with using my iPhone to set the elevation.  I'm having trouble getting SDR# to run on my little Pi, and I can't figure out how to analyze the data on SDR++ yet.  Once I figure that out I'll look at individual spectrums. I'm running it tonight (and maybe through tomorrow night, weather permitting, at 372000 samples/integration

However, I believe I'm going to have to rebuild my waveguide minus the wood to help with permittivity as suggested by Dr. Herrmann.  The power seems to be concentrated +0.75 MHz and doesn't show the expected path of the hydrogen band of the galaxy, as far as I can tell.  Just too much garbage in the signal, I believe.

Lots of work, but I like it!

Thanks,

Nathan Butts

Wannabe Astronomer

Bowling Green, KY  USA

[Nathaniel Butts]

Howdy Folks!

Bit of an update.  Finally saw a good signal from the horn!  From the graph attached, you can see the hydrogen of the galactic arm sweep across.

This was version 1.2 of my build.

1.0 used the styrofoam as bought and a waveguide built from plywood with aluminum foil covering, an SMA connector and a piece of bare copper wire.

1.1 saw the horn covered in aluminum foil when I found out the shiny side of the insulation I bought was a lie and had zero conductivity!

1.2  saw me rebuild the waveguide from styrofoam and aluminum foil.  Dr. Wolfgang was right (surprise, right?) and the wood was absorbing some of the signal.

I ran Ted Cline’s ezRA with 186,000 samples/integration at az=180, el=30.  This number of integrations took about 58 seconds to integrate, so I guess about 0.25 degree of sky per integration….?  Also, I’m in a tight neighborhood inside of an urban setting, so the fact I could see anything tickles me pink.

I could have kept going with this setup, but the styrofoam was way too fragile, especially as many times as I had it apart and retaped it.  I was able to find true aluminized foam insulation boards at Home Depot and have rebuilt the horn.  I wasn’t too keen on buying a can of paint thinner to build the can type waveguide: I don’t really need any thinner and definitely don’t want to waste it.  But I also able to build a waveguide out of galvanized flashing.  I’m testing it now and may do a write up on it soon.

Thank you all for your help with this.  I’m going to work on a quieter spot to setup so I can do a bigger survey.  I’m also going to do some testing to help with the development of ezRA, especially for smaller computers and telescopes.  

[Nathaniel Butts]

Some pics of original setup.

[Nathaniel Butts]

Without any justification whatsoever, I've decided to name my telescope SOKY-RAD, just to convince myself I'm cooler than my kids and neighbors think me to be....

Anyways, have first comparison of v1.2 of my original scope to v2.0.  Night and day difference.

Electronically speaking, V1.2 and V2.0 are the same.  However V1.2 was built with styrofoam coated in aluminum foil, all the way down to the waveguide.  V2.0 utilizes actually aluminized, closed cell foam board and a waveguide built from galvanized steel flashing sourced from Home Depot.  Special note, my soldering iron wasn't big enough to solder the brass sma to the galvanized steel so it is entirely connected using aluminum tape, also from Home Depot.

Thanks for looking!

Thanks,

Nathan Butts

Wannabe Astronomer

Bowling Green, KY