Cantenna feed dimensions & SDR comparison for H-line

[Ayushman Tripathi]

Hi all,

I'm building a cantenna feed for my 2.4m dish for hydrogen line observation. Using the Changpuak calculator with 150mm diameter can at 1420 MHz, I get:

• Cutoff TE11: 1171 MHz 
• Guide wavelength (Lg): 373.5mm
• Probe position from back: 93mm (0.25 × Lg)
• Probe length: 53mm (λ/4)
• Minimum can length: 280mm (0.75 × Lg)

I have two 19cm containers I can join, giving me either:

• 190mm (single container - below minimum)
• 380mm (both joined - well above minimum)

Can anyone please advise should I use the full 380mm length, or would cutting it down closer to 280mm give better performance, as maybe 380mm is a bit more extra? Does 100mm extra length beyond 0.75 × Lg help?

Also I'm using 6mm copper tube for the probe instead of the recommended 2.6mm - is that fine?

Also, I have both HackRF and Nooelec Smartee SDR (RTL-SDR). From my testing, the Nooelec seems to give better sensitivity, but HackRF has a flatter/more linear spectrum (though filled with its own internal spikes). Nooelec picks up more external RFI. After calibration this shouldn't matter, but on my system the RTL calibrated spectrum is still not as linear as HackRF's. Can anyone advise which would be better for H-line work? (I've HackRF One r10, Original,  2024)

Also, I'm observing that even with my LNA and filter chain, the RTL is still able to pick up FM radio. I tried ferrites on the cables - it helped a bit. I tried placing the RTL in a copper box - it was still able to pick up the music. So my conclusion is it's coming through the antenna coax cable from the Discovery Dish feed. This doesn't happen with HackRF - it can't pick up FM music. Though I've already grounded the HackRF, I tried the same with RTL and it didn't help much.

Thanks!


The 2nd much more spiky one is the HackRF (but almost always more linear than the RTL).

[Alex P]

Hello Ayushman,

Is there an FM radio station nearby ?

As an experiment, try over wrapping ~ 2m of the Coax/SDR/USB cable with aluminum foil.

       ========================= FOIL==============================

--------- Coax--------------SDR------------USB_Cable-----------

       ============================================================

( later today/tomorrow I will run some actual tests with a cantenna using 280mm and 380mm tube lengths )

Regards,

Alex P

[Mike Otte]

To block local FM stations , try what is called an "open Stub".  We used on our repeater because the FM station was on the same tower as the repeater.

Open stub is cut to the frequency of the FM station to notch it out.  Lots of video's on how to make one. the components are just a tee in your feed line and a length of coax hanging off of it.  VNA can be used to tune it. 

--
--
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 visit https://groups.google.com/d/msgid/sara-list/6e15128c-64ab-4a96-8da4-960ad2b16f45n%40googlegroups.com.

--

Mike Otte W9YS

[b alex pettit jr]

Mike,

Can you design one and post it as an example ?

Thanks.

              Alex

[Mike Otte]

Google and I am sure AI does all the work now days!  I hope i am not going to be charged for this!

Text was markup and would not copy so i took screen shots (see attached)

--
--
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 visit https://groups.google.com/d/msgid/sara-list/2129585495.1557541.1765030001932%40mail.yahoo.com.

--

Mike Otte W9YS

[b alex pettit jr]

S21 Spectral Response Tests of the normal 280mm and an extended 390mm tube 

          21cm Tuned Cylindrical Waveguide

==================================================================

 280 mm

280mm In Axis

280mm Out of Axis

=====================================================================================

390mm

390mm in axis

390mm out of axis

=============================================================================

Tx Antenna

CHEERS,

                               Alex P

[Don Latham]

And. . . don't forget to seal the open end of the coax!

Don

---

From: "Mike Otte" <mike....@gmail.com>
To: "sara" <sara...@googlegroups.com>
Sent: Saturday, December 6, 2025 7:28:12 AM
Subject: Re: [SARA] Cantenna feed dimensions & SDR comparison for H-line

To view this discussion visit https://groups.google.com/d/msgid/sara-list/CA%2BG65tT6AeK%2B6a5KuU8kvZp35-%3DUa0-KPxTwCV-pHcrK2JhrkA%40mail.gmail.com.

--

------------
Don Latham
PO Box 404,
Frenchtown, MT, 59846
406-626-4304

[Don Latham]

Nice! I clean forgot about using those log periodic antennas as test sources!!!!!!

Don

---

From: "sara" <sara...@googlegroups.com>
To: "sara" <sara...@googlegroups.com>
Sent: Saturday, December 6, 2025 7:58:56 AM
Subject: Re: [SARA] Re: Cantenna feed dimensions & SDR comparison for H-line

--
--
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 visit https://groups.google.com/d/msgid/sara-list/1698371448.830526.1765033137776%40mail.yahoo.com.

[Marcus D. Leech]

On 2025-12-06 08:53, Mike Otte wrote:

To block local FM stations , try what is called an "open Stub".  We used on our repeater because the FM station was on the same tower as the repeater.

Open stub is cut to the frequency of the FM station to notch it out.  Lots of video's on how to make one. the components are just a tee in your feed line and a length of coax hanging off of it.  VNA can be used to tune it.

I'll note that an open 1/4L stub has a response that repeats on odd multiples of the design frequency.  This may be fine in some applications, and unpleasant in others.

To view this discussion visit https://groups.google.com/d/msgid/sara-list/CA%2BG65tSe5XQWAThajWY0bFWaTLzbAOoB5D6LKLySiad8LiBuWw%40mail.gmail.com.

[Andrew Thornett]

My main advice is cut the monopole probe a little bit too long and then tune it once cantenna made using a NanoVNA and slowly cutting tiny bits off end monopole until maximum VSWR (at least less than 2:1).

Sent from Outlook for Android

---

From: sara...@googlegroups.com <sara...@googlegroups.com> on behalf of Marcus D. Leech <patchv...@gmail.com>
Sent: Saturday, December 6, 2025 6:22:35 PM
To: sara...@googlegroups.com <sara...@googlegroups.com>
Subject: Re: [SARA] Cantenna feed dimensions & SDR comparison for H-line

 

To view this discussion visit https://groups.google.com/d/msgid/sara-list/a017c2cc-478a-49cb-b43b-59edc57a8ece%40gmail.com.

[Ayushman Tripathi]

Hi all,

Thanks a lot.

Hi Alex,

Yes, I wrapped the full coax cable with aluminum foil just now, and the FM broadcast stopped. As you can see in the image, there is no music now. Thanks a lot. I also placed the RTL in the copper box again. Yes, it seems I do have an FM tower 6.5 miles from my place.

Sure, Mike, Don - I'll try an open stub soon and will report the observations.

Thanks, Alex, for testing both cantennas; both look very similar at 1420 MHz. Do you think 6mm for the probe is fine for the cantenna?

Okay, Andrew, I'll do that. Thanks for the advice, though I don't have a NanoVNA yet. Airspy SDR is on the list next :). I will try to get a NanoVNA after that.

Also, I got my 2.4m dish; I'll set it up by the first week of January with the Cantenna.

[b alex pettit jr]

I wrapped the full coax cable with aluminum foil just now, and the FM broadcast stopped. 

How About That !  

Good News.

6mm is okay, but the 0.95 * WL/4 rule of thumb will be a bit too long .. 

The surface area at the open end will cause a  'capacitance hat' effect of a few mm.

SO, rather than 210mm/4 * 0.95 = 49.8 mm yours may wind up being .. 47 mm 

Alex

==========================================================================

On Saturday, December 6, 2025 at 03:38:54 PM EST, Ayushman Tripathi <ayushmantr...@gmail.com> wrote:

Hi all,

Thanks a lot.

Hi Alex,

Yes, I wrapped the full coax cable with aluminum foil just now, and the FM broadcast stopped. 

As you can see in the image, there is no music now. Thanks a lot. 

I also placed the RTL in the copper box again. Yes, it seems I do have an FM tower 6.5 miles from my place.

[Andrew Thornett]

Hi Don

What do you mean seal the open end of the coax?

Andy

Sent from Outlook for Android

---

From: sara...@googlegroups.com <sara...@googlegroups.com> on behalf of Don Latham <d...@montana.com>
Sent: Saturday, December 6, 2025 6:18:31 PM
To: sara <sara...@googlegroups.com>

To view this discussion visit https://groups.google.com/d/msgid/sara-list/280444314.11019813.1765041511736.JavaMail.zimbra%40blackfoot.net.

[b alex pettit jr]

A bit more elegant solution than  Al  Foil will be to use Copper Braid

[Don Latham]

just put some silicone seal or other stuff to keep water out. Water loves coax due to capillary action.

---

From: "sara" <sara...@googlegroups.com>
To: "sara" <sara...@googlegroups.com>
Sent: Saturday, December 6, 2025 2:10:12 PM

To view this discussion visit https://groups.google.com/d/msgid/sara-list/FRWP194MB275249358247D22088228767FFA4A%40FRWP194MB2752.EURP194.PROD.OUTLOOK.COM.

[Ayushman Tripathi]

Okay, thanks.

[Alex P]

Please ReRead Marcus' comment :

I'll note that an open 1/4L stub has a response that repeats on odd multiples of the design frequency.  

This may be fine in some applications, and unpleasant in others. 

IF you have removed the RFI with the shield, I'd consider the problem Solved.

[Ayushman Tripathi]

Hi Alex,

Yes, thanks. That's why I'm not trying the Stub method. As for now, the FM is gone. The next task is finding the source of Harmonic RFI.

Will be much better at the remote site with the bigger dish, as there's very little RFI there, only cell towers, FM, no WiFi routers, no PCs, LEDs, etc.

Thanks!

[Ayushman Tripathi]

Hi, I found the source of this harmonic RFI; it was a nearby monitor, and it had a very big effect on the RFI. 

[b alex pettit jr]

YES, they can create a Lot of RFI ..  Typically,  the least radiation is Edgewise to the screen ..

so rotating it 90 degrees may have a dramatic effect !

Alex

[Ayushman Tripathi]

Hi,

Using my Discovery Dish feed (dual QPL9547 LNA, ~36 dB total gain) with RTL-SDR for hydrogen line observations. I noticed that:

• At low RTL gain (0-10 dB): DC spike and a nearby RFI are clearly visible, and the spectrum has good dynamic range.

• At high RTL gain (40+ dB): DC spike disappears completely

My question: Does this disappearance of the DC spike/RFI mean I'm overloading/saturating the ADC?

In other words, should I set the gain to the highest level where the DC spike remains visible?

Attached images show the spectrum at different gain settings.

Is there anyone who made it work with DD feed, and can you please advise what settings you used? I'm trying to find my best settings, using both SDR# and IF AVG (Windows) & also the VIRGO/DSPIRA Spectrometer on Linux.

Soon I'll be setting up a bigger dish with a Nooelec LNA, but still I want to make this setup work, as I have two systems: an HackRF/RTL with a DD Dish Feed and a WiFi Grid.

Also Airspy Mini with 2.4m Dish and Nooelec Sawbird.

I will run both the system at different locations, the bigger dish for SETI, and this one for H-line observations.

Any advice will be appreciated.

Thanks,

[b alex pettit jr]

The mid-band spike is an artifact of that SDR  

Use Max Gain

I use max gain on the AS mini also ..

Here is info on the Air Spy mini that references the problem

Alex

=======================================================

On Tuesday, December 9, 2025 at 06:01:25 PM EST, Ayushman Tripathi <ayushmantr...@gmail.com> wrote:

Hi,

Using my Discovery Dish feed (dual QPL9547 LNA, ~36 dB total gain) with RTL-SDR for hydrogen line observations. I noticed that:

[Stephen Arbogast]

I am using the  Discovery Dish with AirSpy R2,  GNU Radio (not  DSPIRA)  for  data  collection then  I use some developmental ezRA  software  from  Ted  Cline  (still in test)  for  post  processing.  I run  the gain of the  AirSpyR2  at  full (45).  Here is some  very preliminary  results  for a  short 5 hour  drift  scan...  hope this  helps  for  comparison.

[Ayushman Tripathi]

Hi Alex,

Sure, Thanks.

Also, can you please advise best ways to calibrate my system? Currently, I'm using cold sky calibration by pointing the dish towards the cold sky.

But as you can see from the calibrated spectrum, it looks nothing like the linear spectrum in the blog posts using the same setup (apart from the different LNA).
This is where I'm pointed at right now (Perseus Arm), not the galactic centre.

Thanks.

[Ayushman Tripathi]

Hi Stephen,

Thanks a lot! , Good to hear someone has done it with DD Feed!

And I'm not the first one trying. ;)

Thanks, your data helps a lot!

[b alex pettit jr]

THAT plot is reasonable .. and will work fine for a Background Corr.

I purchased the DD feed and will be making tests over the next week or so.

I've fab'd a small 'box' of Conductive Foam as a Background Correction reference ... results TBD..

Alex

========================================================================

On Tuesday, December 9, 2025 at 06:48:14 PM EST, Ayushman Tripathi <ayushmantr...@gmail.com> wrote:

Hi Alex,

Sure, Thanks.

Also, can you please advise best ways to calibrate my system? Currently, I'm using cold sky calibration by pointing the dish towards the cold sky.

But as you can see from the calibrated spectrum, it looks nothing like the linear spectrum in the blog posts using the same setup (apart from the different LNA).
This is where I'm pointed at right now (Perseus Arm), not the galactic centre.

Thanks.

[Ayushman Tripathi]

Okay Alex, thanks. 

[Ayushman Tripathi]

Hi,

I've fixed my system and am now getting much better results. These plots were captured while observing the Cassiopeia/Perseus Arm region.

The key improvement was adding aluminum foil shielding to the back of my dish, fully covering it to block RFI pickup from behind. Within an hour of making this change, I started seeing clear improvements.

My next step is to optimize pointing accuracy. Currently I'm using a phone compass, but it jumps around even after calibration—magnetic compasses aren't stable near metal structures. I'm looking into better solutions for more precise azimuth measurement.

For my larger 2.4 m dish (once it's set up), I plan to use drift scanning rather than daily repointing.

The last plot is from a few hours earlier when I was not pointing at the Cassiopeia region.

Thanks!

[andrew....@googlemail.com]

Drift scanning works well for hydrogen line work, even with small antennas or dishes and I would strongly recommend starting with that. With regards to pointing/compass, make line on the ground before putting your telescope down, and measure the direction of the line to define azimuth and then put telescope down aligned with that. Azimuth then does not change and best to leave the telescope out all the time so no errors due to movement. If you can’t do this then put some markers on the ground so that the telescope can be aligned exactly the same way each time you put it out.

 

For altitude measurements, use a digital inclinometer on the same part of the antenna/dish each time.

 

Both compasses and digital inclinometers are available free of charge as apps on your phone – but they are also so cheap to buy that sometimes easier to get at least the inclinometer as separate device (but not necessary).

 

The effect of metal on compasses can not be removed by buying expensive devices – only by measurement without the metal present, so you need intermediate markers.

 

If you really want to be able to point the telescope in different directions or have projects that require this ability then you need to either use a compass some distance from the metal structure (e.g. on a long boom), or use encoders or other means of identifying angle changes without using a compass.

 

I realise, Ayushman, that you have already done a lot and so moving to these more advanced areas is relevant to you, but for the sake of other readers who might be starting off, I would emphasise the importance of starting simply and with a smaller dish or antenna and building up, rather than jumping in at the deep end – good chance that will lead to failure and frustration, and might even lead newcomers to giving up on radio astronomy completely.

 

………..For all those people who are now raising an eyebrow at the above, thinking of how they advised me to do the same thing, and I ignored them and made a fool of myself, my only response is…..I know, and I was wrong!!!

 

Andy

 

 

From: sara...@googlegroups.com <sara...@googlegroups.com> On Behalf Of Ayushman Tripathi
Sent: 12 December 2025 22:55
To: Society of Amateur Radio Astronomers <sara...@googlegroups.com>
Subject: Re: [SARA] Cantenna feed dimensions & SDR comparison for H-line

 

Hi,

I've fixed my system and am now getting much better results. These plots were captured while observing the Cassiopeia/Perseus Arm region.

The key improvement was adding aluminum foil shielding to the back of my dish, fully covering it to block RFI pickup from behind. Within an hour of making this change, I started seeing clear improvements.

My next step is to optimize pointing accuracy. Currently I'm using a phone compass, but it jumps around even after calibration—magnetic compasses aren't stable near metal structures. I'm looking into better solutions for more precise azimuth measurement.

For my larger 2.4 m dish (once it's set up), I plan to use drift scanning rather than daily repointing.

The last plot is from a few hours earlier when I was not pointing at the Cassiopeia region.

Thanks!

On Tuesday, December 9, 2025 at 8:26:42 PM UTC-5 Ayushman Tripathi wrote:

Okay Alex, thanks. 

On Wednesday, December 10, 2025 at 5:29:05 AM UTC+5:30 b alex pettit jr wrote:

THAT plot is reasonable .. and will work fine for a Background Corr.

 

I purchased the DD feed and will be making tests over the next week or so.

 

I've fab'd a small 'box' of Conductive Foam as a Background Correction reference ... results TBD..

 

Alex

 

========================================================================

 

On Tuesday, December 9, 2025 at 06:48:14 PM EST, Ayushman Tripathi <ayushmantr...@gmail.com> wrote:

 

 

Hi Alex,

Sure, Thanks.

Also, can you please advise best ways to calibrate my system? Currently, I'm using cold sky calibration by pointing the dish towards the cold sky.

But as you can see from the calibrated spectrum, it looks nothing like the linear spectrum in the blog posts using the same setup (apart from the different LNA).
This is where I'm pointed at right now (Perseus Arm), not the galactic centre.

Thanks.

 

--

--
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 visit https://groups.google.com/d/msgid/sara-list/774cfdd9-48f2-4110-9e91-7350fb929133n%40googlegroups.com.

[Ayushman Tripathi]

Hi Andy,

Thanks for the detailed advice!

Sure, I'll try drift scanning for my setup. The marker/line idea for azimuth is clever - much better than fighting with a compass near metal.

I already have a digital inclinometer, so elevation is sorted. It's just azimuth that's been giving me trouble.

And yes, starting simple was key - the foil shielding trick only came after lots of trial and error!

Thanks,
Ayushman

[Andrew Thornett]

Sounds like brilliant idea and fits in with Alex's advice to remove ground noise which is relevant to all of us - so you've demonstrated how effective that is - timely!

Andy

Sent from Outlook for Android

---

From: sara...@googlegroups.com <sara...@googlegroups.com> on behalf of Ayushman Tripathi <ayushmantr...@gmail.com>
Sent: Friday, December 12, 2025 11:50:30 PM

To view this discussion visit https://groups.google.com/d/msgid/sara-list/eb5c9bc5-e86e-415a-9a0f-d40ae18cd5f2n%40googlegroups.com.

[b alex pettit jr]

Ideally, you want to keep the dish pointed as 'high' in the sky as possible ..

Meaning, point along the Meridian & let the sky drift by

[Steve Berl]

Magnetic compasses have to be corrected for both variation (difference between true north and magnetic north at your location and date), and deviation (error introduced by local magnetic fields from metal objects nearby). It's hard to do that without a true reference direction of some kind.

An accurate way to get an azimuth reference is use the azimuth of the sun, which you can get from any one of several online calculators. One example is https://gml.noaa.gov/grad/solcalc/azel.html

Alternately you can find the time of local apparent noon for your date and location, and at that moment the azimuth of the sun is exactly 180°, due south.

Steve

--
--
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 visit https://groups.google.com/d/msgid/sara-list/828227956.201160.1765584260516%40mail.yahoo.com.

--

-steve

[b alex pettit jr]

If you pick a fixed location for your setup,

Go out on a clear night, use a star chart pgm to find where :Polaris is WRT True North, and

put two reference stakes in the ground pointing to that position for daytime alignment.  

( Without the planetarium software you'll be not more than  ~ 1/2 Degree from True N )

[andrew....@googlemail.com]

How important is this Steve for small dishes – given the beamwidth of small dishes the exact azimuth pointing direction is perhaps not vital? Or am I wrong? Also, if we are using the same direction with drift scans, then slight errors in pointing can be corrected later in the analysis?

 

…Just thoughts – open for discussion.

 

Andy

 

From: sara...@googlegroups.com <sara...@googlegroups.com> On Behalf Of Steve Berl
Sent: 13 December 2025 19:35
To: sara...@googlegroups.com
Subject: Re: [SARA] Cantenna feed dimensions & SDR comparison for H-line

 

Magnetic compasses have to be corrected for both variation (difference between true north and magnetic north at your location and date), and deviation (error introduced by local magnetic fields from metal objects nearby). It's hard to do that without a true reference direction of some kind.

To view this discussion visit https://groups.google.com/d/msgid/sara-list/CAB4gGneN4nFeG_gmjLZ%3D5WnyU_y1tQQbDs_pdgJnAJQ8OpRm4w%40mail.gmail.com.

[andrew....@googlemail.com]

Of if you want to be really accurate then polar align an astrophotography mount and use that to get an absolute direction – but again how accurate do we need to be for various sizes of dishes, e.g. 100cm, 150cm, 200cm, 300cm dishes?

Andy

 

From: 'b alex pettit jr' via Society of Amateur Radio Astronomers <sara...@googlegroups.com>
Sent: 13 December 2025 19:59
To: sara...@googlegroups.com
Subject: Re: [SARA] Cantenna feed dimensions & SDR comparison for H-line

 

If you pick a fixed location for your setup,

--

--
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 visit https://groups.google.com/d/msgid/sara-list/1853444121.287245.1765655953637%40mail.yahoo.com.

[b alex pettit jr]

1) with a small dish with beamwidths of 10-15 degrees, not that important ..

2) how would you correct in the analysis if you don't know where it was originally pointed ?

========================================================================

[b alex pettit jr]

I Did Polar align my equatorial mount to True N and have alignment pipes in the ground for the tripod feet to return the mount to the identical location each time,

but that does not mean the Antenna beam matches the axis of the mount..

[Steve Berl]

How close do you need to get to a reference true north? As others have pointed out, it depends.

Depending on where you live, magnetic variation (the location and date dependent difference between magnetic north and true north) can be as big as 30°, so you definitely need to correct for that. 

If you use the included iPhone compass app, there is a setting to show either true north or magnetic north and for astronomy purposes you definitely want true north.

I'm not sure what the default setting is, so check the Settings on your phone.

Visual light astronomers don't typically use the sun, but use Polaris instead, because they typically are out at night and well, it's hard to see the sun then.

But during the day, the easiest thing is going to be to go out at your local apparent noon (from the online calculator, and an accurate time of day clock) and look at the shadow of a stick in the ground. The shadow will point true north. Make sure the stick is as vertical and straight as possible.

--
--
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 visit https://groups.google.com/d/msgid/sara-list/1549920466.285365.1765656681659%40mail.yahoo.com.

--

-steve

[andrew....@googlemail.com]

You could see how the plots align with what it should look like and adjust the angle so they overlap…..or is that bad practice?

 

From: 'b alex pettit jr' via Society of Amateur Radio Astronomers <sara...@googlegroups.com>
Sent: 13 December 2025 20:07
To: andrew.thornett via Society of Amateur Radio Astronomers <sara...@googlegroups.com>
Subject: Re: [SARA] Cantenna feed dimensions & SDR comparison for H-line

 

1) with a small dish with beamwidths of 10-15 degrees, not that important ..

--

--
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 visit https://groups.google.com/d/msgid/sara-list/1238281216.284667.1765656446327%40mail.yahoo.com.

[andrew....@googlemail.com]

Good point! I’ve noticed that too on my RTs!

 

From: 'b alex pettit jr' via Society of Amateur Radio Astronomers <sara...@googlegroups.com>
Sent: 13 December 2025 20:11
To: andrew.thornett via Society of Amateur Radio Astronomers <sara...@googlegroups.com>
Subject: Re: [SARA] Cantenna feed dimensions & SDR comparison for H-line

 

I Did Polar align my equatorial mount to True N and have alignment pipes in the ground for the tripod feet to return the mount to the identical location each time,

--

--
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 visit https://groups.google.com/d/msgid/sara-list/1549920466.285365.1765656681659%40mail.yahoo.com.

[andrew....@googlemail.com]

Not around here – you can hardly ever see the sun during the day even in the summer!!

 

From: sara...@googlegroups.com <sara...@googlegroups.com> On Behalf Of Steve Berl
Sent: 13 December 2025 20:55
To: sara...@googlegroups.com
Subject: Re: [SARA] Cantenna feed dimensions & SDR comparison for H-line

 

How close do you need to get to a reference true north? As others have pointed out, it depends.

To view this discussion visit https://groups.google.com/d/msgid/sara-list/CAB4gGndkR%2B5oQfVKeq-UrwhZp22eT3V-4wj3LaMFkh5X7ZU2LQ%40mail.gmail.com.

[Ayushman Tripathi]

Thank you Alex, Steve, and Andrew.

For azimuth, I compared readings across three different phones to rule out magnetometer issues, which helped. I've also been making small adjustments (2-8 degrees) based on trial and error while observing the results.

I've marked the dish position using nearby buildings and towers as reference points for consistent alignment.

I'm using the drift scanning approach rather than frequent repointing - just minor tweaks occasionally. The elevation is adjustable from 20° to 45°, currently fixed at 34°.

I'm using Virgo software running natively on a Raspberry Pi 5. I'm testing this setup on smaller dish before deploying it with my larger 2.4m dish at a remote location - running locally should avoid the latency issues that come with RTL-TCP over network. These observations were captured during 5 and 30 minute integrations. 

I'm also working on a cantenna feed for the larger dish and will share progress soon. 

Here's the plot I recorded today:
Thanks,
Ayushman

[Ayushman Tripathi]

All plots are from the same Cassiopeia/Perseus Arm region.

[b alex pettit jr]

" Cassiopeia "  is in the Dec +60 Deg RA 01:00 Region

Not a particularly bright area  for system evaluation

=================================================================================================

[Ayushman Tripathi]

Hi Alex,

Yes, from my balcony currently only the Perseus arm region is visible due to surrounding buildings - but this won't be an issue with my bigger dish at the remote location where there's a clear horizon.

Based on my plots, is there anything you'd suggest I improve or try differently? Any feedback would be helpful!

Thanks

[b alex pettit jr]

Hello Ayushman,

Honestly, I can not eval the data as I am totally unfamiliar with your software ..

Alex

=====================================

[b alex pettit jr]

With regards to Software :

I have run linearity , calibration, and signal range tests on the AirSpy SDR# Studio with the Kaminski IF_avg plugin software.

They function correctly. 

I have not seen similar tests performed on other software....

Alex

==========================

[Ayushman Tripathi]

Hi Alex,

Sure, no problem! I'll try capturing some data using SDR# with the IF Average plugin and share that tomorrow. I found a version of the plugin that has calibration save/read options, which should help.

Thanks,

Ayushman

[b alex pettit jr]

Hi Ayushman,

The Huge advantage of IF-Avg is its capability of background correcting all data As it is taken ..

Meaning, you can see small details in the Hline spectrum and not have the display overwhelmed with the Giant filter pass-band M Curve !

( this is probably Not the same data, but it illustrates the problem )

And, Yes, if you can capture a clean background file,  that might help.

Regards,

        Alex

========================================================================================

[Ayushman Tripathi]

Hi Alex,

Yes, thanks. I'm trying the SDR# IF Average plugin but running into a problem. I did cold calibration by disconnecting the antenna from the feed, and it acquired the background fine - showed "Corrected background!" with a flat, leveled spectrum. But as soon as I reconnected the antenna, the spectrum shot above what the IF Average window can display even with plugin's max gain. The same calibration method works fine on Virgo.

Settings: SDR# v1.0.0.1920 via RTL-SDR TCP, IF Average v2.8, RTL Gain max, FFT 1024, Intermediate Average 1000, Plugin Gain 400 (max), Level 1000, Dynamic Averaging 902000, Center 1420.010 MHz, Sample Rate 2.048 MSPS and 3.2 MSPS

The same calibration method with the same setting works perfectly on Virgo—it shows a clear H1 bump.

Can you please advise if I should reduce RF gain before calibration? Or record anyway since it might just be a display scaling issue? Any other settings to adjust?

I tried the same thing twice today, and it led to the same issue.

Screenshots attached.

Thanks

[b alex pettit jr]

The Level & Gain are Display parameters only and have no effect on the data files.

Adjust the (Display) Gain as required = less than 400

On Sunday, December 14, 2025 at 04:09:40 PM EST, Ayushman Tripathi <ayushmantr...@gmail.com> wrote:

Hi Alex,

Yes, thanks. I'm trying the SDR# IF Average plugin but running into a problem. I did cold calibration by disconnecting the antenna from the feed, and it acquired the background fine - showed "Corrected background!" with a flat, leveled spectrum. But as soon as I reconnected the antenna, the spectrum shot above what the IF Average window can display even with plugin's max gain. The same calibration method works fine on Virgo.

Settings: SDR# v1.0.0.1920 via RTL-SDR TCP, IF Average v2.8, RTL Gain max, FFT 1024, Intermediate Average 1000, Plugin Gain 400 (max), Level 1000, Dynamic Averaging 902000, Center 1420.010 MHz, Sample Rate 2.048 MSPS and 3.2 MSPS

The same calibration method with the same setting works perfectly on Virgo—it shows a clear H1 bump.

Can you please advise if I should reduce RF gain before calibration? Or record anyway since it might just be a display scaling issue? Any other settings to adjust?

I tried the same thing twice today, and it led to the same issue.

Screenshots attached.

Thanks

[b alex pettit jr]

Ayushman,

You are using the Disco Feed on the Nooelec Grid Antenna ?

Alex

[Ayushman Tripathi]

Hi Alex,

Yes, these observations are from a ~1 meter WiFi dish (with the back wrapped in 2 layers of aluminum foil) + Discovery Dish Feed.

I think the Calibrated Spectrum Display of IF Avg. somehow isn't usable for my observation today. I noticed the higher the plugin display gain, the more of the spectrum it shows. Earlier, without good calibration, I was able to view the calibrated spectrum at around 335 gain—if the display was full, I increased gain to see more of the spectrum above the noise floor. But now it's at maximum (400), and still the top of the spectrum isn't visible. Maybe I'll need to reduce the RTL gain itself? for the display to show live updates.

Anyway, I'll still record the data as a test since even if the display doesn't work properly, the saved data should be fine.

Also making progress on my cantenna feed for the bigger dish—got the 2 containers welded today and kept the inner surface smooth.

Thanks

[b alex pettit jr]

The Discovery Dish Feed is a problem :  You need to either use a non Hline region of the sky 

OR ( I should get around to testing this in the next few days ) an earth noise simulator with conductive foam

Keep the SDR gain at max, just teak the IF-avg display.

Alex

--
--
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/oVp2CXRN5XQ/unsubscribe.
To unsubscribe from this group and all its topics, send an email to sara-list+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/sara-list/1160ddf4-25ae-4dff-956f-3cd6e57657abn%40googlegroups.com.

[Ayushman Tripathi]

Hi Alex,

Thanks. Yes, for calibration I wrapped the whole feed tightly in aluminum foil (2 layers) to block it from picking up anything, as it can easily pick up RFI otherwise.

After doing that calibration, at least I'm able to see a clean H1 bump even in weak regions like Cassiopeia with just 5 minutes of observation.

Thanks.

[Robert Meade]

Can you share how you plan on using foam to simulate noise only input with no HI line? A hat on the antenna feed or just putting absorber foam in front of the whole aperture? 

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 visit https://groups.google.com/d/msgid/sara-list/739994732.401798.1765755081511%40mail.yahoo.com.

[b alex pettit jr]

Give me a day or two ... probably enclosing the plastic feed box = a variation on this cover I developed for Loop and Cantenna .Feeds

Alex

========================================

[Ayushman Tripathi]

Hi, Alex.

Thanks. Yes, I found the optimal distance from the feed box to the base of my dish to be ~278 mm, but I'll test more, it will depend on the specific WiFi dish.
I'll run some tests with SDR# soon and share the results.

I'm getting good results with Virgo so should be same with SDR# IF Avg.

Yes, I used the foil like this for callibration (2 layers):

Thanks.

On Tuesday, December 16, 2025 at 8:26:06 AM UTC-5 b alex pettit jr wrote:

Discovery 1420 MHz Feed on Nooelec 1m WiFi Dish

Some preliminary info .... More To Come

( will put this together as a new post )

TakeAways :

1) The optimal spacing from the dish to the center of the Feed Box = 28 cm ( will re-verify )

2) Aluminum Foil equals the performance in creating an IF_avg Background Correction File as does Conductive Foam

3) I was Impressed by the Exceptional Performance of the  Discovery 1420 MHz Feed w/ integral LNA  on the Nooelec 1m WiFi Dish

4) I was Unimpressed by the Dismal Performance of the Nooelec 1420 MHz Feed w/ Nooelec SAWbird LNA on the Nooelec 1m WiFi Dish

         Happy Holidays,

              Alex P

[Stephen Arbogast]

I  don't  use Virgo..   it is  outdated.   I don't use  SDR# since I cannot  see   how it  works....    with    IF  Averaging..    What  is   IF Averaging?

[Stephen Arbogast]

From   Marcus  Leech    a simple   single  pole infinite impulse   response  filter works  for me......

[Ayushman Tripathi]

Hi Stephen,

Yes, I noticed the Virgo GitHub has been inactive for about 3 years. Despite that, I'm still getting good results with it. I did try the IF Average plugin in SDR#, and while I could see the H-line bump, it wasn't nearly as clear as what Virgo produces—likely due to Virgo's median filtering and calibration division approach.

With Virgo, I can detect the H-line in just 1 minute of integration even on weaker regions. The SDR# results might also be affected by my RFI environment—I have quite a lot of interference here, and much of it was visible in SDR#. I'll try to address that later this week. The setup differences could also play a role—I'm running SDR# through RTL-TCP, whereas Virgo runs directly on the Raspberry Pi 5.

Do you use a fully custom GNU Radio pipeline? I'm planning to deploy this setup at a remote location with automated daily observations uploading to the cloud, so I'm evaluating my options. For now, Virgo seems like the only viable option for remote unattended operation—even if the internet goes down or becomes slow, it will keep recording locally and upload later. Ideally, I'd like to use something that can easily be processed into a turboSETI-compatible format (Filterbank .fil or .h5).

For now, I'd rather not experiment with building custom Python or GNU Radio scripts myself—I could easily make a mistake that affects data accuracy, and I wouldn't want to discover that months later. So I'm sticking with known tools for the time being, though I did make a simple GUI wrapper for Virgo, but it's not fully tested yet. Down the line, I'd like to build something more custom, possibly by extending the Virgo codebase.

The attached plot is from Cassiopeia—a weaker region—captured today in just 5 minutes.

Thanks

[Stephen Arbogast]

Hi Ayushman ,

I  like  having access to the source code so I can  see what is  being done.  GNU Radio  flow graphs  make  this even easier.    No, I have not  written  any software from scratch  for  collecting  H-Line  signals.  I make  use of the work of others, open  source  with permission of course,  and  I modify it for what  I  want or need.  All  of the  software  I have tried works fine....     there  are many ways  to achieve  the goal.  The important thing  is the Physics!!!   (for me)

I understand that some people  don't  want to dig into the software...  that is ok.

Have  fun.

[Stephen Arbogast]

Ayushman,

 It is  just scaling  your plots of collected data.. you  can make them look  any way  you want......

[b alex pettit jr]

Some Software

Corrects for band pass filter Gain non-linearity

Removes for  LSR Doppler Frequency ( Velocity ) error

Normalizes background noise level shifts

 Clips noise  

[Ayushman Tripathi]

Hi Stephen and Alex,

Thanks. I tried the HackRF today for H-line observations and got good results. Interestingly, it picks up less RFI than the RTL-SDR.

I tested on a weak region—even fainter than Cassiopeia (due to bad timing) —and still got a clear detection. Processed 2 observations through ezRA and the plots look good.

I will try SDR# with HackRF tomorrow to collect more data at stronger regions.

Thanks

[Stephen Arbogast]

All,   once you have collected raw data    the  battle  begins with  massaging   it.   My  Discovery Dish LNA  does not have  an  M shaped   curve.  I  want  to warn you that  you  still  have some work to  do....  Data analysis  is  an Art combined with  Science...  I  find   ezRACon   very helpful in sorting this   out.....

Alex  has done  wonderful  work...

[Stephen Arbogast]

Python   runs  on all  platforms.....  for  post data collection  I have found it very helpful  once you have    raw  data.  My  personal  choice  for  collecting   raw data is GNU Radio  flow  graph  since  it handles  thread management   and scheduling  data transfer....   then  I use Ted  Cline's Python software   for   post processing.

Will not be on the  Zoom   tomorrow.....  got a date  with mother  nature in  the  San Juan Mountains...........

[b alex pettit jr]

Hello Ayushman,

Here are some questions/comments on your plot set

Please Explain This Plot

[Ayushman Tripathi]

Hi Stephen, Yes, sure, I'll improve the data processing pipeline. Yesterday I attempted to edit the GNU Radio signal processing flowgraph used in Virgo (on my local system) to fix the DC spike by adding a dc_blocker_cc block and also tried gr-correctiq, though both caused spectrum distortion so I reverted and instead just mask the center frequency in post-processing now.

[Ayushman Tripathi]

Hi Alex,

Thanks for the detailed feedback! To answer your questions:

Filter response curve—Yes, that's my Discovery Dish Feed bandpass combined with the HackRF response. It gets divided out during calibration. Also today I added the center frequency to the masking list so the DC spike disappears even from the RAW spectrum.

Over what time interval is this averaged? - That was only a 5-minute observation. I was testing my system and had limited time while the galaxy was in my beam. Even in 5 minutes I think it shows a clear bump.

Why minute time intervals? Yes, I'll focus on doing longer observations soon.

The ezRA waterfall—that's from two consecutive 5-min observations I ran through ezCon (while I was testing it freshly set up on my system). I believe some of the horizontal bands are galactic hydrogen at different Doppler velocities, though there may be some RFI mixed in too.

I'm planning longer observations now—tonight I'll run 1 hour. What duration would you recommend for my setup 1hr or more ?

Thanks!

[Ayushman Tripathi]

Also, I've almost finished building my cantenna feed and will share progress pics soon! Next week I'm setting it up with the cantenna + SAWbird + H1 on a 2.5 m dish at a remote location with Raspberry Pi 5. After a few months I might try the DD feed on it too as an experiment, though it probably won't illuminate the dish edges well due to its beam pattern.

Thanks!

[b alex pettit jr]

Pointing error is a challenge ,. I can get to within +- a few degrees, but that is from 5 years of experience and having

pipes in the ground to reposition and Polar align the equatorial mount each time.. With an  Alt/Az, config. the challenge becomes doubly hard.

" 5 Minutes "  = One Average with my setup 

>> Set up for 1 degree ( 4 min )  samples =  15 saved files per hour 

Here is a 5 hour data set with my Discovery  Feed ( on the WiFi Dish )

Set it up and let it run 4-6 hours... and when you get your system optimized ( or therabouts ) 24 hr Drift Scan slices 

BTW : that center spike is an artifact of the SDR ..

AirSpy Mini does not have such.

No IQ imbalance, DC offset or 1/F noise at the center of the spectrum that plagues all the other SDRs

==========================================================

[b alex pettit jr]

For Reference,

This is what happens when my 4K monitor goes into power down mode .

It is located ~ 18m from the dish

[Ayushman Tripathi]

Hi Alex,

Here's the data I collected today: 3.5 hours of continuous H-line observation using HackRF & DD Feed.

Attached: waterfall, max spectrum, all 41 spectra overlaid, velocity vs galactic longitude, and VLSR tracking plots.

Fixed pointing at Az=330°, El=34°

Yes, for pointing, I use phone compass and trial and error for correction, watching live SNR to find the signal, and then I keep it fixed.

Also, after I covered the back of the dish, as you advised, the monitor/EcoFlow RFI dropped sharply.

Thanks.

[b alex pettit jr]

Hi Aysushman,

Can you please explain the axis on this scale ?

Why is there a quick level shift between 16 and 28 hrs ?

I have Validated the SDR# > IFavg>HL3D  software sequence. It provides data scaled in calibrated engineering units upon which I can judge Performance ..

 EZxx  is fine for presenting the data in many interesting plots, but known/calibrated data scaling is required for hardware development.

Alex

=================================================

[Ayushman Tripathi]

Hi Alex,

I suspected this was either galactic structure or system drift. I repeated the observation tonight with the same Az/El pointing, the level shift did NOT occur at the same UTC time. It was likely thermal drift or strong RFI, not astronomical.

Axis explanation:

• X-axis: UTC time / sample number
• Y-axis: Doppler offset from center freq.

Tonight's 7.5-hour run shows continuous H-line at -0.2 MHz offset with no sudden level changes.

The thin line at 0.0 MHz is the HackRF DC spike - I have the Airspy Mini arriving this week which should eliminate that meanwhile will try SDR# soon, I'll try to make the HackRF work with a Windows VM running in Ubuntu, so far, it's giving cleaner results than the RTL SDR for me.

Thanks

[Stephen Arbogast]

Hi  Ayushman,

I am not an expert  with  ezRA   by  any  measure   but....   I suspect  you are  still  dealing  with RFI.  The  horizontal  blue  lines  in your  ezRA plots  are RFI.  Here is  my  general  approach to  collecting  and post processing  H-Line.

1.  Collect  raw  data  using  your favorite software  while   minimizing  RFI as  much as is possible and save in a  format for  ezRA.

2.  Practice, practice, practice with  ezCon.py   Very  important  to eliminate  the outliers in your  raw data.

3. Be  patient.

4. Document  everything you do   success or  failure.

Attached  is a  typical  run for  me  using  ezRA   which I  learned from Ted Cline.....

[Stephen Arbogast]

One  can  see  still   some RFI in my heat map  plot  but  I managed  to minimize  it.

[Stephen Arbogast]

Or...    I need  to  take  out  low values...  with  further  post  processing.

[b alex pettit jr]

Ayushman,

Can you move your system to  friend's house, or a park or some place away from any RFI 

and run at least a few hours of acquisition as a baseline ??

What is the sample time per spectra on the plot below ?

Alex

On Monday, December 22, 2025 at 08:14:13 PM EST, Ayushman Tripathi <ayushmantr...@gmail.com> wrote:

Hi Alex,

I suspected this was either galactic structure or system drift. I repeated the observation tonight with the same Az/El pointing, the level shift did NOT occur at the same UTC time. It was likely thermal drift or strong RFI, not astronomical.

Axis explanation:

• X-axis: UTC time / sample number
• Y-axis: Doppler offset from center freq.

Tonight's 7.5-hour run shows continuous H-line at -0.2 MHz offset with no sudden level changes.

The thin line at 0.0 MHz is the HackRF DC spike - I have the Airspy Mini arriving this week which should eliminate that meanwhile will try SDR# soon, I'll try to make the HackRF work with a Windows VM running in Ubuntu, so far, it's giving cleaner results than the RTL SDR for me.

Thanks

[Ayushman Tripathi]

Hi Stephen,

Thanks! Yes, I'll try to improve RFI mitigation and mask the outliers. Sure, I'll practice more with ezCon.py. Your spectrum does look much cleaner.

Thanks.

[Ayushman Tripathi]

Hi Alex,

Yes, Next week I'm setting up a cantenna on the 2.5-meter dish at a remote location with no RFI, using the Airspy Mini.

This current location will continue to have some RFI, as it's in the middle of a city, I can see at least 10 cellular 5G towers within 500 meters of the dish :)

After adding aluminum foil shielding behind the dish as you suggested, the RFI has reduced significantly. It was much worse before, and the HackRF results are cleaner than RTL for me somehow. I'll mask the remaining RFI in software and try additional mitigation methods.

Also, here's the progress on my cantenna. I'm facing an issue: the 6mm copper tube is much wider than the solder point on the N-female connector.

Should I hammer the connector into the tube end or use a narrower solid copper wire (maybe 4 mm) instead of a tube?


Thanks a lot!

[Ayushman Tripathi]

Also, the sample time per spectra was 5 mins in the earlier plot.

[jpett...@gmail.com]

Round Copper Telescopic Tubing Assortment (7 Pieces) – ksmetals

 

Each tube slides inside the next up to a maximum of a ¼ inch. Start small to fit the solder point and work your way up.

 

Jonathan

 

From: sara...@googlegroups.com <sara...@googlegroups.com> On Behalf Of Ayushman Tripathi
Sent: Monday, December 22, 2025 9:46 PM
To: Society of Amateur Radio Astronomers <sara...@googlegroups.com>
Subject: Re: [SARA] Cantenna feed dimensions & SDR comparison for H-line

 

Hi Alex,

Yes, Next week I'm setting up a cantenna on the 2.5-meter dish at a remote location with no RFI, using the Airspy Mini.

This current location will continue to have some RFI, as it's in the middle of a city, I can see at least 10 cellular 5G towers within 500 meters of the dish :)

After adding aluminum foil shielding behind the dish as you suggested, the RFI has reduced significantly. It was much worse before, and the HackRF results are cleaner than RTL for me somehow. I'll mask the remaining RFI in software and try additional mitigation methods.

Also, here's the progress on my cantenna. I'm facing an issue: the 6mm copper tube is much wider than the solder point on the N-female connector.

Should I hammer the connector into the tube end or use a narrower solid copper wire (maybe 4 mm) instead of a tube?


Thanks a lot!

On Tuesday, December 23, 2025 at 9:00:24 AM UTC+5:30 Ayushman Tripathi wrote:

Hi Stephen,

Thanks! Yes, I'll try to improve RFI mitigation and mask the outliers. Sure, I'll practice more with ezCon.py. Your spectrum does look much cleaner.

Thanks.

On Tuesday, December 23, 2025 at 7:39:41 AM UTC+5:30 b alex pettit jr wrote:

Ayushman,

 

Can you move your system to  friend's house, or a park or some place away from any RFI 

and run at least a few hours of acquisition as a baseline ??

 

What is the sample time per spectra on the plot below ?

 

Alex

 

On Monday, December 22, 2025 at 08:14:13 PM EST, Ayushman Tripathi <ayushmantr...@gmail.com> wrote:

 

 

Hi Alex,

I suspected this was either galactic structure or system drift. I repeated the observation tonight with the same Az/El pointing, the level shift did NOT occur at the same UTC time. It was likely thermal drift or strong RFI, not astronomical.

Axis explanation:

• X-axis: UTC time / sample number
• Y-axis: Doppler offset from center freq.

Tonight's 7.5-hour run shows continuous H-line at -0.2 MHz offset with no sudden level changes.

The thin line at 0.0 MHz is the HackRF DC spike - I have the Airspy Mini arriving this week which should eliminate that meanwhile will try SDR# soon, I'll try to make the HackRF work with a Windows VM running in Ubuntu, so far, it's giving cleaner results than the RTL SDR for me.

Thanks

 

--
--
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 visit https://groups.google.com/d/msgid/sara-list/f43d135d-a3ed-4614-97c5-834e0105edb8n%40googlegroups.com.

[b alex pettit jr]

I would not be overly concerned with the wire  diameter..

2-3mm dia is Fine     & will wind up being   48-49mm long when tuned

[b alex pettit jr]

I suggest working on improving the Hardware and using  'stock' Software for the measurements of such

=  SDR# & IFavg  

[b alex pettit jr]

For Example :

If I turn on my monitor and it goes into sleep mode, I go from Left to Center

If I add the Al Foil extension on the reflector, it drops to the Right spectrum

Not Perfect, but it gives you an idea of things to try.

Alex

[Dykes Cupstid]

Alex, when setting the feed at the focal point do you measure to the center of the white box which houses the dipole?  In the photo, to what  dimension in centimeters is the focal distance set? Thanks.  Dykes Cupstid

---

From: 'b alex pettit jr' via Society of Amateur Radio Astronomers <sara...@googlegroups.com>
Sent: Tuesday, December 23, 2025 7:54 AM
To: 'b alex pettit jr' via Society of Amateur Radio Astronomers <sara...@googlegroups.com>

Subject: Re: [SARA] Cantenna feed dimensions & SDR comparison for H-line

--
--
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 visit https://groups.google.com/d/msgid/sara-list/1613677833.1977961.1766505254669%40mail.yahoo.com.

[b alex pettit jr]

Here's the Info.

I have nearly completed a full eval will create a new post .. 

Performance Exceeds anything I've seen in the  'out of the box'  beginner systems .. 

[andrew....@googlemail.com]

This is the Discovery Dish feed I presume?

Andy

 

From: 'b alex pettit jr' via Society of Amateur Radio Astronomers <sara...@googlegroups.com>
Sent: 23 December 2025 21:29
To: 'Dykes Cupstid' via Society of Amateur Radio Astronomers <sara...@googlegroups.com>
Subject: Re: [SARA] Cantenna feed dimensions & SDR comparison for H-line

 

Here's the Info.

--

--
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 visit https://groups.google.com/d/msgid/sara-list/1531180754.2045391.1766525320792%40mail.yahoo.com.

[b alex pettit jr]

Si Senior

[Andrew Thornett]

Thanks.

May I wish you a VERY Happy Christmas!

Andy

Sent from Outlook for Android

---

From: 'b alex pettit jr' via Society of Amateur Radio Astronomers <sara...@googlegroups.com>
Sent: Wednesday, December 24, 2025 12:07:39 AM
To: andrew.thornett via Society of Amateur Radio Astronomers <sara...@googlegroups.com>

Subject: Re: [SARA] Cantenna feed dimensions & SDR comparison for H-line

--
--
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 visit https://groups.google.com/d/msgid/sara-list/880748170.2066761.1766534859549%40mail.yahoo.com.

[Ayushman Tripathi]

Hi Alex,

Thanks! I'll try to improve the RFI shielding. I took a plot today for a few hours the data, but the RFI was worse than previous observations.

Yes, I'll try setting up a Windows VM on my home setup to run SDR# + IF AVG + HACKRF / RTL.

For the remote setup, it'll run on a Raspberry Pi + AIRSPY, so VIRGO seems to be the only option for now, as SDR# needs . NET,  occasionally I can try with the SDR# Linux server and see if it works well remotely.

Also do you think a SMA Female Solder Cup is also fine for the Cantenna ? I was thinking of getting 1 as a backup for N-Female.

Thanks.

[b alex pettit jr]

SMA connectors are physically weak due to their size ..  You can try to use such, but

you may have issues .  a Type N is much stronger ,..

[b alex pettit jr]

Here's an idea :

Use an SMA Bulkhead connector and support the LNA with an L Bracket 

[b alex pettit jr]

Hello Ayushman,

I made this software comparison a few yeas ago of ~ the same slice of sky.

( ezra plot furnished to me ) 

Regards,

Alex

==============================

On Tuesday, December 23, 2025 at 09:11:11 PM EST, Ayushman Tripathi <ayushmantr...@gmail.com> wrote:

Hi Alex,

Thanks! I'll try to improve the RFI shielding. I took a plot today for a few hours the data, but the RFI was worse than previous observations.

[Ayushman Tripathi]

Hi Alex,

Merry Christmas!

Thanks! I did try SDR# today, but I ran into its limitations with my HackRF One. SDR# enforces a minimum sample rate of 8 MHz for HackRF, but the hardware can actually go down to 2 MHz on other SDR apps. For hydrogen line observations, 8 MHz is too wide. I use 2.4 MHz bandwidth, which works well in VIRGO, but SDR# won't allow it.

I also experimented with SDRangel's Radio Astronomy plugin but haven't had enough time to fully understand the settings or get it working properly.

Once I receive my Airspy this weekend, I'll give SDR# another try, as it seems more optimized for Airspy hardware.

Regarding the cantenna, I've decided to skip the scalar ring for now and may revisit it in a few months.

Also, can the H-line 3D Java program work with data collected from other software like VIRGO? I'd like to use it for 3D visualization if I can convert my data to the right format. N

[Stephen Arbogast]

Hi Ayushman,

I would  like  to speak up here.....   just my  observations from my experience...

1. Keep it simple  starting out.  Don't try  SDR Angel   until  you  get  the  basics  down.. it is  advanced  with all the  calibration stuff. We  are not  calibrating   our  stuff  here, just  getting  relative  measurements.

2. I have  used  the   AirSpy R2   for  about  a  year  now   with SDR#  on  Windows....   no problems  but I don't understand the  advice to  double  the sample rate then   decimate by  2.  According  to Nyquist  Sampling Theorem your  sample rate should  be a minimum of twice the  bandwidth you need which    the sample  rate for  Air Spy  should  be   2.4 Meg with no  decimation.   If the   sample  rate  is too low  you  will  experience aliasing     which is  folding     over of  signal into your  desired  band  causing problems.

3.  ezRA  is excellent  software from Ted  Cline but it is  a very  large suit  of  software tools...    which takes time to  learn....   start simple.

4.  About  VIRGO... good stuff but  dated   ...  GNU  Radio  has been updated  since then.  I use  an Infinite  Impulse Response  Single  Pole filter  instead of   the four   pole filter in  VIRGO.  I  won't include results   here because I am using/testing    Ted Cline's  pre-release  software  with my experimental     front  end  GNU  Radio data collection flow graph   which  I have modified  from  original author.

5.  HackRF  SDR  is  not   sensitive  enough  for  H-Line   work....   Hope  you  are using  an LNA  with  low   noise figure.