A quick lashup 2.3 to 3.4 GHz horn


Alan Beard
 

Hi all,

Ridged waveguide, supposedly broad bandwidth horn antennae
are out there, so I had a go at making one.

This is 100 x 50mm OD box section aluminium tube.

What would make it much much easier to build is if the
ridge could be thin, and I mean only one sheet of metal thick. 0.5mm.

ALA Vivaldi style.

Has anyone tried this?


---------------------------------------------------
Alan VK2ZIW
Before the Big Bang, God, Sela.
OpenWebMail 2.53, nothing in the cloud.


Murray Niman
 


You may find that the central ridge/flair ideally needs to extend further to prevent 2.3 cutoff/dispersion and smooth out the impedance change

There is no reason why you cant start by taking one of Kents WA5VJBs flaired pcb vivalidis such as https://www.ebay.co.uk/itm/254170352417  (the hard bit) and then enclose it in a DIY horn surround built from metal or blank fr4 PCB to direct the radiation better

  73

 Murray G6JYB 


Andy G4JNT
 

I found this horn in my junk box.   It's been a bit bashed about, and there is no label, but I just tested it's match, and there is something li[ke -13dB RL at 2.4GHz and it changes if I wave my hand around in front, so appears to be working (ish)

Is it reasonable to estimate its gain from the front area, and if so, what sort of efficiency could I expect?
I'm after an antenna of accurately known gain, although it doesn't have to be too much gain,  at 2.4GHz for some QO100 measurements.

And trying to resist constructing one from tin plate.



On Sat, 5 Jun 2021 at 12:04, Murray Niman <g6jyb@...> wrote:

You may find that the central ridge/flair ideally needs to extend further to prevent 2.3 cutoff/dispersion and smooth out the impedance change

There is no reason why you cant start by taking one of Kents WA5VJBs flaired pcb vivalidis such as https://www.ebay.co.uk/itm/254170352417  (the hard bit) and then enclose it in a DIY horn surround built from metal or blank fr4 PCB to direct the radiation better

  73

 Murray G6JYB 


Andy G4JNT
 


On Sat, 5 Jun 2021 at 12:32, Andy Talbot <andy.g4jnt@...> wrote:
I found this horn in my junk box.   It's been a bit bashed about, and there is no label, but I just tested it's match, and there is something li[ke -13dB RL at 2.4GHz and it changes if I wave my hand around in front, so appears to be working (ish)

Is it reasonable to estimate its gain from the front area, and if so, what sort of efficiency could I expect?
I'm after an antenna of accurately known gain, although it doesn't have to be too much gain,  at 2.4GHz for some QO100 measurements.

And trying to resist constructing one from tin plate.



On Sat, 5 Jun 2021 at 12:04, Murray Niman <g6jyb@...> wrote:

You may find that the central ridge/flair ideally needs to extend further to prevent 2.3 cutoff/dispersion and smooth out the impedance change

There is no reason why you cant start by taking one of Kents WA5VJBs flaired pcb vivalidis such as https://www.ebay.co.uk/itm/254170352417  (the hard bit) and then enclose it in a DIY horn surround built from metal or blank fr4 PCB to direct the radiation better

  73

 Murray G6JYB 


Tim Hague M0AFJ
 

Looks like a 1-18GHz horn similar to the ETS Lindgren / Emco 3115.

Best Regards
Tim Hague


On 5 Jun 2021, at 12:36, Andy G4JNT <andy.g4jnt@...> wrote:


Would have helped if I'd attached some photos.

<Vhorn1.jpg>


<Vhorn2.jpg>




On Sat, 5 Jun 2021 at 12:32, Andy Talbot <andy.g4jnt@...> wrote:
I found this horn in my junk box.   It's been a bit bashed about, and there is no label, but I just tested it's match, and there is something li[ke -13dB RL at 2.4GHz and it changes if I wave my hand around in front, so appears to be working (ish)

Is it reasonable to estimate its gain from the front area, and if so, what sort of efficiency could I expect?
I'm after an antenna of accurately known gain, although it doesn't have to be too much gain,  at 2.4GHz for some QO100 measurements.

And trying to resist constructing one from tin plate.



On Sat, 5 Jun 2021 at 12:04, Murray Niman <g6jyb@...> wrote:

You may find that the central ridge/flair ideally needs to extend further to prevent 2.3 cutoff/dispersion and smooth out the impedance change

There is no reason why you cant start by taking one of Kents WA5VJBs flaired pcb vivalidis such as https://www.ebay.co.uk/itm/254170352417  (the hard bit) and then enclose it in a DIY horn surround built from metal or blank fr4 PCB to direct the radiation better

  73

 Murray G6JYB 


SAM JEWELL
 

To which I add I have a number of the 1.5 -15GHz ones in stock.

Sam, G4DDK






On 5 Jun 2021, at 12:04, Murray Niman <g6jyb@...> wrote:


You may find that the central ridge/flair ideally needs to extend further to prevent 2.3 cutoff/dispersion and smooth out the impedance change

There is no reason why you cant start by taking one of Kents WA5VJBs flaired pcb vivalidis such as https://www.ebay.co.uk/itm/254170352417  (the hard bit) and then enclose it in a DIY horn surround built from metal or blank fr4 PCB to direct the radiation better

  73

 Murray G6JYB 


Andy G4JNT
 

Looks a bit like that, but it's smaller
Mouth dimensions (internal)  104 x 69mm



On Sat, 5 Jun 2021 at 13:16, Tim Hague M0AFJ via groups.io <m0afj=yahoo.co.uk@groups.io> wrote:
Looks like a 1-18GHz horn similar to the ETS Lindgren / Emco 3115.

Best Regards
Tim Hague



Alan Beard
 

Hi all,

Thanks guys for the pictures of commercial wide band horns but...

Can somebody answer the question please?

The ridge thickness, can it be thin, one sheet of sheet metal, say 1mm ?

(Just like the Vivaldi)

Does this affect the feed impedance?

Here is my SAAV2 scan.

Alan VK2ZIW


On Sat, 5 Jun 2021 13:31:05 +0100, Andy G4JNT wrote
> Looks a bit like that, but it's smaller
> Mouth dimensions (internal)  104 x 69mm
>
> Andy
> www.g4jnt.com
>
> On Sat, 5 Jun 2021 at 13:16, Tim Hague M0AFJ via groups.io <m0afj=yahoo.co.uk@groups.io> wrote:
>

> Looks like a 1-18GHz horn similar to the ETS Lindgren / Emco 3115.
>
> Best Regards
> Tim Hague
>
>


---------------------------------------------------
Alan VK2ZIW
Before the Big Bang, God, Sela.
OpenWebMail 2.53, nothing in the cloud.


KENT BRITAIN
 

I have perhaps a dozen Ridged horns of different sub-species.

Most have a ridge several millimeters wide.

But my 1-18 GHz one has the ridge ground to a fine edge along the top.

I think the fine edge helps the upper frequency limit.   How close you can make the 
gap between the top and bottom ridges has a major effect on the upper frequency limit of the antenna.
Kent.

On Saturday, June 5, 2021, 05:13:38 PM CDT, Alan Beard <beardal@...> wrote:


Hi all,

Thanks guys for the pictures of commercial wide band horns but...

Can somebody answer the question please?

The ridge thickness, can it be thin, one sheet of sheet metal, say 1mm ?

(Just like the Vivaldi)

Does this affect the feed impedance?

Here is my SAAV2 scan.

Alan VK2ZIW


On Sat, 5 Jun 2021 13:31:05 +0100, Andy G4JNT wrote
> Looks a bit like that, but it's smaller
> Mouth dimensions (internal)  104 x 69mm
>
> Andy
> www.g4jnt.com
>
> On Sat, 5 Jun 2021 at 13:16, Tim Hague M0AFJ via groups.io <m0afj=yahoo.co.uk@groups.io> wrote:
>

> Looks like a 1-18GHz horn similar to the ETS Lindgren / Emco 3115.
>
> Best Regards
> Tim Hague
>
>


---------------------------------------------------
Alan VK2ZIW
Before the Big Bang, God, Sela.
OpenWebMail 2.53, nothing in the cloud.


Alan Beard
 

Thanks Kent,

Can you do a Return Loss plot to determine if the feed impedance is 50 ohms?

I can readily get the gap to between 1 to 2 mm witch surely is close enough for the 4GHz 
or so maximum frequency I intend to use with my HackRF One.

It's a practical thing. I can cut with tin-snips a smooth curve. But to fold metal to simulate
a thicker piece, I cannot achieve a curve, only a straight edge as you see in my image
in the previous email.
From:
https://www.slideshare.net/mariasheebap/ridge-guide
 
My thought, using a very thin ridge puts the  a2/a1 down at 0.01 where the graph becomes obscure.
And the impedances get high.

If the ridge were 1/5th the waveguide broadface width and the  gap was 1/20 the waveguide narrowface,
the impedance would be around 50 ohms. (for their 0.7 a/b waveguide)


And, looking at a Return Loss plot of a 1 - 18 GHz, it was not that good, around 10 to 15db.
With those results, I'm putting my horn on my antenna pole and giving it a go.

Alan VK2ZIW
 


On Sat, 5 Jun 2021 23:05:42 +0000 (UTC), KENT BRITAIN wrote
> I have perhaps a dozen Ridged horns of different sub-species.
>
> Most have a ridge several millimeters wide.
>
> But my 1-18 GHz one has the ridge ground to a fine edge along the top.
>
> I think the fine edge helps the upper frequency limit.   How close you can make the 
> gap between the top and bottom ridges has a major effect on the upper frequency limit of the antenna.
> Kent.
>
> On Saturday, June 5, 2021, 05:13:38 PM CDT, Alan Beard <beardal@...> wrote:
>
> Hi all,
>

> Thanks guys for the pictures of commercial wide band horns but...
>

> Can somebody answer the question please?
>

> The ridge thickness, can it be thin, one sheet of sheet metal, say 1mm ?
>

> (Just like the Vivaldi)
>

> Does this affect the feed impedance?
>

> Here is my SAAV2 scan.
>

> Alan VK2ZIW
>

>

> On Sat, 5 Jun 2021 13:31:05 +0100, Andy G4JNT wrote
> Looks a bit like that, but it's smaller
> Mouth dimensions (internal)  104 x 69mm
>
> Andy
> www.g4jnt.com
>
> On Sat, 5 Jun 2021 at 13:16, Tim Hague M0AFJ via groups.io <m0afj=yahoo.co.uk@groups.io> wrote:
>

> Looks like a 1-18GHz horn similar to the ETS Lindgren / Emco 3115.
>
> Best Regards
> Tim Hague
>
>


---------------------------------------------------
Alan VK2ZIW
Before the Big Bang, God, Sela.
OpenWebMail 2.53, nothing in the cloud.


>

---------------------------------------------------
Alan VK2ZIW
Before the Big Bang, God, Sela.
OpenWebMail 2.53, nothing in the cloud.


KENT BRITAIN
 

Hi Alan

Most of these horns are used on my antenna range and have Network Analyzer S11 and Gain tables laminated on them.   Yes, they work nicely in 50 Ohm systems.

Traditionally on an exponentiation Slot or Vivaldi antenna the feed point was 1/4 wave in front of a short.
Later is was learned that replacing the 1/4 wave slot with a circle greatly broadened the frequency response of the antenna.     As for the ridges themselves.   The size of the opening determines the lowest frequency they will work at. How close you can position the slots determines the highest frequency.   Basically it's the structure behind the feedpoint that is helping to set the impedance.

Kent

PS  I like to think to the Vivaldi antennas as a ridged horn without the horn! hihi

On Sunday, June 6, 2021, 04:32:47 AM CDT, Alan Beard <beardal@...> wrote:


Thanks Kent,

Can you do a Return Loss plot to determine if the feed impedance is 50 ohms?

I can readily get the gap to between 1 to 2 mm witch surely is close enough for the 4GHz 
or so maximum frequency I intend to use with my HackRF One.

It's a practical thing. I can cut with tin-snips a smooth curve. But to fold metal to simulate
a thicker piece, I cannot achieve a curve, only a straight edge as you see in my image
in the previous email.
From:
https://www.slideshare.net/mariasheebap/ridge-guide
 
My thought, using a very thin ridge puts the  a2/a1 down at 0.01 where the graph becomes obscure.
And the impedances get high.

If the ridge were 1/5th the waveguide broadface width and the  gap was 1/20 the waveguide narrowface,
the impedance would be around 50 ohms. (for their 0.7 a/b waveguide)


And, looking at a Return Loss plot of a 1 - 18 GHz, it was not that good, around 10 to 15db.
With those results, I'm putting my horn on my antenna pole and giving it a go.

Alan VK2ZIW
 


On Sat, 5 Jun 2021 23:05:42 +0000 (UTC), KENT BRITAIN wrote
> I have perhaps a dozen Ridged horns of different sub-species.
>
> Most have a ridge several millimeters wide.
>
> But my 1-18 GHz one has the ridge ground to a fine edge along the top.
>
> I think the fine edge helps the upper frequency limit.   How close you can make the 
> gap between the top and bottom ridges has a major effect on the upper frequency limit of the antenna.
> Kent.
>
> On Saturday, June 5, 2021, 05:13:38 PM CDT, Alan Beard <beardal@...> wrote:
>
> Hi all,
>

> Thanks guys for the pictures of commercial wide band horns but...
>

> Can somebody answer the question please?
>

> The ridge thickness, can it be thin, one sheet of sheet metal, say 1mm ?
>

> (Just like the Vivaldi)
>

> Does this affect the feed impedance?
>

> Here is my SAAV2 scan.
>

> Alan VK2ZIW
>

>

> On Sat, 5 Jun 2021 13:31:05 +0100, Andy G4JNT wrote
> Looks a bit like that, but it's smaller
> Mouth dimensions (internal)  104 x 69mm
>
> Andy
> www.g4jnt.com
>
> On Sat, 5 Jun 2021 at 13:16, Tim Hague M0AFJ via groups.io <m0afj=yahoo.co.uk@groups.io> wrote:
>

> Looks like a 1-18GHz horn similar to the ETS Lindgren / Emco 3115.
>
> Best Regards
> Tim Hague
>
>


---------------------------------------------------
Alan VK2ZIW
Before the Big Bang, God, Sela.
OpenWebMail 2.53, nothing in the cloud.


>

---------------------------------------------------
Alan VK2ZIW
Before the Big Bang, God, Sela.
OpenWebMail 2.53, nothing in the cloud.


Alan Beard
 

Hi Kent and all,

On looking at the table in my previous email, I've rebuilt the feed.
The waveguide here is 100 x 50mm OD or 47 x 97mm ID.
So I went for a 20mm wide ridge (start) and because I can't fold a curve,
I elected to widen the ridge as we move away from the feed point.

2300.1  and 3398.1 MHz are my targets.

RF measurement with the NanoVNA V2_2 SAAV2 is very difficult without
good adapters and a good low return-loss terminator in BNC. 

(So many of the BNC terminators here are from "Thin-Net" 10Mbits)

I find BNC hardware so practical.

Anyway, the horn is ready for service.

Alan VK2ZIW

On Sun, 6 Jun 2021 13:53:32 +0000 (UTC), KENT BRITAIN wrote
> Hi Alan
>
> Most of these horns are used on my antenna range and have Network Analyzer S11 and Gain tables laminated on them.   Yes, they work nicely in 50 Ohm systems.
>
> Traditionally on an exponentiation Slot or Vivaldi antenna the feed point was 1/4 wave in front of a short.
> Later is was learned that replacing the 1/4 wave slot with a circle greatly broadened the frequency response of the antenna.      As for the ridges themselves.   The size of the opening determines the lowest frequency they will work at. How close you can position the slots determines the highest frequency.   Basically it's the structure behind the feedpoint that is helping to set the impedance.
>
> Kent
>
> PS  I like to think to the Vivaldi antennas as a ridged horn without the horn! hihi
>
> On Sunday, June 6, 2021, 04:32:47 AM CDT, Alan Beard <beardal@...> wrote:
>
> Thanks Kent,
>



---------------------------------------------------
Alan VK2ZIW
Before the Big Bang, God, Sela.
OpenWebMail 2.53, nothing in the cloud.