10GHz WG transition


John Fell
 

Found this design by G4JNT just now .(see attached)
I have made several and they are easy to adjust using cross coupler and diode detector .

The back short was milled from 1"x 1/2" Brass bar into a T section length and then cut into 1/2" lengths , providing a solid contact around the inner guide faces .

The short is inserted partially and return loss measured - if required the short is squeezed further into the guide using a vice or thumped with a hammer and measurement repeated .

As shown the socket is based on a flanged N type with extended spill.An SMA version is fine , with the spill extended using the outer of solid jacket cable to within a mm of the inner guide wall .

73
John
G0API


Andy G4JNT
 

'snot my design.    If I gave it to you,  'twould have been nicked from somewhere else. 
Remember decades ago hearing a rule of thumb along the lines of  "a fifth of a wavelength probe at a fifth of a wavelength from the back short".



On Thu, 10 Jun 2021 at 10:33, John Fell <john.g0api@...> wrote:
Found this design by G4JNT just now .(see attached)
I have made several and they are easy to adjust using cross coupler and diode detector .

The back short was milled from 1"x 1/2" Brass bar into a T section length and then cut into 1/2" lengths , providing a solid contact around the inner guide faces .

The short is inserted partially and return loss measured - if required the short is squeezed further into the guide using a vice or thumped with a hammer and measurement repeated .

As shown the socket is based on a flanged N type with extended spill.An SMA version is fine , with the spill extended using the outer of solid jacket cable to within a mm of the inner guide wall .

73
John
G0API


John Fell
 

Fairy snuff .
It was based on one you made and I swopped for a T bar , rear of guide socket based thingy - long time ago .

I also like the 3 stub /screw design using brass or copper plated brass screws for a more defined bandwidth .

73
John
G0API

On Thu, 10 Jun 2021 at 10:48, Andy G4JNT <andy.g4jnt@...> wrote:
'snot my design.    If I gave it to you,  'twould have been nicked from somewhere else. 
Remember decades ago hearing a rule of thumb along the lines of  "a fifth of a wavelength probe at a fifth of a wavelength from the back short".



On Thu, 10 Jun 2021 at 10:33, John Fell <john.g0api@...> wrote:
Found this design by G4JNT just now .(see attached)
I have made several and they are easy to adjust using cross coupler and diode detector .

The back short was milled from 1"x 1/2" Brass bar into a T section length and then cut into 1/2" lengths , providing a solid contact around the inner guide faces .

The short is inserted partially and return loss measured - if required the short is squeezed further into the guide using a vice or thumped with a hammer and measurement repeated .

As shown the socket is based on a flanged N type with extended spill.An SMA version is fine , with the spill extended using the outer of solid jacket cable to within a mm of the inner guide wall .

73
John
G0API


Alan Beard
 

Hi John,

"T" bar, can the bar be replaced by a flat strip?

Are there any simulations to give us some guidelines?

I'm looking for "no tune" transitions for 1, 2 and 3GHz.
Where the waveguide will be extruded aluminium and
all the added bits will be brass with the coax soldered to them. 
Also, to reduce the inductance where the "inner goes to the bar",
bring the coax right into the waveguide, the coax outer being a larger
diameter.

I have a 3.5GHz "T" bar transition here from 1950 perhaps, beautifully made
but a custom, never seen anywhere, coax connector.
WG 2 3/8" x 1 1/8" or 60mm x 29mm and the bar 1/2" or 12.5mm 5/8" or 16mm
from the back-short, 3/8 " or 10mm centre up from the WG broad face. 
Thus the gap down to the broad face is about 4mm.

Alan VK2ZIW

On Thu, 10 Jun 2021 11:06:27 +0100, John Fell wrote
> Fairy snuff .
> It was based on one you made and I swopped for a T bar , rear of guide socket based thingy - long time ago .
>
> I also like the 3 stub /screw design using brass or copper plated brass screws for a more defined bandwidth .
>
> 73
> John
> G0API
>
> On Thu, 10 Jun 2021 at 10:48, Andy G4JNT <andy.g4jnt@...> wrote:
>

>
> 'snot my design.    If I gave it to you,  'twould have been nicked from somewhere else. 
> Remember decades ago hearing a rule of thumb along the lines of  "a fifth of a wavelength probe at a fifth of a wavelength from the back short".
>
> Andy
> www.g4jnt.com
>
> On Thu, 10 Jun 2021 at 10:33, John Fell <john.g0api@...> wrote:
>

> Found this design by G4JNT just now .(see attached)
> I have made several and they are easy to adjust using cross coupler and diode detector .
>
> The back short was milled from 1"x 1/2" Brass bar into a T section length and then cut into 1/2" lengths , providing a solid contact around the inner guide faces .
>
> The short is inserted partially and return loss measured - if required the short is squeezed further into the guide using a vice or thumped with a hammer and measurement repeated .
>
> As shown the socket is based on a flanged N type with extended spill.An SMA version is fine , with the spill extended using the outer of solid jacket cable to within a mm of the inner guide wall .
>
> 73
> John
> G0API
>
>

>
>


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


John Fell
 

Hi Alan,
Probably but I have no data .
At the low frequencies you quote I would use a conventional 1/4 wave dipole structure mounted on a suitable coaxial socket .For no tune expand the dipole diameter and it should give a wideband performance .
Material mix could give you long term intermetallic issues in a damp environment .

73
John
G0API   

On Fri, 11 Jun 2021 at 06:45, Alan Beard <beardal@...> wrote:
Hi John,

"T" bar, can the bar be replaced by a flat strip?

Are there any simulations to give us some guidelines?

I'm looking for "no tune" transitions for 1, 2 and 3GHz.
Where the waveguide will be extruded aluminium and
all the added bits will be brass with the coax soldered to them. 
Also, to reduce the inductance where the "inner goes to the bar",
bring the coax right into the waveguide, the coax outer being a larger
diameter.

I have a 3.5GHz "T" bar transition here from 1950 perhaps, beautifully made
but a custom, never seen anywhere, coax connector.
WG 2 3/8" x 1 1/8" or 60mm x 29mm and the bar 1/2" or 12.5mm 5/8" or 16mm
from the back-short, 3/8 " or 10mm centre up from the WG broad face. 
Thus the gap down to the broad face is about 4mm.

Alan VK2ZIW

On Thu, 10 Jun 2021 11:06:27 +0100, John Fell wrote
> Fairy snuff .
> It was based on one you made and I swopped for a T bar , rear of guide socket based thingy - long time ago .
>
> I also like the 3 stub /screw design using brass or copper plated brass screws for a more defined bandwidth .
>
> 73
> John
> G0API
>
> On Thu, 10 Jun 2021 at 10:48, Andy G4JNT <andy.g4jnt@...> wrote:
>

>
> 'snot my design.    If I gave it to you,  'twould have been nicked from somewhere else. 
> Remember decades ago hearing a rule of thumb along the lines of  "a fifth of a wavelength probe at a fifth of a wavelength from the back short".
>
> Andy
> www.g4jnt.com
>
> On Thu, 10 Jun 2021 at 10:33, John Fell <john.g0api@...> wrote:
>

> Found this design by G4JNT just now .(see attached)
> I have made several and they are easy to adjust using cross coupler and diode detector .
>
> The back short was milled from 1"x 1/2" Brass bar into a T section length and then cut into 1/2" lengths , providing a solid contact around the inner guide faces .
>
> The short is inserted partially and return loss measured - if required the short is squeezed further into the guide using a vice or thumped with a hammer and measurement repeated .
>
> As shown the socket is based on a flanged N type with extended spill.An SMA version is fine , with the spill extended using the outer of solid jacket cable to within a mm of the inner guide wall .
>
> 73
> John
> G0API
>
>

>
>


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


Alan Beard
 

Hi all,

I asked about the "T" bar transition a while back, has anybody tried a flat strip replacing the round cross bar?

Could somebody do a simulation to give use guidelines for a "simple to build" transition?

Alan VK2ZIW

On Fri, 11 Jun 2021 09:57:23 +0100, John Fell wrote
> Hi Alan,
> Probably but I have no data .
> At the low frequencies you quote I would use a conventional 1/4 wave dipole structure mounted on a suitable coaxial socket .For no tune expand the dipole diameter and it should give a wideband performance .
> Material mix could give you long term intermetallic issues in a damp environment .
>
> 73
> John
> G0API   
>
> On Fri, 11 Jun 2021 at 06:45, Alan Beard <beardal@...> wrote:
>

> Hi John,
>
>

> "T" bar, can the bar be replaced by a flat strip?
>
>

> Are there any simulations to give us some guidelines?
>
>

> I'm looking for "no tune" transitions for 1, 2 and 3GHz.
> Where the waveguide will be extruded aluminium and
> all the added bits will be brass with the coax soldered to them. 
> Also, to reduce the inductance where the "inner goes to the bar",
> bring the coax right into the waveguide, the coax outer being a larger
> diameter.
>
>

> I have a 3.5GHz "T" bar transition here from 1950 perhaps, beautifully made
> but a custom, never seen anywhere, coax connector.
> WG 2 3/8" x 1 1/8" or 60mm x 29mm and the bar 1/2" or 12.5mm 5/8" or 16mm
> from the back-short, 3/8 " or 10mm centre up from the WG broad face. 
> Thus the gap down to the broad face is about 4mm.
>
>

> Alan VK2ZIW
>
>

> On Thu, 10 Jun 2021 11:06:27 +0100, John Fell wrote
> > Fairy snuff .
> > It was based on one you made and I swopped for a T bar , rear of guide socket based thingy - long time ago .
> >
> > I also like the 3 stub /screw design using brass or copper plated brass screws for a more defined bandwidth .
> >
> > 73
> > John
> > G0API
> >
> > On Thu, 10 Jun 2021 at 10:48, Andy G4JNT <andy.g4jnt@...> wrote:
> >

> >
> > 'snot my design.    If I gave it to you,  'twould have been nicked from somewhere else. 
> > Remember decades ago hearing a rule of thumb along the lines of  "a fifth of a wavelength probe at a fifth of a wavelength from the back short".
> >
> > Andy
> > www.g4jnt.com
> >
> > On Thu, 10 Jun 2021 at 10:33, John Fell <john.g0api@...> wrote:
> >

> > Found this design by G4JNT just now .(see attached)
> > I have made several and they are easy to adjust using cross coupler and diode detector .
> >
> > The back short was milled from 1"x 1/2" Brass bar into a T section length and then cut into 1/2" lengths , providing a solid contact around the inner guide faces .
> >
> > The short is inserted partially and return loss measured - if required the short is squeezed further into the guide using a vice or thumped with a hammer and measurement repeated .
> >
> > As shown the socket is based on a flanged N type with extended spill.An SMA version is fine , with the spill extended using the outer of solid jacket cable to within a mm of the inner guide wall .
> >
> > 73
> > John
> > G0API
> >
> >

> >
> >

>
> ---------------------------------------------------
> 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.