Waveguide switch replacement


John Fell
 

I have a defective WG20 switch and a working WG22 switch , both 4 port Ledex activated at 20V d.c. 
My question is , could I use the WG22 device on 24GHZ using suitable adaptors ?
If so has anyone got any data on suitable adaptors ?

73
John
G0API


Paul G8KFW
 

Hi john

Regarding WG 22 the normal band is 26.5 to 40 Ghz  the lower cut off is about 21.1 ghz

 

I regularly use  WG 22 components for testing at 24 Ghz and all works OK

Also other amateurs use  WG 22 for a dual band unit for  24  and 47 Ghz

 

Best of luck with the metal work for any transitions

Regards Paul B

 


From: UKMicrowaves@groups.io [mailto:UKMicrowaves@groups.io] On Behalf Of John Fell
Sent: 01 February 2019 11:14
To: ukmicrowaves@groups.io
Subject: [UKMicrowaves] Waveguide switch replacement

 

I have a defective WG20 switch and a working WG22 switch , both 4 port Ledex activated at 20V d.c. 

My question is , could I use the WG22 device on 24GHZ using suitable adaptors ?

If so has anyone got any data on suitable adaptors ?

 

73

John

G0API

No virus found in this message.
Checked by AVG - www.avg.com
Version: 2016.0.8048 / Virus Database: 4793/15884 - Release Date: 08/14/18
Internal Virus Database is out of date.


--
Paul Bicknell G8KFW   South Coast UK


Mike Willis
 

WG22 is recommended for 26GHz-40GHz but it is still well above cut-off at 24GHz. I think the New Zealand relays  Iam using are WG22. I simply connected them together directly, by filing slots in the WG20 flange because the hole spacing is different. I used 3 fine screws to tune out the miss-match in the usual manner. It should work OK for narrow bandwidths. If you don't have a 24 GHz VSWR meter or VNA, just use a directional coupler and a sensitive detector (e.g. a spectrum analyser or a crystal detector). The old 22GHz HP8559 analysers work fine at 24GHz, they are un-calibrated but you are only looking for a null.
--
Mike G0MJW


Paul G8KFW
 

Hi john

I think Mike and I have hade the same positive experience with using WG 22 at 24 Ghz

 

Regarding spectrum analysers

Yes the 21 GHz    HP 8559 analysers work fine at 24GHz, as that is what I  used for the 24 ghz  FM  TV link I demonstrated at a BATC convention some 30 plus years ago

The old  HP 18 ghz  in the  140 main frame also works OK at 24 ghz using a WG 22 external mixer  

 

Regarding power measurements the WG 22  HP power sensors are useable at 24 Ghz

 

Just as a side question what is your approximate location  

Paul B

 


From: UKMicrowaves@groups.io [mailto:UKMicrowaves@groups.io] On Behalf Of Mike Willis
Sent: 01 February 2019 11:52
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] Waveguide switch replacement

 

WG22 is recommended for 26GHz-40GHz but it is still well above cut-off at 24GHz. I think the New Zealand relays  Iam using are WG22. I simply connected them together directly, by filing slots in the WG20 flange because the hole spacing is different. I used 3 fine screws to tune out the miss-match in the usual manner. It should work OK for narrow bandwidths. If you don't have a 24 GHz VSWR meter or VNA, just use a directional coupler and a sensitive detector (e.g. a spectrum analyser or a crystal detector). The old 22GHz HP8559 analysers work fine at 24GHz, they are un-calibrated but you are only looking for a null.
--
Mike G0MJW

No virus found in this message.
Checked by AVG - www.avg.com
Version: 2016.0.8048 / Virus Database: 4793/15884 - Release Date: 08/14/18
Internal Virus Database is out of date.


--
Paul Bicknell G8KFW   South Coast UK


John Fell
 

Thanks Paul .
The bandwidth quoted on the Flann site indicated 26.4  -  40.1GHZ  , so good to know it above cutoff still at 24GHZ.

73
John
G0API

On Fri, 1 Feb 2019 at 11:44, Paul Bicknell <paul@...> wrote:

Hi john

Regarding WG 22 the normal band is 26.5 to 40 Ghz  the lower cut off is about 21.1 ghz

 

I regularly use  WG 22 components for testing at 24 Ghz and all works OK

Also other amateurs use  WG 22 for a dual band unit for  24  and 47 Ghz

 

Best of luck with the metal work for any transitions

Regards Paul B

 


From: UKMicrowaves@groups.io [mailto:UKMicrowaves@groups.io] On Behalf Of John Fell
Sent: 01 February 2019 11:14
To: ukmicrowaves@groups.io
Subject: [UKMicrowaves] Waveguide switch replacement

 

I have a defective WG20 switch and a working WG22 switch , both 4 port Ledex activated at 20V d.c. 

My question is , could I use the WG22 device on 24GHZ using suitable adaptors ?

If so has anyone got any data on suitable adaptors ?

 

73

John

G0API

No virus found in this message.
Checked by AVG - www.avg.com
Version: 2016.0.8048 / Virus Database: 4793/15884 - Release Date: 08/14/18
Internal Virus Database is out of date.


--
Paul Bicknell G8KFW   South Coast UK


John Fell
 

Approx location is IO80XS77JS .

73
John
G0API

On Fri, 1 Feb 2019 at 12:31, Paul Bicknell <paul@...> wrote:

Hi john

I think Mike and I have hade the same positive experience with using WG 22 at 24 Ghz

 

Regarding spectrum analysers

Yes the 21 GHz    HP 8559 analysers work fine at 24GHz, as that is what I  used for the 24 ghz  FM  TV link I demonstrated at a BATC convention some 30 plus years ago

The old  HP 18 ghz  in the  140 main frame also works OK at 24 ghz using a WG 22 external mixer  

 

Regarding power measurements the WG 22  HP power sensors are useable at 24 Ghz

 

Just as a side question what is your approximate location  

Paul B

 


From: UKMicrowaves@groups.io [mailto:UKMicrowaves@groups.io] On Behalf Of Mike Willis
Sent: 01 February 2019 11:52
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] Waveguide switch replacement

 

WG22 is recommended for 26GHz-40GHz but it is still well above cut-off at 24GHz. I think the New Zealand relays  Iam using are WG22. I simply connected them together directly, by filing slots in the WG20 flange because the hole spacing is different. I used 3 fine screws to tune out the miss-match in the usual manner. It should work OK for narrow bandwidths. If you don't have a 24 GHz VSWR meter or VNA, just use a directional coupler and a sensitive detector (e.g. a spectrum analyser or a crystal detector). The old 22GHz HP8559 analysers work fine at 24GHz, they are un-calibrated but you are only looking for a null.
--
Mike G0MJW

No virus found in this message.
Checked by AVG - www.avg.com
Version: 2016.0.8048 / Virus Database: 4793/15884 - Release Date: 08/14/18
Internal Virus Database is out of date.


--
Paul Bicknell G8KFW   South Coast UK


John Fell
 

Thanks Mike - metalwork is my thing so it can transition outwards to 10mm heating pipe waveguide .

We have used both 27 and 25MHZ Octagon twins for EME at FRARS .

27MHZ have always worked using Leo B dual or Minis but the 25MHZ Xtals are low spec and usually outside injection lock range .Cure is to remove the Xtal at 25MHZ and take in the external reference via the unused F type to the Xtal track revealed under it , when it is lifted .
A good 25MHZ Xtal can then be used in its own housing in shack and injection locked from a Leo B - that provided similar low noise reference as the basic Xtal to the LNB Synth - or use a Morrion OCXO direct .

Polarity changes hinge around 16V .With F types coming out of LNB horizontally and below 16V supply , the polarity will be Horizontal .

73
John
G0API

On Fri, 1 Feb 2019 at 11:52, Mike Willis <willis.mj@...> wrote:
WG22 is recommended for 26GHz-40GHz but it is still well above cut-off at 24GHz. I think the New Zealand relays  Iam using are WG22. I simply connected them together directly, by filing slots in the WG20 flange because the hole spacing is different. I used 3 fine screws to tune out the miss-match in the usual manner. It should work OK for narrow bandwidths. If you don't have a 24 GHz VSWR meter or VNA, just use a directional coupler and a sensitive detector (e.g. a spectrum analyser or a crystal detector). The old 22GHz HP8559 analysers work fine at 24GHz, they are un-calibrated but you are only looking for a null.
--
Mike G0MJW


Chris Bartram G4DGU
 

Hello John

I described a fixed, no amateur twiddle-fiddle, transition which does exactly that - and can be made by hand - in Scatterpoint about a decade ago. I believe that one or two people have duplicated it with good results.  GW3TKF commented in conversation that his had achieved around 30dB RL. That concurs pretty well with my own measurements using a (Cornish made!) Flann coupler, and load. The load shows around -37dB RL btw with little sensitivity to load phase.

Like Mike I use one of the Relcom relays on 24 with WG20, and I have the transition on all three ports. I see about 0.3dB IL for the combination of the relay and transitions in both TX and RX. They are still in use, currently mainly for /P operation, but I hope to resume my interrupted work on 24EME later this year when the house is fully mangled!

73

Chris

G4DGU

On 01/02/2019 11:13, John Fell wrote:
I have a defective WG20 switch and a working WG22 switch , both 4 port Ledex activated at 20V d.c.
My question is , could I use the WG22 device on 24GHZ using suitable adaptors ?
If so has anyone got any data on suitable adaptors ?
--
Chris Bartram
Luxulyan, Bodmin, Cornwall

chris@chris-bartram.co.uk


Paul G8KFW
 

Hi Mike

Is that pool  I am in west Sussex

Paul B


From: UKMicrowaves@groups.io [mailto:UKMicrowaves@groups.io] On Behalf Of John Fell
Sent: 01 February 2019 13:54
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] Waveguide switch replacement

 

Approx location is IO80XS77JS .

 

73   John   G0API

 

On Fri, 1 Feb 2019 at 12:31, Paul Bicknell <paul@...> wrote:

Hi john

I think Mike and I have hade the same positive experience with using WG 22 at 24 Ghz 

Regarding spectrum analysers

Yes the 21 GHz    HP 8559 analysers work fine at 24GHz, as that is what I  used for the 24 ghz  FM  TV link I demonstrated at a BATC convention some 30 plus years ago

The old  HP 18 ghz  in the  140 main frame also works OK at 24 ghz using a WG 22 external mixer    

Regarding power measurements the WG 22  HP power sensors are useable at 24 Ghz  

Just as a side question what is your approximate location  

Paul B 


From: UKMicrowaves@groups.io [mailto:UKMicrowaves@groups.io] On Behalf Of Mike Willis
Sent: 01 February 2019 11:52
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] Waveguide switch replacement

WG22 is recommended for 26GHz-40GHz but it is still well above cut-off at 24GHz. I think the New Zealand relays  Iam using are WG22. I simply connected them together directly, by filing slots in the WG20 flange because the hole spacing is different. I used 3 fine screws to tune out the miss-match in the usual manner. It should work OK for narrow bandwidths. If you don't have a 24 GHz VSWR meter or VNA, just use a directional coupler and a sensitive detector (e.g. a spectrum analyser or a crystal detector). The old 22GHz HP8559 analysers work fine at 24GHz, they are un-calibrated but you are only looking for a null.
--
Mike G0MJW

No virus found in this message.
Checked by AVG - www.avg.com
Version: 2016.0.8048 / Virus Database: 4793/15884 - Release Date: 08/14/18
Internal Virus Database is out of date.

Paul Bicknell G8KFW   South Coast UK

_._,_._,_

 


--
Paul Bicknell G8KFW   South Coast UK


John Fell
 

Thanks Chris ,
Will email Keith and see if he has the original data - spent a few Hrs hand cutting/drilling some flange blanks this afternoon from 6mm copper plate .
73 from the Dorset Snowfields ....err well about 25mm .

John
G0API

On Fri, 1 Feb 2019 at 15:10, Chris Bartram G4DGU <chris@...> wrote:
Hello John

I described a fixed, no amateur twiddle-fiddle, transition which does
exactly that - and can be made by hand - in Scatterpoint about a decade
ago. I believe that one or two people have duplicated it with good
results.  GW3TKF commented in conversation that his had achieved around
30dB RL. That concurs pretty well with my own measurements using a
(Cornish made!) Flann coupler, and load. The load shows around -37dB RL
btw with little sensitivity to load phase.

Like Mike I use one of the Relcom relays on 24 with WG20, and I have the
transition on all three ports. I see about 0.3dB IL for the combination
of the relay and transitions in both TX and RX. They are still in use,
currently mainly for /P operation, but I hope to resume my interrupted
work on 24EME later this year when the house is fully mangled!

73

Chris

G4DGU


On 01/02/2019 11:13, John Fell wrote:
> I have a defective WG20 switch and a working WG22 switch , both 4 port
> Ledex activated at 20V d.c.
> My question is , could I use the WG22 device on 24GHZ using suitable
> adaptors ?
> If so has anyone got any data on suitable adaptors ?
>
--
Chris Bartram
Luxulyan, Bodmin, Cornwall

chris@...





John Fell
 

I am in Corfe Mullen , aprox  5 miles NW of Poole .
John
G0API

On Fri, 1 Feb 2019 at 15:20, Paul Bicknell <paul@...> wrote:

Hi Mike

Is that pool  I am in west Sussex

Paul B


From: UKMicrowaves@groups.io [mailto:UKMicrowaves@groups.io] On Behalf Of John Fell
Sent: 01 February 2019 13:54
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] Waveguide switch replacement

 

Approx location is IO80XS77JS .

 

73   John   G0API

 

On Fri, 1 Feb 2019 at 12:31, Paul Bicknell <paul@...> wrote:

Hi john

I think Mike and I have hade the same positive experience with using WG 22 at 24 Ghz 

Regarding spectrum analysers

Yes the 21 GHz    HP 8559 analysers work fine at 24GHz, as that is what I  used for the 24 ghz  FM  TV link I demonstrated at a BATC convention some 30 plus years ago

The old  HP 18 ghz  in the  140 main frame also works OK at 24 ghz using a WG 22 external mixer    

Regarding power measurements the WG 22  HP power sensors are useable at 24 Ghz  

Just as a side question what is your approximate location  

Paul B 


From: UKMicrowaves@groups.io [mailto:UKMicrowaves@groups.io] On Behalf Of Mike Willis
Sent: 01 February 2019 11:52
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] Waveguide switch replacement

WG22 is recommended for 26GHz-40GHz but it is still well above cut-off at 24GHz. I think the New Zealand relays  Iam using are WG22. I simply connected them together directly, by filing slots in the WG20 flange because the hole spacing is different. I used 3 fine screws to tune out the miss-match in the usual manner. It should work OK for narrow bandwidths. If you don't have a 24 GHz VSWR meter or VNA, just use a directional coupler and a sensitive detector (e.g. a spectrum analyser or a crystal detector). The old 22GHz HP8559 analysers work fine at 24GHz, they are un-calibrated but you are only looking for a null.
--
Mike G0MJW

No virus found in this message.
Checked by AVG - www.avg.com
Version: 2016.0.8048 / Virus Database: 4793/15884 - Release Date: 08/14/18
Internal Virus Database is out of date.

Paul Bicknell G8KFW   South Coast UK


Chris Bartram G4DGU
 

Hello Kent

I've just been deleting some old posts and I noticed your comment about PCB materials. I think you are being a little hard on the reputable PCB manufacturers! Most of the good ones - both those who make specialist microwave laminates, and many of those who produce lower cost materials, do have good data on the dispersive change of Er with frequency, as this now needs to be taken account of in the design of fairly ordinary PCBs. Some modern laminates, such as FR408, now have surprisingly stable and reproducible, Er figures up to 10GHz and even beyond! Losses are not that much improved though - but with care they are acceptable for much work for <=10GHz.

Of course, anyone using stock FR4 is at the mercy of low cost manufacture, but even there, I've been able to a good job in the 10GHz region - although printed filters are out, unless ...  :-) Watch this space when I get a bit more time!

73

Chris

G4DGU


--
Chris Bartram
Luxulyan, Bodmin, Cornwall

chris@chris-bartram.co.uk


KENT BRITAIN
 

Hi Chris

I get about 10-12,000 boards made a year
I have my own files on the Er of the local materials vs frequency and
try to stay with the same PCB houses.

My comments are to the 'Students' so make complex models with
data to the .001% for many materials with Er values they took off the
Internet.      The Rogers and Isola vendors know me well.

But few of the low cost PCB houses have any idea what the factors are
for their materials except for the 1 kHz data.

Just saying, be very careful designing RF boards believing the 1 kHz data
is valid above a MHz.    Kent

On Friday, February 1, 2019, 4:01:20 PM CST, Chris Bartram G4DGU <chris@...> wrote:


Hello Kent

I've just been deleting some old posts and I noticed your comment about
PCB materials. I think you are being a little hard on the reputable PCB
manufacturers! Most of the good ones - both those who make specialist
microwave laminates, and many of those who produce lower cost materials,
do have good data on the dispersive change of Er with frequency, as this
now needs to be taken account of in the design of fairly ordinary PCBs.
Some modern laminates, such as FR408, now have surprisingly stable and
reproducible, Er figures up to 10GHz and even beyond! Losses are not
that much improved though - but with care they are acceptable for much
work for <=10GHz.

Of course, anyone using stock FR4 is at the mercy of low cost
manufacture, but even there, I've been able to a good job in the 10GHz
region - although printed filters are out, unless ...  :-) Watch this
space when I get a bit more time!

73

Chris

G4DGU


--
Chris Bartram
Luxulyan, Bodmin, Cornwall

chris@...





Chris Bartram G4DGU
 

Hello, again, Kent.

My comments are to the 'Students' so make complex models with
data to the .001% for many materials with Er values they took off the
Internet.      The Rogers and Isola vendors know me well.
I couldn't agree more! CAE can be a very satisfying, but ultimately misleading, way of wasting time! It requires a lot of experience and knowledge to use the software effectively.

But few of the low cost PCB houses have any idea what the factors are
for their materials except for the 1 kHz data.
Don't I know!! Again, circumspection is needed ...

Just saying, be very careful designing RF boards believing the 1 kHz data
is valid above a MHz.
Having learned how to use a slotted line for making microwave dielectric measurements when I worked (designing test equipment) for an antenna manufacturer back in the '70s I'm very aware of the pitfalls. However, it is possible to design reliably on bog-standard FR4 at microwaves, providing you are not too ambitious!

73

Chris

G4DGU

--
Chris Bartram
Luxulyan, Bodmin, Cornwall

chris@chris-bartram.co.uk


KENT BRITAIN
 

I have used fiberglass boards up to 35 GHz, but it was very thin
and a low Q circuit.  With a very low Q, frequency shift and losses
are quite minimal.

When it comes to modeling, I must quote Tom Clark, ex-president of AMSAT.

"Why be approximately correct when you can be precisely wrong!"



On Friday, February 1, 2019, 6:26:36 PM CST, Chris Bartram G4DGU <chris@...> wrote:


Hello, again, Kent.

> My comments are to the 'Students' so make complex models with
> data to the .001% for many materials with Er values they took off the
> Internet.      The Rogers and Isola vendors know me well.

I couldn't agree more! CAE can be a very satisfying, but ultimately
misleading, way of wasting time! It requires a lot of experience and
knowledge to use the software effectively.

> But few of the low cost PCB houses have any idea what the factors are
> for their materials except for the 1 kHz data.

Don't I know!! Again, circumspection is needed ...

> Just saying, be very careful designing RF boards believing the 1 kHz data
> is valid above a MHz.

Having learned how to use a slotted line for making microwave dielectric
measurements when I worked (designing test equipment) for an antenna
manufacturer back in the '70s I'm very aware of the pitfalls. However,
it is possible to design reliably on bog-standard FR4 at microwaves,
providing you are not too ambitious!

73

Chris

G4DGU

--
Chris Bartram
Luxulyan, Bodmin, Cornwall

chris@...





Mike Willis
 

On Fri, Feb 1, 2019 at 04:26 PM, Chris Bartram G4DGU wrote:
Having learned how to use a slotted line for making microwave dielectric measurements when I worked (designing test equipment) for an antenna manufacturer back in the '70s I'm very aware of the pitfalls. However, it is possible to design reliably on bog-standard FR4 at microwaves, providing you are not too ambitious!
You can always meet the requirements if you are free to change the specification in any way you wish.
 
--
Mike G0MJW


Richard <richard@...>
 

What's very important is to teach students to be able to identify
where their designs have gone wrong.
Variations in pcbs has always been a pain in early stages of new design
to manufacture, and how many have heard the phrase" it worked OK in
simulation" and found the person uttering the comment hasn't got a clue
to sort it out.
I'm sure Chris has been there

On Sat, 2019-02-02 at 00:26 +0000, Chris Bartram G4DGU wrote:
Hello, again, Kent.

My comments are to the 'Students' so make complex models with
data to the .001% for many materials with Er values they took off
the
Internet. The Rogers and Isola vendors know me well.
I couldn't agree more! CAE can be a very satisfying, but ultimately
misleading, way of wasting time! It requires a lot of experience and
knowledge to use the software effectively.

But few of the low cost PCB houses have any idea what the factors
are
for their materials except for the 1 kHz data.
Don't I know!! Again, circumspection is needed ...

Just saying, be very careful designing RF boards believing the 1
kHz data
is valid above a MHz.
Having learned how to use a slotted line for making microwave
dielectric
measurements when I worked (designing test equipment) for an antenna
manufacturer back in the '70s I'm very aware of the pitfalls.
However,
it is possible to design reliably on bog-standard FR4 at microwaves,
providing you are not too ambitious!

73

Chris

G4DGU


Mike Willis
 

We have all been there and many seem to think that when the theory and measurement diverge, it is the measurement that is wrong.
--
Mike G0MJW


Ian White
 

Generalizing even further, "If theory and practice diverge, you need more and better theory, more and better practice; or very probably, both."

73 from Ian GM3SEK
On 02/02/2019 10:22, Mike Willis wrote:

We have all been there and many seem to think that when the theory and measurement diverge, it is the measurement that is wrong.
--
Mike G0MJW


alwyn.seeds1
 

Dear All,

It’s Ian’s maxim, I’ll go with. At work we have been comparing measured with modelled (CST) photodiode parameters above 60 GHz and the main uncertainty is with the VNA coplanar calibration standards, despite their very considerable price tag. If you need to work in CPW above 100 GHz things get worse.

Regards,

Alwyn


From: 
Ian White
Date: Sat, 02 Feb 2019 02:37:19 PST 

Generalizing even further, "If theory and practice diverge, you need more and better theory, more and better practice; or very probably, both."

73 from Ian GM3SEK

_____________________________________________________

Alwyn Seeds, Director
SynOptika Ltd.,
114 Beaufort Street,
London,
SW3 6BU,
England.

Tel.: +44 (0) 20 7376 4110


SynOptika Ltd., Registered in England and Wales: No. 04606737
Registered Office: 114 Beaufort Street, London, SW3 6BU, United Kingdom.
_____________________________________________________