Retuning a 23cm duplexer, any hints please?


Alan Beard
 

Hi all,

I've just attempted to retune this duplexer from Radio Frequency Systems P/L.
Model: CP1800-I42

Originally on 1710 and 1910 MHz. It tuned down to 1273 and 1293 MHz OK but
kept the same "channel" bandwidth of about 20 MHz. (the width of the pass band)

So, it's skirts only attenuate 45db or so the Tx carrier and/or the
Tx spurii.

I note the inter digital coupling screws do nothing and I mean nothing
to reduce the coupling and thus the filter bandwidth. (the three on the
left and right ends of the top cover)

Has anyone any thoughts on making it useful for our 20MHz split 23cm
repeaters?


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


KENT BRITAIN
 

For what it's worth, we often retune 450-470 MHz duplexers to 430 MHz or so, then fine tune them on their 3rd Harmonic.   Many of the Texas 23 cm repeaters are using UHF duplexers. 

On Sunday, April 4, 2021, 12:44:57 AM CDT, Alan Beard via groups.io <beardal@...> wrote:


Hi all,

I've just attempted to retune this duplexer from Radio Frequency Systems P/L.
Model: CP1800-I42

Originally on 1710 and 1910 MHz. It tuned down to 1273 and 1293 MHz OK but
kept the same "channel" bandwidth of about 20 MHz. (the width of the pass band)

So, it's skirts only attenuate 45db or so the Tx carrier and/or the
Tx spurii.

I note the inter digital coupling screws do nothing and I mean nothing
to reduce the coupling and thus the filter bandwidth. (the three on the
left and right ends of the top cover)

Has anyone any thoughts on making it useful for our 20MHz split 23cm
repeaters?


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







Robin Szemeti - G1YFG
 

Depending on how it is set up, you may find the top screws are the "tuning" screws and the side screws are the coupling scews ... there are a multitude of different setups for duplexers, but if you put a sweeper on it or an analyzer with a tracking generator, you should find that as well as the pass band peaks, there may also be notches ... you may find tweaking the coupling screws effects the notches while the tuning screws shift the passbands. The passbands are usually fairly broad, tweaking those is simple enough, althoug a VNA will help you optimise the match ... the notches (if the design has notches ... ) are usually very sharp and sensitive!



On Sun, 4 Apr 2021 at 11:58, KENT BRITAIN <WA5VJB@...> wrote:
For what it's worth, we often retune 450-470 MHz duplexers to 430 MHz or so, then fine tune them on their 3rd Harmonic.   Many of the Texas 23 cm repeaters are using UHF duplexers. 

On Sunday, April 4, 2021, 12:44:57 AM CDT, Alan Beard via groups.io <beardal=unixservice.com.au@groups.io> wrote:


Hi all,

I've just attempted to retune this duplexer from Radio Frequency Systems P/L.
Model: CP1800-I42

Originally on 1710 and 1910 MHz. It tuned down to 1273 and 1293 MHz OK but
kept the same "channel" bandwidth of about 20 MHz. (the width of the pass band)

So, it's skirts only attenuate 45db or so the Tx carrier and/or the
Tx spurii.

I note the inter digital coupling screws do nothing and I mean nothing
to reduce the coupling and thus the filter bandwidth. (the three on the
left and right ends of the top cover)

Has anyone any thoughts on making it useful for our 20MHz split 23cm
repeaters?


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







--
Robin Szemeti - G1YFG


Alan Beard
 

Thanks Kent,

We have a "parrot" 23cm repeater over in the Wollongong area and changing it to Full duplex will require
cavity filters. I acquired this one, at a good price a few years ago. Perhaps spares never used.
Using 70cm cavities in 3rd overtone is a great idea. 

Robin, 

My test gear here is the new VNA the SAA V2 and a Wiltron 640 from around 1985. What I'm asking is:
Noting that both sides are just bandpass. And, as this inter-digital design is common.

How does one reduce the coupling so as to reduce the bandwidth?

As mentioned, the top screws do absolutely nothing, an imperceptible change.
I need to achieve somewhere between 5 and 10 MHz, not 20MHz.

Has anyone here any experience, either practical or with simulation?

BTW: The SAA V2 is magic.

Also, I don't know who the designer was in RFS but it's a bugger to tune with those M3 brass screws
wobbling about inside the finger ends with only 1/2mm clearance or less.

Keep smiling

Alan VK2ZIW


On Sun, 4 Apr 2021 15:28:59 +0100, Robin Szemeti - G1YFG wrote
> Depending on how it is set up, you may find the top screws are the "tuning" screws and the side screws are the coupling scews ... there are a multitude of different setups for duplexers, but if you put a sweeper on it or an analyzer with a tracking generator, you should find that as well as the pass band peaks, there may also be notches ... you may find tweaking the coupling screws effects the notches while the tuning screws shift the passbands. The passbands are usually fairly broad, tweaking those is simple enough, althoug a VNA will help you optimise the match ... the notches (if the design has notches ... ) are usually very sharp and sensitive!
>
> On Sun, 4 Apr 2021 at 11:58, KENT BRITAIN <WA5VJB@...> wrote:
>

>
> For what it's worth, we often retune 450-470 MHz duplexers to 430 MHz or so, then fine tune them on their 3rd Harmonic.   Many of the Texas 23 cm repeaters are using UHF duplexers. 
>
> On Sunday, April 4, 2021, 12:44:57 AM CDT, Alan Beard via groups.io <beardal=unixservice.com.au@groups.io> wrote:
>
> Hi all,
>
> I've just attempted to retune this duplexer from Radio Frequency Systems P/L.
>
> Model: CP1800-I42
>
> Originally on 1710 and 1910 MHz. It tuned down to 1273 and 1293 MHz OK but
>
> kept the same "channel" bandwidth of about 20 MHz. (the width of the pass band)
>
> So, it's skirts only attenuate 45db or so the Tx carrier and/or the
>
> Tx spurii.
>
> I note the inter digital coupling screws do nothing and I mean nothing
>
> to reduce the coupling and thus the filter bandwidth. (the three on the
>
> left and right ends of the top cover)
>
> Has anyone any thoughts on making it useful for our 20MHz split 23cm
>
> repeaters?
>
> ---------------------------------------------------
>
> Alan VK2ZIW
>
> Before the Big Bang, God, Sela.
>
> OpenWebMail 2.53, nothing in the cloud.
>
>

> --
> Robin Szemeti - G1YFG

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


KENT BRITAIN
 

Interestingly the 3rd harmonic is not exactly 3X the 430 MHz frequency, but close.
Often easier to do the first tuning at the 430 MHz frequency,m then fine tune at the 23 cm frequency.

Quite often we see the 450 MHz filters from commercial mobile radios surplus.  Most of those tune 23 cm as well.   Good luck with your project,  Kent WA5VJB/G8EMY

On Sunday, April 4, 2021, 4:43:42 PM CDT, Alan Beard via groups.io <beardal@...> wrote:


Thanks Kent,

We have a "parrot" 23cm repeater over in the Wollongong area and changing it to Full duplex will require
cavity filters. I acquired this one, at a good price a few years ago. Perhaps spares never used.
Using 70cm cavities in 3rd overtone is a great idea. 

Robin, 

My test gear here is the new VNA the SAA V2 and a Wiltron 640 from around 1985. What I'm asking is:
Noting that both sides are just bandpass. And, as this inter-digital design is common.

How does one reduce the coupling so as to reduce the bandwidth?

As mentioned, the top screws do absolutely nothing, an imperceptible change.
I need to achieve somewhere between 5 and 10 MHz, not 20MHz.

Has anyone here any experience, either practical or with simulation?

BTW: The SAA V2 is magic.

Also, I don't know who the designer was in RFS but it's a bugger to tune with those M3 brass screws
wobbling about inside the finger ends with only 1/2mm clearance or less.

Keep smiling

Alan VK2ZIW


On Sun, 4 Apr 2021 15:28:59 +0100, Robin Szemeti - G1YFG wrote
> Depending on how it is set up, you may find the top screws are the "tuning" screws and the side screws are the coupling scews ... there are a multitude of different setups for duplexers, but if you put a sweeper on it or an analyzer with a tracking generator, you should find that as well as the pass band peaks, there may also be notches ... you may find tweaking the coupling screws effects the notches while the tuning screws shift the passbands. The passbands are usually fairly broad, tweaking those is simple enough, althoug a VNA will help you optimise the match ... the notches (if the design has notches ... ) are usually very sharp and sensitive!
>
> On Sun, 4 Apr 2021 at 11:58, KENT BRITAIN <WA5VJB@...> wrote:
>

>
> For what it's worth, we often retune 450-470 MHz duplexers to 430 MHz or so, then fine tune them on their 3rd Harmonic.   Many of the Texas 23 cm repeaters are using UHF duplexers. 
>
> On Sunday, April 4, 2021, 12:44:57 AM CDT, Alan Beard via groups.io <beardal=unixservice.com.au@groups.io> wrote:
>
> Hi all,
>
> I've just attempted to retune this duplexer from Radio Frequency Systems P/L.
>
> Model: CP1800-I42
>
> Originally on 1710 and 1910 MHz. It tuned down to 1273 and 1293 MHz OK but
>
> kept the same "channel" bandwidth of about 20 MHz. (the width of the pass band)
>
> So, it's skirts only attenuate 45db or so the Tx carrier and/or the
>
> Tx spurii.
>
> I note the inter digital coupling screws do nothing and I mean nothing
>
> to reduce the coupling and thus the filter bandwidth. (the three on the
>
> left and right ends of the top cover)
>
> Has anyone any thoughts on making it useful for our 20MHz split 23cm
>
> repeaters?
>
> ---------------------------------------------------
>
> Alan VK2ZIW
>
> Before the Big Bang, God, Sela.
>
> OpenWebMail 2.53, nothing in the cloud.
>
>

> --
> Robin Szemeti - G1YFG

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


Alan Beard
 

Hi all,
Here are my mods to this 1800 MHz duplexer so far:
> www.unixservice.com.au/hamradio/repeaters/duplexers/23CMpix/
IMG_3809c.JPG
Tuning is now much easier with the M3 screws further out, out from the 4mm ID tube.

My next question is:

Can the inter-digital filter be made BandPass and BandReject?

(as we do for 2m repeaters) (photos, drawings please)

Our 20 MHz split for 23cm repeaters is pretty tight.

I see nothing on the web.

Alan VK2ZIW


On Sun, 4 Apr 2021 22:05:14 +0000 (UTC), KENT BRITAIN wrote
> Interestingly the 3rd harmonic is not exactly 3X the 430 MHz frequency, but close.
> Often easier to do the first tuning at the 430 MHz frequency,m then fine tune at the 23 cm frequency.
>
> Quite often we see the 450 MHz filters from commercial mobile radios surplus.  Most of those tune 23 cm as well.   Good luck with your project,  Kent WA5VJB/G8EMY
>
> On Sunday, April 4, 2021, 4:43:42 PM CDT, Alan Beard via groups.io <beardal@...> wrote:
>
> Thanks Kent,
>

> We have a "parrot" 23cm repeater over in the Wollongong area and changing it to Full duplex will require
> cavity filters. I acquired this one, at a good price a few years ago. Perhaps spares never used.
> Using 70cm cavities in 3rd overtone is a great idea. 
>



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


Alan Beard
 

Hi all,

Thomas Eddison 'tinkered' until the breakthrough...

The figure of 8 loop did the trick.


Where are the 'tinkerers' out there????

I've achieved a super steep slope on the high side of a pass, now
to see if I can get another super steep slope on the lower frequency 
side of the high pass. 

Come on guys, where's your imagination?

Alan VK2ZIW

On Tue, 6 Apr 2021 07:44:42 +1000, Alan Beard via groups.io wrote
> Hi all,
> Here are my mods to this 1800 MHz duplexer so far:
>
> > www.unixservice.com.au/hamradio/repeaters/duplexers/23CMpix/
> IMG_3809c.JPG
> Tuning is now much easier with the M3 screws further out, out from the 4mm ID tube.
>
> My next question is:
>
> Can the inter-digital filter be made BandPass and BandReject?
>
> (as we do for 2m repeaters) (photos, drawings please)
>
> Our 20 MHz split for 23cm repeaters is pretty tight.
>
> I see nothing on the web.
>
> Alan VK2ZIW
>
>

>
>

> On Sun, 4 Apr 2021 22:05:14 +0000 (UTC), KENT BRITAIN wrote
> > Interestingly the 3rd harmonic is not exactly 3X the 430 MHz frequency, but close.
> > Often easier to do the first tuning at the 430 MHz frequency,m then fine tune at the 23 cm frequency.
> >
> > Quite often we see the 450 MHz filters from commercial mobile radios surplus.  Most of those tune 23 cm as well.   Good luck with your project,  Kent WA5VJB/G8EMY
> >
> > On Sunday, April 4, 2021, 4:43:42 PM CDT, Alan Beard via groups.io <beardal@...> wrote:
> >
> > Thanks Kent,
> >

> > We have a "parrot" 23cm repeater over in the Wollongong area and changing it to Full duplex will require
> > cavity filters. I acquired this one, at a good price a few years ago. Perhaps spares never used.
> > Using 70cm cavities in 3rd overtone is a great idea. 
> >

>
> ---------------------------------------------------
> 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 all,

I gave up on the BandPass-BandReject and went with 25mm x 25mm dividers
as in the attached pixture. This reduced the pass bandwidth to 10MHz centered on
1274 MHz and 1294 MHz with 50db or so attenuation of the other frequency.
The input RL is -16db or better.

Measured by the SAA-V2.

Alan VK2ZIW

On Wed, 7 Apr 2021 07:28:52 +1000, Alan Beard via groups.io wrote
> Hi all,
>
>

> Thomas Eddison 'tinkered' until the breakthrough...
>
>

> The figure of 8 loop did the trick.
>
>

>
>

> Where are the 'tinkerers' out there????
>
> I've achieved a super steep slope on the high side of a pass, now
> to see if I can get another super steep slope on the lower frequency 
> side of the high pass. 
>
>

> Come on guys, where's your imagination?
>
>

> Alan VK2ZIW
>
>

> On Tue, 6 Apr 2021 07:44:42 +1000, Alan Beard via groups.io wrote
> > Hi all,
> > Here are my mods to this 1800 MHz duplexer so far:
> >
> > > www.unixservice.com.au/hamradio/repeaters/duplexers/23CMpix/
> > IMG_3809c.JPG
> > Tuning is now much easier with the M3 screws further out, out from the 4mm ID tube.
> >
> > My next question is:
> >


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