Topics

Anyone got any spare WG15 / WR112 waveguide


g4cch_1
 
Edited

I'm looking for a short length of WG15/WR112 to make an evanescent mode filter for 13cm

If anyone has some to spare, please email me off line.

Thanks and 73
Howard, G4CCH


Andy G4JNT
 

I think I'd be more inclined to look at WG14 for that rather than WG15.

At 3.4GHz WG16 was at about the limit that could be used, screw penetration was nearly right across the guide - although it worked very well indeed, and I've made three repeatable units for that band.  http://g4jnt.com/EVANFILT.pdf

Looking at dimensions, linear scaling suggests WG14 wouldn't be too far off the mark

It would be nice to get a filter that covers 2300 - 2420 with a good transition band - especially on the LF side for image rejection.   SO may be worth looking at a 4th order design.  
Evanescent filters are so easy to throw together, and drilling too many tapped holes while just playing around wont hurt the final result.



On Fri, 29 Jan 2021 at 21:35, g4cch_1 <g4cch_1@...> wrote:
I'm looking for a short length of WG15 WR112 to make an evanescent mode filter for 13cm

If anyone has some to spare, please email me off line.

Thanks and 73
Howard, G4CCH


g4cch_1
 
Edited

Thanks Andy, and I will take your advice...

I build stuff, and experiment, even though I don't always know what I am doing.
Eventually I get there, well most of the time...

So I am looking for WG14 (WR137), but will take WG15 (WR112) if anyone has any to spare.
I have some WG16, but not enough to waste, so may need some more of anyone has some to spare.

I built a filter using WG17 just because I had some, and used dimensions for this QEX article:
http://www.w1ghz.org/filter/QEX_Mar_2010_p23-29.pdf

I realised later that the WG17 was for a 5.7GHz filter.

Anyway, the filter did eventually tune down to 2.3GHz, but I had to add some concentric tubes to the face opposite the tuning screws.
The filter looks like this:
https://g4cch.com//images/13cm/Evanescent_%20mode_filter/WG17/20210130_105850_1.jpg

https://g4cch.com//images/13cm/Evanescent_%20mode_filter/WG17/20210130_105911_1.jpg

The tuning screws and nuts came from some surplus Nokia filters the at I have.

This was a useful project to try out my nanovna v2 plus 4, the passband looks like this:
https://g4cch.com//images/13cm/Evanescent_%20mode_filter/WG17/20210130_111150_1.jpg

The filter is too narrow for my application, and it's too lossy.
As you say - it would be nice to get a filter that covers 2300 - 2420 with a good transition band.
Actually, I don't need to worry about image rejection, as I have a 7 pole filter on the 13cm side of the mixer.

The evanescent mode filter is intended to be placed between my G4DDK VLNA (at feed point) and the RF stage (mga86576) in my transverter.
This is part of an effort to reduce wifi qrm which I can see on 2320, and of course much worse on 2400.
I'm sure that the RF stage in the transverter is being overloaded, and tests using a pipecap filter proved that was the case, and the qrm could be eliminated.

In addition to the above, I changed the RF stage in the transverter to a PHA-83w, and this has almost eliminated the problem with out adding any extra filtering.
No qrm on 2320, 2300, 2304, 2301, and greatly reduced on 2400 (was S9 on K3S meter). I can now receive JA's again on 2400 EME.

Almost there, but would like to build another filter using the right waveguide this time.

Thanks and 73
Howard, G4CCH



John Fell
 

Nice plots Howard .
I dug out some EVOs that I made for various EME related projects .Couple of pics below showing 2400MHz in WG16 , 3400MHz in WG 16 and the 5700MHz one that Andy made up for me in WR28 .

The insertion loss of the 2400MHz was 8dB , which between a DDK first stage and  an external secondary stage at EME feed area , would be fine .

Much lower insertion loss on 3400MHz and 5.7GHz , so Ok for LO and RF duty.

73
John
G0API

It am raining here too

On Sat, 30 Jan 2021 at 12:19, g4cch_1 <g4cch_1@...> wrote:

[Edited Message Follows]

Thanks Andy, and I will take your advice...

I build stuff, and experiment, even though I don't always know what I am doing.
Eventually I get there, well most of the time...

So I am looking for WG14 (WR137), but will take WG15 (WR112) if anyone has any to spare.
I have some WG16, but not enough to waste, so may need some more of anyone has some to spare.

I built a filter using WG17 just because I had some, and used dimensions for this QEX article:
http://www.w1ghz.org/filter/QEX_Mar_2010_p23-29.pdf

I realised later that the WG17 was for a 5.7GHz filter.

Anyway, the filter did eventually tune down to 2.3GHz, but I had to add some concentric tubes to the face opposite the tuning screws.
The filter looks like this:
https://g4cch.com/images/13cm/Evanescent_%20mode_filter/WG17/20210130_105850_1.jpg
https://g4cch.com//images/13cm/Evanescent_%20mode_filter/WG17/20210130_105911_1.jpg
The tuning screws and nuts came from some surplus Nokia filters the at I have.

This was a useful project to try out my nanovna v2 plus 4, the passband looks like this:
https://g4cch.com//images/13cm/Evanescent_%20mode_filter/WG17/20210130_111150_1.jpg

The filter is too narrow for my application, and it's too lossy.
As you say - it would be nice to get a filter that covers 2300 - 2420 with a good transition band.
Actually, I don't need to worry about image rejection, as I have a 7 pole filter on the 13cm side of the mixer.

The evanescent mode filter is intended to be placed between my G4DDK VLNA (at feed point) and the RF stage (mga86576) in my transverter.
This is part of an effort to reduce wifi qrm which I can see on 2320, and of course much worse on 2400.
I'm sure that the RF stage in the transverter is being overloaded, and tests using a pipecap filter proved that was the case, and the qrm could be eliminated.

In addition to the above, I changed the RF stage in the transverter to a PHA-83w, and this has almost eliminated the problem with out adding any extra filtering.
No qrm on 2320, 2300, 2304, 2301, and greatly reduced on 2400 (was S9 on K3S meter). I can now receive JA's again on 2400 EME.

Almost there, but would like to build another filter using the right waveguide this time.

Thanks and 73
Howard, G4CCH



g4cch_1
 
Edited

I'm in touch with Mark GM4ISM about WG14, so no need to look further.

Howard, G4CCH