10GHz SMA to WG16


Gareth G4XAT
 

I'm hoping to build a SMA to WG16 launcher/feeder by adapting an old Gunn diode radar housing. I have read the article in August 07 Radcom which has dimensions ect, and I have some brass WG16 and an end-stop along with suitable SMA long-pin to be the probe. I'm wondering of my two options which is best. i.e. adapt the Gunn diode housing to SMA probe feed or fit a SMA to the bare brass WG16 which would then need a flange making/fitting for horn mount. Both will also bolt up to a conical WG16 to 22mm copper dish feed that I bought about 20 years ago from Bob Platts. I have a new BSB small dish that it can fir directly to.
I'd assume that the Gunn diode is in the 'right place' already in the WG16 housing. It also has a tuning screw and a detector diode (might this be useful for measuring my own power - only mW at the moment from a Pluto SDR). I have another Gunn cavity with a good size horn , tuning screw and an 'aperture' (not sure what it does). Photos supplement the above....
So which to build or modify?
Guidance and suggestions very welcome...
Thanks
Gareth


 

On Tue, 8 Jun 2021 at 14:43, Gareth G4XAT via groups.io <g4xat=ntlworld.com@groups.io> wrote:
I'm hoping to build a SMA to WG16 launcher/feeder by adapting an old Gunn diode radar housing. I have read the article in August 07 Radcom which has dimensions ect, and I have some brass WG16 and an end-stop along with suitable SMA long-pin to be the probe. I'm wondering of my two options which is best. i.e. adapt the Gunn diode housing to SMA probe feed or fit a SMA to the bare brass WG16 which would then need a flange making/fitting for horn mount. Both will also bolt up to a conical WG16 to 22mm copper dish feed that I bought about 20 years ago from Bob Platts. I have a new BSB small dish that it can fir directly to.
I'd assume that the Gunn diode is in the 'right place' already in the WG16 housing. It also has a tuning screw and a detector diode (might this be useful for measuring my own power - only mW at the moment from a Pluto SDR). I have another Gunn cavity with a good size horn , tuning screw and an 'aperture' (not sure what it does). Photos supplement the above....
So which to build or modify?
Guidance and suggestions very welcome...
Thanks
Gareth
I would expect the Gunn diode to have a major effect on the cavity, so its removal, and replacement by an SMA connector, is in my opinion unlikely to be optimal. That said, if you have a  tuning screw, it may be possible to compensate.

I recall reading an article in an amateur radio publication about someone combining the output of two Gunn diodes and getting more power than the sum of the two individual diodes. I recall discussing this with a couple of people at a meeting held at Keysight. One of the people was an academic building THz Gunn diodes. His thoughts were it was utter crap - totally impossible. Perpetual motion. Another person I spoke to about this was someone from Virginia diodes, who did make THz extenders for Keysight VNAs. He was not surprised by this at all and said that by combining the output of two diodes you have completely changed the cavity they are working in. So therefore if you completely change the cavity, you would expect things to behave very differently. It's his comments that make me feel that removing Gunn diodes and detectors, and expecting an SMA connector to work in the same position, is just hoping for too much. You might be lucky, but are more likely to be unlucky.

I would not think trying to measure your own power in the cavity with the detector diode is a good idea. I would couple some of the RF out some distance from the transition.

I don't know what the RadCom article is like, but W1GHz done a lot of work on optimisation of transitions using HFSS. That would be worth looking at.

Dave


John E. Beech
 

I could imagine a scenario where two Gunn diodes give more power than the sum - one could be acting as an
oscillator and the second as an amplifier with perhaps a slightly different bias on it giving a combined output of more
than two oscillators.

de John G8SEQ

-------Original Message-------
From: Dr. David Kirkby, Kirkby Microwave Ltd <drkirkby@kirkbymicrowave.co.uk>
To: UKMicrowaves@groups.io <UKMicrowaves@groups.io>
Subject: Re: [UKMicrowaves] 10GHz SMA to WG16
Sent: Jun 08 '21 20:37

On Tue, 8 Jun 2021 at 14:43, Gareth G4XAT via groups.io
<g4xat=ntlworld.com@groups.io> wrote:

> I'm hoping to build a SMA to WG16 launcher/feeder by adapting an old
> Gunn diode radar housing. I have read the article in August 07
> Radcom which has dimensions ect, and I have some brass WG16 and an
> end-stop along with suitable SMA long-pin to be the probe. I'm
> wondering of my two options which is best. i.e. adapt the Gunn diode
> housing to SMA probe feed or fit a SMA to the bare brass WG16 which
> would then need a flange making/fitting for horn mount. Both will
> also bolt up to a conical WG16 to 22mm copper dish feed that I
> bought about 20 years ago from Bob Platts. I have a new BSB small
> dish that it can fir directly to.
> I'd assume that the Gunn diode is in the 'right place' already in
> the WG16 housing. It also has a tuning screw and a detector diode
> (might this be useful for measuring my own power - only mW at the
> moment from a Pluto SDR). I have another Gunn cavity with a good
> size horn , tuning screw and an 'aperture' (not sure what it does).
> Photos supplement the above....
> So which to build or modify?
> Guidance and suggestions very welcome...
> Thanks
> Gareth

I would expect the Gunn diode to have a major effect on the cavity, so
its removal, and replacement by an SMA connector, is in my opinion
unlikely to be optimal. That said, if you have a tuning screw, it may
be possible to compensate.

I recall reading an article in an amateur radio publication about
someone combining the output of two Gunn diodes and getting more power
than the sum of the two individual diodes. I recall discussing this
with a couple of people at a meeting held at Keysight. One of the
people was an academic building THz Gunn diodes. His thoughts were it
was utter crap - totally impossible. Perpetual motion. Another person
I spoke to about this was someone from Virginia diodes, who did make
THz extenders for Keysight VNAs. He was not surprised by this at all
and said that by combining the output of two diodes you have
COMPLETELY changed the cavity they are working in. So therefore if you
completely change the cavity, you would expect things to behave very
differently. It's his comments that make me feel that removing Gunn
diodes and detectors, and expecting an SMA connector to work in the
same position, is just hoping for too much. You might be lucky, but
are more likely to be unlucky.

I would not think trying to measure your own power in the cavity with
the detector diode is a good idea. I would couple some of the RF out
some distance from the transition.

I don't know what the RadCom article is like, but W1GHz done a lot of
work on optimisation of transitions using HFSS. That would be worth
looking at.

Dave

>


Wilko
 

I would personally go for the "brass route". I have
built a number of these sma-wg transitions. Measurements have shown that they are not at all bad. I lack equipment to measure things like
S11 at this frequency so I had to work based on mechanical dimensions and hope for the best.

Obviously the measurement results led to some smugness on my part :-)

Wilko
PA1WBU


Gareth G4XAT
 

Brass it was, only room for two tuning screws, one opposite the probe (hole already there) and one 9.3mm upstream. Working well across the shack. Going to try the garden next. Any excuse for a bit of sunshine....
Gareth