G/T for parabolic dishes


John Lemay
 

One thing I like about Dubus magazine is that I can revisit old copies and
find articles of interest, which I the time of publication I pretty much
ignored.

In issue 2/2015 there's a paper by W1GHZ which examines the potential for
improving G/T by under illuminating the parabolic surface. In 2015 I was
using yagis for 23cms - now I'm using a dish and the information is very
interesting indeed.

All of the information presented relates to EME use of a dish. My question
relates to improving G/T for tropo use - is it worthwhile ? Given that
arguably the dish is seeing half and half sky noise and ground noise, maybe
the answer to my question is "yes, but only half the improvement" ?

Comments and views welcome !

John G4ZTR



--
This email has been checked for viruses by AVG.
https://www.avg.com


Chris Bartram G4DGU
 

Hello John,

Analytically there isn't any real difference between Yagi-Uda antennas, dishes (and optical telescopes ... look-up 'Airy Disc') when it comes to optimising performance. They are simply different ways of focussing the E-M energy towards a distant point. You get optimum 'gain' when the first sidelobes are close to -17.5dB. Reducing the sidelobe levels is possible, but at the expense of gain and although some of the amateur antenna designers and manufacturers have been playing numbers games with G/T ratio, but I can't see how they escape from the constraints of the physics! The advantage of dishes as you go up in frequency is that they are less critical in terms of dimensions, and there is less to go wrong! They are also a LOT easier to design. Compared to dishes, y-u antennas are a MTBF nightmare ...

In the amateur literature, the late K2RIW wrote a very nice paper on this aeons ago ('MAXIMUM DISH EFFICIENCY AND THE BEST ANTENNA SIDELOBE LEVELS'' Dick Knadle, K2RIW 6/04/05') I think it can be found on the net for download. I have a copy, but I'm not sure of the copyright issues so I won't distribute it. Also the text mentioned by K2RIW - Hect's 'Optics' - is also worth a look. But be prepared for some heavy maths! It gave me some serious headaches!

Matching the feed design to the dish so as to make the edge illumination of around -10dB will give the best G/T - which it is why it's the usual choice of antenna designers. Increasing the feed illumination taper to -11 or -12dB can provide some noise advantage if you have a bit too much dish gain; I'd love to be in that state one day!

73

Chris G4DGU


Robin Szemeti - G1YFG
 

My take on the matter is that it has changed in the last 20 years, due to improvements in receiver front end design.  

20+ years ago, the focus (if you will pardon the pun) was on gain,  as many receiver LNA devices were relatively noisy and you could not hear that much ground/sky noise easily.  QRO was relatively difficult to generate and therefore a lot of the antennas were optimised for gain, to deliver as much signal as possible at the target, and the additional sidelobes did not generate a significant degradation to the overall receiver noise performance.

These days, power is cheap and easy, and receiver front end noise figures have improved, and the additional noise contributed by the sidelobes is audible and significant.  It is possible to sacrifice a little gain and optimise for receive noise floor, and minimum sidelobes is now the design goal.

The physics of course has not changed, but maybe the design goalposts have moved a little?


On Thu, 16 Jul 2020 at 10:41, Chris Bartram G4DGU <chris@...> wrote:
Hello John,

Analytically there isn't any real difference between Yagi-Uda antennas,
dishes (and optical telescopes ... look-up 'Airy Disc') when it comes to
optimising performance. They are simply different ways of focussing the
E-M energy towards a distant point. You get optimum 'gain' when the
first sidelobes are close to -17.5dB. Reducing the sidelobe levels is
possible, but at the expense of gain and although some of the amateur
antenna designers and manufacturers have been playing numbers games with
G/T ratio, but I can't see how they escape from the constraints of the
physics! The advantage of dishes as you go up in frequency is that they
are less critical in terms of dimensions, and there is less to go wrong!
They are also a LOT easier to design. Compared to dishes, y-u antennas
are a MTBF nightmare ...

In the amateur literature, the late K2RIW wrote a very nice paper on
this aeons ago ('MAXIMUM DISH EFFICIENCY AND THE BEST ANTENNA SIDELOBE
LEVELS'' Dick Knadle, K2RIW 6/04/05') I think it can be found on the net
for download. I have a copy, but I'm not sure of the copyright issues so
I won't distribute it. Also the text mentioned by K2RIW - Hect's
'Optics' - is also worth a look. But be prepared for some heavy maths!
It gave me some serious headaches!

Matching the feed design to the dish so as to make the edge illumination
of around -10dB will give the best G/T - which it is why it's the usual
choice of antenna designers. Increasing the feed illumination taper to
-11 or -12dB can provide some noise advantage if you have a bit too much
dish gain; I'd love to be in that state one day!

73

Chris G4DGU







Chris Bartram G4DGU
 

Hello Robin,

I'm not sure that I can agree with you - particularly for tropo, which, I think, is John's main interest.

On 1296 and up the receiver sensitivity on the horizon in a quiet environment is set by a mixture of thermal noise from the the terrestrial environment in the antenna main lobe - which comes mainly from thermal radiation from the ground and receiver front-end noise. The first, major, sidelobes from a properly functioning antenna - be it Yagi-Uda or a dish - will be close to -17dB wrt the main lobe, the other sidelobes will be smaller. As radiation from the ground will be of the order of 300K and the main lobe of the antenna will see around half of that ie. the main lobe will add ~150K to the noise seen by the antenna. With a properly designed antenna, the sidelobes will probably only contribute only a relatively small extra extra amount of noise, maybe 30 - 40K.

Front-end improvement contributions to tropo receiver design haven't actually been massive in the last 30years. We were already at a point in the later part of the 1980s where some of us were making 23cm receivers with sub-50K noise temperatures. Ground  and sky noise were both clearly detectable! In fact, I was doing that ten years earlier when setting-up preamps. GaAsFETs, and later GaAsHEMTs were readily available. A big turning point came when a group of radioastronomers published a paper on 'source feedback' LNAs. It's now pretty routine to make 0.3dBNF preamplifiers for 1.3GHz using that technique. So there's little more to be gained.

If we could completely strip the sidelobe contributions from the equation we'd see a small improvement in sensitivity, but that would inevitably lead to a loss in antenna gain and the sensitivity of the system would be compromised!

As usual, there ain't no free lunch.

The real problem we have with receivers is not sensitivity. It's linearity, but that's a very different discussion!

73

Chris G4DGU



On 16/07/2020 15:31, Robin Szemeti - G1YFG wrote:
My take on the matter is that it has changed in the last 20 years, due to improvements in receiver front end design.  

20+ years ago, the focus (if you will pardon the pun) was on gain,  as many receiver LNA devices were relatively noisy and you could not hear that much ground/sky noise easily.  QRO was relatively difficult to generate and therefore a lot of the antennas were optimised for gain, to deliver as much signal as possible at the target, and the additional sidelobes did not generate a significant degradation to the overall receiver noise performance.

These days, power is cheap and easy, and receiver front end noise figures have improved, and the additional noise contributed by the sidelobes is audible and significant.  It is possible to sacrifice a little gain and optimise for receive noise floor, and minimum sidelobes is now the design goal.

The physics of course has not changed, but maybe the design goalposts have moved a little?

On Thu, 16 Jul 2020 at 10:41, Chris Bartram G4DGU <chris@...> wrote:
Hello John,

Analytically there isn't any real difference between Yagi-Uda antennas,
dishes (and optical telescopes ... look-up 'Airy Disc') when it comes to
optimising performance. They are simply different ways of focussing the
E-M energy towards a distant point. You get optimum 'gain' when the
first sidelobes are close to -17.5dB. Reducing the sidelobe levels is
possible, but at the expense of gain and although some of the amateur
antenna designers and manufacturers have been playing numbers games with
G/T ratio, but I can't see how they escape from the constraints of the
physics! The advantage of dishes as you go up in frequency is that they
are less critical in terms of dimensions, and there is less to go wrong!
They are also a LOT easier to design. Compared to dishes, y-u antennas
are a MTBF nightmare ...

In the amateur literature, the late K2RIW wrote a very nice paper on
this aeons ago ('MAXIMUM DISH EFFICIENCY AND THE BEST ANTENNA SIDELOBE
LEVELS'' Dick Knadle, K2RIW 6/04/05') I think it can be found on the net
for download. I have a copy, but I'm not sure of the copyright issues so
I won't distribute it. Also the text mentioned by K2RIW - Hect's
'Optics' - is also worth a look. But be prepared for some heavy maths!
It gave me some serious headaches!

Matching the feed design to the dish so as to make the edge illumination
of around -10dB will give the best G/T - which it is why it's the usual
choice of antenna designers. Increasing the feed illumination taper to
-11 or -12dB can provide some noise advantage if you have a bit too much
dish gain; I'd love to be in that state one day!

73

Chris G4DGU







Robin Szemeti - G1YFG
 

Good points, noted.  I do think though there has been a shift from "high gain" to "low noise" in what are seen as "good" antenna designs ... and you are right, that it was possible to get low <0.5dB noise figures 30 years ago, but it was relatively uncommon in the wider amateur community.  Perhaps part of the problem was that "antenna gain" was a headline figure easily sold to the un-discerning operator, and having a nice big "21.2dB gain" sticker on your sales brochure was what sold antennas ... perhaps now, the remaining operators are a little more educated on what  good design might be?

On Thu, 16 Jul 2020 at 17:10, Chris Bartram G4DGU <chris@...> wrote:

Hello Robin,

I'm not sure that I can agree with you - particularly for tropo, which, I think, is John's main interest.

On 1296 and up the receiver sensitivity on the horizon in a quiet environment is set by a mixture of thermal noise from the the terrestrial environment in the antenna main lobe - which comes mainly from thermal radiation from the ground and receiver front-end noise. The first, major, sidelobes from a properly functioning antenna - be it Yagi-Uda or a dish - will be close to -17dB wrt the main lobe, the other sidelobes will be smaller. As radiation from the ground will be of the order of 300K and the main lobe of the antenna will see around half of that ie. the main lobe will add ~150K to the noise seen by the antenna. With a properly designed antenna, the sidelobes will probably only contribute only a relatively small extra extra amount of noise, maybe 30 - 40K.

Front-end improvement contributions to tropo receiver design haven't actually been massive in the last 30years. We were already at a point in the later part of the 1980s where some of us were making 23cm receivers with sub-50K noise temperatures. Ground  and sky noise were both clearly detectable! In fact, I was doing that ten years earlier when setting-up preamps. GaAsFETs, and later GaAsHEMTs were readily available. A big turning point came when a group of radioastronomers published a paper on 'source feedback' LNAs. It's now pretty routine to make 0.3dBNF preamplifiers for 1.3GHz using that technique. So there's little more to be gained.

If we could completely strip the sidelobe contributions from the equation we'd see a small improvement in sensitivity, but that would inevitably lead to a loss in antenna gain and the sensitivity of the system would be compromised!

As usual, there ain't no free lunch.

The real problem we have with receivers is not sensitivity. It's linearity, but that's a very different discussion!

73

Chris G4DGU



On 16/07/2020 15:31, Robin Szemeti - G1YFG wrote:
My take on the matter is that it has changed in the last 20 years, due to improvements in receiver front end design.  

20+ years ago, the focus (if you will pardon the pun) was on gain,  as many receiver LNA devices were relatively noisy and you could not hear that much ground/sky noise easily.  QRO was relatively difficult to generate and therefore a lot of the antennas were optimised for gain, to deliver as much signal as possible at the target, and the additional sidelobes did not generate a significant degradation to the overall receiver noise performance.

These days, power is cheap and easy, and receiver front end noise figures have improved, and the additional noise contributed by the sidelobes is audible and significant.  It is possible to sacrifice a little gain and optimise for receive noise floor, and minimum sidelobes is now the design goal.

The physics of course has not changed, but maybe the design goalposts have moved a little?

On Thu, 16 Jul 2020 at 10:41, Chris Bartram G4DGU <chris@...> wrote:
Hello John,

Analytically there isn't any real difference between Yagi-Uda antennas,
dishes (and optical telescopes ... look-up 'Airy Disc') when it comes to
optimising performance. They are simply different ways of focussing the
E-M energy towards a distant point. You get optimum 'gain' when the
first sidelobes are close to -17.5dB. Reducing the sidelobe levels is
possible, but at the expense of gain and although some of the amateur
antenna designers and manufacturers have been playing numbers games with
G/T ratio, but I can't see how they escape from the constraints of the
physics! The advantage of dishes as you go up in frequency is that they
are less critical in terms of dimensions, and there is less to go wrong!
They are also a LOT easier to design. Compared to dishes, y-u antennas
are a MTBF nightmare ...

In the amateur literature, the late K2RIW wrote a very nice paper on
this aeons ago ('MAXIMUM DISH EFFICIENCY AND THE BEST ANTENNA SIDELOBE
LEVELS'' Dick Knadle, K2RIW 6/04/05') I think it can be found on the net
for download. I have a copy, but I'm not sure of the copyright issues so
I won't distribute it. Also the text mentioned by K2RIW - Hect's
'Optics' - is also worth a look. But be prepared for some heavy maths!
It gave me some serious headaches!

Matching the feed design to the dish so as to make the edge illumination
of around -10dB will give the best G/T - which it is why it's the usual
choice of antenna designers. Increasing the feed illumination taper to
-11 or -12dB can provide some noise advantage if you have a bit too much
dish gain; I'd love to be in that state one day!

73

Chris G4DGU







KENT BRITAIN
 

Hi Robin

I'm afraid the 'Low Noise' is a marketing term, not a function of the antenna design.

Kent WA5VJB







On Thursday, July 16, 2020, 11:28:10 AM CDT, Robin Szemeti - G1YFG <robin@...> wrote:


Good points, noted.  I do think though there has been a shift from "high gain" to "low noise" in what are seen as "good" antenna designs ... and you are right, that it was possible to get low <0.5dB noise figures 30 years ago, but it was relatively uncommon in the wider amateur community.  Perhaps part of the problem was that "antenna gain" was a headline figure easily sold to the un-discerning operator, and having a nice big "21.2dB gain" sticker on your sales brochure was what sold antennas ... perhaps now, the remaining operators are a little more educated on what  good design might be?

On Thu, 16 Jul 2020 at 17:10, Chris Bartram G4DGU <chris@...> wrote:

Hello Robin,

I'm not sure that I can agree with you - particularly for tropo, which, I think, is John's main interest.

On 1296 and up the receiver sensitivity on the horizon in a quiet environment is set by a mixture of thermal noise from the the terrestrial environment in the antenna main lobe - which comes mainly from thermal radiation from the ground and receiver front-end noise. The first, major, sidelobes from a properly functioning antenna - be it Yagi-Uda or a dish - will be close to -17dB wrt the main lobe, the other sidelobes will be smaller. As radiation from the ground will be of the order of 300K and the main lobe of the antenna will see around half of that ie. the main lobe will add ~150K to the noise seen by the antenna. With a properly designed antenna, the sidelobes will probably only contribute only a relatively small extra extra amount of noise, maybe 30 - 40K.

Front-end improvement contributions to tropo receiver design haven't actually been massive in the last 30years. We were already at a point in the later part of the 1980s where some of us were making 23cm receivers with sub-50K noise temperatures. Ground  and sky noise were both clearly detectable! In fact, I was doing that ten years earlier when setting-up preamps. GaAsFETs, and later GaAsHEMTs were readily available. A big turning point came when a group of radioastronomers published a paper on 'source feedback' LNAs. It's now pretty routine to make 0.3dBNF preamplifiers for 1.3GHz using that technique. So there's little more to be gained.

If we could completely strip the sidelobe contributions from the equation we'd see a small improvement in sensitivity, but that would inevitably lead to a loss in antenna gain and the sensitivity of the system would be compromised!

As usual, there ain't no free lunch.

The real problem we have with receivers is not sensitivity. It's linearity, but that's a very different discussion!

73

Chris G4DGU



On 16/07/2020 15:31, Robin Szemeti - G1YFG wrote:
My take on the matter is that it has changed in the last 20 years, due to improvements in receiver front end design.  

20+ years ago, the focus (if you will pardon the pun) was on gain,  as many receiver LNA devices were relatively noisy and you could not hear that much ground/sky noise easily.  QRO was relatively difficult to generate and therefore a lot of the antennas were optimised for gain, to deliver as much signal as possible at the target, and the additional sidelobes did not generate a significant degradation to the overall receiver noise performance.

These days, power is cheap and easy, and receiver front end noise figures have improved, and the additional noise contributed by the sidelobes is audible and significant.  It is possible to sacrifice a little gain and optimise for receive noise floor, and minimum sidelobes is now the design goal.

The physics of course has not changed, but maybe the design goalposts have moved a little?

On Thu, 16 Jul 2020 at 10:41, Chris Bartram G4DGU <chris@...> wrote:
Hello John,

Analytically there isn't any real difference between Yagi-Uda antennas,
dishes (and optical telescopes ... look-up 'Airy Disc') when it comes to
optimising performance. They are simply different ways of focussing the
E-M energy towards a distant point. You get optimum 'gain' when the
first sidelobes are close to -17.5dB. Reducing the sidelobe levels is
possible, but at the expense of gain and although some of the amateur
antenna designers and manufacturers have been playing numbers games with
G/T ratio, but I can't see how they escape from the constraints of the
physics! The advantage of dishes as you go up in frequency is that they
are less critical in terms of dimensions, and there is less to go wrong!
They are also a LOT easier to design. Compared to dishes, y-u antennas
are a MTBF nightmare ...

In the amateur literature, the late K2RIW wrote a very nice paper on
this aeons ago ('MAXIMUM DISH EFFICIENCY AND THE BEST ANTENNA SIDELOBE
LEVELS'' Dick Knadle, K2RIW 6/04/05') I think it can be found on the net
for download. I have a copy, but I'm not sure of the copyright issues so
I won't distribute it. Also the text mentioned by K2RIW - Hect's
'Optics' - is also worth a look. But be prepared for some heavy maths!
It gave me some serious headaches!

Matching the feed design to the dish so as to make the edge illumination
of around -10dB will give the best G/T - which it is why it's the usual
choice of antenna designers. Increasing the feed illumination taper to
-11 or -12dB can provide some noise advantage if you have a bit too much
dish gain; I'd love to be in that state one day!

73

Chris G4DGU







Robin Szemeti - G1YFG
 

That's true, perhaps "minimal sidelobe" is a better term.  Perhaps it is more to do with the noisy RF environment in which we now operate, where there are strong noise sources in some directions?  Possibly it is the ready availability of computer based antenna optimisation software such as mmana-gal that has driven the change in apparent emphasis?

On Thu, 16 Jul 2020 at 17:36, KENT BRITAIN <WA5VJB@...> wrote:
Hi Robin

I'm afraid the 'Low Noise' is a marketing term, not a function of the antenna design.

Kent WA5VJB







On Thursday, July 16, 2020, 11:28:10 AM CDT, Robin Szemeti - G1YFG <robin@...> wrote:


Good points, noted.  I do think though there has been a shift from "high gain" to "low noise" in what are seen as "good" antenna designs ... and you are right, that it was possible to get low <0.5dB noise figures 30 years ago, but it was relatively uncommon in the wider amateur community.  Perhaps part of the problem was that "antenna gain" was a headline figure easily sold to the un-discerning operator, and having a nice big "21.2dB gain" sticker on your sales brochure was what sold antennas ... perhaps now, the remaining operators are a little more educated on what  good design might be?

On Thu, 16 Jul 2020 at 17:10, Chris Bartram G4DGU <chris@...> wrote:

Hello Robin,

I'm not sure that I can agree with you - particularly for tropo, which, I think, is John's main interest.

On 1296 and up the receiver sensitivity on the horizon in a quiet environment is set by a mixture of thermal noise from the the terrestrial environment in the antenna main lobe - which comes mainly from thermal radiation from the ground and receiver front-end noise. The first, major, sidelobes from a properly functioning antenna - be it Yagi-Uda or a dish - will be close to -17dB wrt the main lobe, the other sidelobes will be smaller. As radiation from the ground will be of the order of 300K and the main lobe of the antenna will see around half of that ie. the main lobe will add ~150K to the noise seen by the antenna. With a properly designed antenna, the sidelobes will probably only contribute only a relatively small extra extra amount of noise, maybe 30 - 40K.

Front-end improvement contributions to tropo receiver design haven't actually been massive in the last 30years. We were already at a point in the later part of the 1980s where some of us were making 23cm receivers with sub-50K noise temperatures. Ground  and sky noise were both clearly detectable! In fact, I was doing that ten years earlier when setting-up preamps. GaAsFETs, and later GaAsHEMTs were readily available. A big turning point came when a group of radioastronomers published a paper on 'source feedback' LNAs. It's now pretty routine to make 0.3dBNF preamplifiers for 1.3GHz using that technique. So there's little more to be gained.

If we could completely strip the sidelobe contributions from the equation we'd see a small improvement in sensitivity, but that would inevitably lead to a loss in antenna gain and the sensitivity of the system would be compromised!

As usual, there ain't no free lunch.

The real problem we have with receivers is not sensitivity. It's linearity, but that's a very different discussion!

73

Chris G4DGU



On 16/07/2020 15:31, Robin Szemeti - G1YFG wrote:
My take on the matter is that it has changed in the last 20 years, due to improvements in receiver front end design.  

20+ years ago, the focus (if you will pardon the pun) was on gain,  as many receiver LNA devices were relatively noisy and you could not hear that much ground/sky noise easily.  QRO was relatively difficult to generate and therefore a lot of the antennas were optimised for gain, to deliver as much signal as possible at the target, and the additional sidelobes did not generate a significant degradation to the overall receiver noise performance.

These days, power is cheap and easy, and receiver front end noise figures have improved, and the additional noise contributed by the sidelobes is audible and significant.  It is possible to sacrifice a little gain and optimise for receive noise floor, and minimum sidelobes is now the design goal.

The physics of course has not changed, but maybe the design goalposts have moved a little?

On Thu, 16 Jul 2020 at 10:41, Chris Bartram G4DGU <chris@...> wrote:
Hello John,

Analytically there isn't any real difference between Yagi-Uda antennas,
dishes (and optical telescopes ... look-up 'Airy Disc') when it comes to
optimising performance. They are simply different ways of focussing the
E-M energy towards a distant point. You get optimum 'gain' when the
first sidelobes are close to -17.5dB. Reducing the sidelobe levels is
possible, but at the expense of gain and although some of the amateur
antenna designers and manufacturers have been playing numbers games with
G/T ratio, but I can't see how they escape from the constraints of the
physics! The advantage of dishes as you go up in frequency is that they
are less critical in terms of dimensions, and there is less to go wrong!
They are also a LOT easier to design. Compared to dishes, y-u antennas
are a MTBF nightmare ...

In the amateur literature, the late K2RIW wrote a very nice paper on
this aeons ago ('MAXIMUM DISH EFFICIENCY AND THE BEST ANTENNA SIDELOBE
LEVELS'' Dick Knadle, K2RIW 6/04/05') I think it can be found on the net
for download. I have a copy, but I'm not sure of the copyright issues so
I won't distribute it. Also the text mentioned by K2RIW - Hect's
'Optics' - is also worth a look. But be prepared for some heavy maths!
It gave me some serious headaches!

Matching the feed design to the dish so as to make the edge illumination
of around -10dB will give the best G/T - which it is why it's the usual
choice of antenna designers. Increasing the feed illumination taper to
-11 or -12dB can provide some noise advantage if you have a bit too much
dish gain; I'd love to be in that state one day!

73

Chris G4DGU







Neil Smith G4DBN
 

My terrestrial 3cm dish has some sidelobes around 18dB down. I get around 5dB excess noise when pointing the dish at my tree belt, but there is no trace of the tree in the sidelobe even when I point at the sky a few degrees above the horizon.

What I have seen is a tiny extra noise when the dish is elevated 40 degrees or so, as there is a bit of overillumination and I get a faint whiff of the wall and roof of the big house next door. As this dish is primarily for terrestrial and is usually 10 to 13m above ground, the over spill is usually just cold sky, although once when beaming to GM, I found a weird moving noise source that turned out to be the Sun behind the dish at just the right elevation.

There is a very hot noise source at the other end of the village and I think I am seeing it via a sidelobe when beaming at the horizon. I need to set the dish so it can point down a bit more to test that hypothesis.

With a dish up well above local clutter, especially an offset feed, I don't think under-illumination makes sense, but if there are hot things behind a prime focus terrestrial dish, or noisy things in the sidelobe pattern, it might help to under-illuminate a bit. 

Elevation control of a dish is very useful when examining sidelobe performance.

Neil G4DBN

PS 
Apologies to everyone waiting for machining jobs, Caroline is very ill at present, and I am on caring duties, plus the carriers lost a pair of 10GHz Pickett-Potter WA6KBL design horns, and I've had to spend 8 hours of scarce machine shop time making replacement parts. On top of that, two suppliers sent material that has been lost by the couriers. Quality.



 

On Thu, 16 Jul 2020 at 10:41, Chris Bartram G4DGU <chris@...> wrote:
Hello John,

Analytically there isn't any real difference between Yagi-Uda antennas,
dishes (and optical telescopes ... look-up 'Airy Disc') when it comes to
optimising performance. They are simply different ways of focussing the
E-M energy towards a distant point. You get optimum 'gain' when the
first sidelobes are close to -17.5dB. Reducing the sidelobe levels is
possible, but at the expense of gain and although some of the amateur
antenna designers and manufacturers have been playing numbers games with
G/T ratio, but I can't see how they escape from the constraints of the
physics!

Chris,
what physics comes into play here? It's either something I never knew, or have forgotten. 

73

Chris G4DGU

Dave, G8WRB


John Lemay
 

Gentlemen

 

Many thanks for your thoughtful input. It’s given me food for thought and I’ve taken the opportunity to read the K2RIW paper, which I’d not seen before.

 

The background to my enquiry arises from the annoyance I have on 23cms from a local noise source. It’s around “S”7 and occurs as a warble every 100kHz or so across the band with a variety of noise and mush in-between. I suspect the local pub, because there is not much else in that direction.

 

Thinking through the content of your replies I decided to carefully check the beam pattern of this noise source and found that I could easily hear it on the first and second side lobes. But the penny has dropped, I think, in that the noise could be any polarisation and it could be re-radiated from cables within the building.

 

I’ve also carefully checked the only beacon here which is strong enough for tests; Martlesham. That isn’t as simple as it should be because I think there are some reflections on the path. But to the best of my ability, I concluded that the first side lobes are 16dB down and they are at +/- 20 degrees from the main lobe. My feed is a dual band loop feed and on 13cms I decided that the side lobes are also at -16dB and at around +/- 15 degrees from the main lobe.

 

The dish is a 2m diameter prime focus home made construction by the way. The feed will be modified to include 9cms shortly, and that’s another reason for my enquiry, because if the dish is not performing optimally on 23cms it certainly won’t play ball on 9cms !

 

Thanks again

 

John G4ZTR

 

 

From: UKMicrowaves@groups.io [mailto:UKMicrowaves@groups.io] On Behalf Of Robin Szemeti - G1YFG
Sent: 16 July 2020 18:29
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] G/T for parabolic dishes

 

That's true, perhaps "minimal sidelobe" is a better term.  Perhaps it is more to do with the noisy RF environment in which we now operate, where there are strong noise sources in some directions?  Possibly it is the ready availability of computer based antenna optimisation software such as mmana-gal that has driven the change in apparent emphasis?

 

On Thu, 16 Jul 2020 at 17:36, KENT BRITAIN <WA5VJB@...> wrote:

Hi Robin

 

I'm afraid the 'Low Noise' is a marketing term, not a function of the antenna design.

 

Kent WA5VJB

 

 

 

 

 

 

 

On Thursday, July 16, 2020, 11:28:10 AM CDT, Robin Szemeti - G1YFG <robin@...> wrote:

 

 

Good points, noted.  I do think though there has been a shift from "high gain" to "low noise" in what are seen as "good" antenna designs ... and you are right, that it was possible to get low <0.5dB noise figures 30 years ago, but it was relatively uncommon in the wider amateur community.  Perhaps part of the problem was that "antenna gain" was a headline figure easily sold to the un-discerning operator, and having a nice big "21.2dB gain" sticker on your sales brochure was what sold antennas ... perhaps now, the remaining operators are a little more educated on what  good design might be?

 

On Thu, 16 Jul 2020 at 17:10, Chris Bartram G4DGU <chris@...> wrote:

Hello Robin,

I'm not sure that I can agree with you - particularly for tropo, which, I think, is John's main interest.

On 1296 and up the receiver sensitivity on the horizon in a quiet environment is set by a mixture of thermal noise from the the terrestrial environment in the antenna main lobe - which comes mainly from thermal radiation from the ground and receiver front-end noise. The first, major, sidelobes from a properly functioning antenna - be it Yagi-Uda or a dish - will be close to -17dB wrt the main lobe, the other sidelobes will be smaller. As radiation from the ground will be of the order of 300K and the main lobe of the antenna will see around half of that ie. the main lobe will add ~150K to the noise seen by the antenna. With a properly designed antenna, the sidelobes will probably only contribute only a relatively small extra extra amount of noise, maybe 30 - 40K.

Front-end improvement contributions to tropo receiver design haven't actually been massive in the last 30years. We were already at a point in the later part of the 1980s where some of us were making 23cm receivers with sub-50K noise temperatures. Ground  and sky noise were both clearly detectable! In fact, I was doing that ten years earlier when setting-up preamps. GaAsFETs, and later GaAsHEMTs were readily available. A big turning point came when a group of radioastronomers published a paper on 'source feedback' LNAs. It's now pretty routine to make 0.3dBNF preamplifiers for 1.3GHz using that technique. So there's little more to be gained.

If we could completely strip the sidelobe contributions from the equation we'd see a small improvement in sensitivity, but that would inevitably lead to a loss in antenna gain and the sensitivity of the system would be compromised!

As usual, there ain't no free lunch.

The real problem we have with receivers is not sensitivity. It's linearity, but that's a very different discussion!

73

Chris G4DGU

 

 

On 16/07/2020 15:31, Robin Szemeti - G1YFG wrote:

My take on the matter is that it has changed in the last 20 years, due to improvements in receiver front end design.  

 

20+ years ago, the focus (if you will pardon the pun) was on gain,  as many receiver LNA devices were relatively noisy and you could not hear that much ground/sky noise easily.  QRO was relatively difficult to generate and therefore a lot of the antennas were optimised for gain, to deliver as much signal as possible at the target, and the additional sidelobes did not generate a significant degradation to the overall receiver noise performance.

 

These days, power is cheap and easy, and receiver front end noise figures have improved, and the additional noise contributed by the sidelobes is audible and significant.  It is possible to sacrifice a little gain and optimise for receive noise floor, and minimum sidelobes is now the design goal.

 

The physics of course has not changed, but maybe the design goalposts have moved a little?

 

On Thu, 16 Jul 2020 at 10:41, Chris Bartram G4DGU <chris@...> wrote:

Hello John,

Analytically there isn't any real difference between Yagi-Uda antennas,
dishes (and optical telescopes ... look-up 'Airy Disc') when it comes to
optimising performance. They are simply different ways of focussing the
E-M energy towards a distant point. You get optimum 'gain' when the
first sidelobes are close to -17.5dB. Reducing the sidelobe levels is
possible, but at the expense of gain and although some of the amateur
antenna designers and manufacturers have been playing numbers games with
G/T ratio, but I can't see how they escape from the constraints of the
physics! The advantage of dishes as you go up in frequency is that they
are less critical in terms of dimensions, and there is less to go wrong!
They are also a LOT easier to design. Compared to dishes, y-u antennas
are a MTBF nightmare ...

In the amateur literature, the late K2RIW wrote a very nice paper on
this aeons ago ('MAXIMUM DISH EFFICIENCY AND THE BEST ANTENNA SIDELOBE
LEVELS'' Dick Knadle, K2RIW 6/04/05') I think it can be found on the net
for download. I have a copy, but I'm not sure of the copyright issues so
I won't distribute it. Also the text mentioned by K2RIW - Hect's
'Optics' - is also worth a look. But be prepared for some heavy maths!
It gave me some serious headaches!

Matching the feed design to the dish so as to make the edge illumination
of around -10dB will give the best G/T - which it is why it's the usual
choice of antenna designers. Increasing the feed illumination taper to
-11 or -12dB can provide some noise advantage if you have a bit too much
dish gain; I'd love to be in that state one day!

73

Chris G4DGU






Virus-free. www.avg.com


Chris Bartram G4DGU
 

Hello Dave,

what physics comes into play here? It's either something I never knew, or have forgotten.
It's all to do with diffraction. Optical telescope designers know all about it, but I'd have to go back to a text to refresh my limited understanding in order to say more.

73

Chris G4DGU


Andy G4JNT
 


On Thu, 16 Jul 2020 at 21:57, Chris Bartram G4DGU <chris@...> wrote:
Hello Dave,

> what physics comes into play here? It's either something I never knew,
> or have forgotten.

It's all to do with diffraction. Optical telescope designers know all
about it, but I'd have to go back to a text to refresh my limited
understanding in order to say more.

73

Chris G4DGU





alwyn.seeds1
 

Dear John,

For contesting the low sidelobes of underilluminated dishes can be a disadvantage. With Yagi arrays a good few CQ responses come off side-lobes, and as Chris says if the main beam sees the ground, ultra-low side lobes are not required.

Your 23cm noise source interests me- I have had some problems from radiated harmonics off poorly installed 1000Base-T wiring to outdoor 5GHz links, but this has quite broad noise characteristics; had to explain to the installers that the cable screening does need to be connected to shielded plugs at each end!

Let us know the exact source, when you have tracked it down, please.

Regards,

Alwyn G8DOH


_____________________________________________________

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


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


Robin Szemeti - G1YFG
 

I think the advantages/disadvantages of sidelobes depend to some extent on the contest .... on a relatively sparsely populated 23cm contest, being heard off a sidelobe would be a distinct advantage.  

In a good location on a busy 2m contest, you almost certainly will not have the frequency to yourself, and being able to hear just the stuff you are pointing at, with minimal sidelobes may work better.


On Fri, 17 Jul 2020 at 10:39, alwyn.seeds1 <a.seeds@...> wrote:
Dear John,

For contesting the low sidelobes of underilluminated dishes can be a disadvantage. With Yagi arrays a good few CQ responses come off side-lobes, and as Chris says if the main beam sees the ground, ultra-low side lobes are not required.

Your 23cm noise source interests me- I have had some problems from radiated harmonics off poorly installed 1000Base-T wiring to outdoor 5GHz links, but this has quite broad noise characteristics; had to explain to the installers that the cable screening does need to be connected to shielded plugs at each end!

Let us know the exact source, when you have tracked it down, please.

Regards,

Alwyn G8DOH


_____________________________________________________

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


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


Brian Howie GM4DIJ
 

It's not quite as bad as that. The optical side lobes assumes the limiting aperture (lens etc) is uniformly illuminated. In RF land the dish is illuminated by a spatially non-uniform field. I imagine the illumination is some sort of sine function from a horn. The level at the edges of is meant to be ( rule of thumb) about -10db from the centre. So finger in the air, the sidelobes are going to to be 10dB lower than the Airy Disc solution. Of course you get sidelobes from both the horn and dish edge.

The maths must be horrible.

I used to do this sort of stuff for laser beam diffraction, The difference between a top-Hat beam and a truncated Gaussian beam is about 12dB for an example I have. Infinite Gaussian has the lowest sidelobe, but you need a big lens for that ;-)

Maths for this is a lot easier, an decent ray-tracing programs can do it.

Brian GM4DIJ

On 16/07/2020 22:09, Andy G4JNT wrote:
Good pictures of optical sidelobes
https://www.photokonnexion.com/definition-diffraction-pinhole-diffraction/
http://hyperphysics.phy-astr.gsu.edu/hbase/phyopt/cirapp2.html
image.png
Andy
www.g4jnt.com <http://www.g4jnt.com>
On Thu, 16 Jul 2020 at 21:57, Chris Bartram G4DGU <chris@chris-bartram.co.uk <mailto:chris@chris-bartram.co.uk>> wrote:
Hello Dave,

> what physics comes into play here? It's either something I never
knew,
> or have forgotten.
It's all to do with diffraction. Optical telescope designers know all
about it, but I'd have to go back to a text to refresh my limited
understanding in order to say more.
73
Chris G4DGU
--
Brian


 

On Thu, 16 Jul 2020 at 21:57, Chris Bartram G4DGU <chris@...> wrote:
Hello Dave,

> what physics comes into play here? It's either something I never knew,
> or have forgotten.

It's all to do with diffraction. Optical telescope designers know all
about it, but I'd have to go back to a text to refresh my limited
understanding in order to say more.

73

Chris G4DGU

I’m not convinced that one can arrive at the same levels for a Yagi-Uda though, which I believe was mentioned earlier.  You can compute a capture area, which is itself not circular. 

I have thought about trying to reduce the first sidelobe on a Yagi-Uda type antenna, perhaps with a small bend on the ends of the elements. I personally find that sidelobe annoying as if one knows the general direction of a station one can rotate the beam and think one has the peak but doesn’t.

You can argue sidelobes are advantageous in a contest. 

Anyway, I will not take the argument too far from the dish. Yagis are a quite different beast. 

Dave
--
Dr. David Kirkby,
Kirkby Microwave Ltd,
drkirkby@...
https://www.kirkbymicrowave.co.uk/
Telephone 01621-680100./ +44 1621 680100

Registered in England & Wales, company number 08914892.
Registered office:
Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United Kingdom


KENT BRITAIN
 

There has been some interesting work with very long, very narrow banded, log periodics
with excellent gain and low sidelobes.    But all the work I have seen was simulations,
looking forward to the day someone brings one to the antenna range.   Kent WA5VJB-2E0VAA

On Friday, July 17, 2020, 2:16:45 PM CDT, Dr. David Kirkby, Kirkby Microwave Ltd <drkirkby@...> wrote:


On Thu, 16 Jul 2020 at 21:57, Chris Bartram G4DGU <chris@...> wrote:
Hello Dave,

> what physics comes into play here? It's either something I never knew,
> or have forgotten.

It's all to do with diffraction. Optical telescope designers know all
about it, but I'd have to go back to a text to refresh my limited
understanding in order to say more.

73

Chris G4DGU

I’m not convinced that one can arrive at the same levels for a Yagi-Uda though, which I believe was mentioned earlier.  You can compute a capture area, which is itself not circular. 

I have thought about trying to reduce the first sidelobe on a Yagi-Uda type antenna, perhaps with a small bend on the ends of the elements. I personally find that sidelobe annoying as if one knows the general direction of a station one can rotate the beam and think one has the peak but doesn’t.

You can argue sidelobes are advantageous in a contest. 

Anyway, I will not take the argument too far from the dish. Yagis are a quite different beast. 

Dave
--
Dr. David Kirkby,
Kirkby Microwave Ltd,
drkirkby@...
https://www.kirkbymicrowave.co.uk/
Telephone 01621-680100./ +44 1621 680100

Registered in England & Wales, company number 08914892.
Registered office:
Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United Kingdom


 

On Fri, 17 Jul 2020 at 20:19, KENT BRITAIN <WA5VJB@...> wrote:

There has been some interesting work with very long, very narrow banded, log periodics
with excellent gain and low sidelobes.    But all the work I have seen was simulations,
looking forward to the day someone brings one to the antenna range.   Kent WA5VJB-2E0VAA

If it’s very long, the chances are it’s impossible to hold the elements in position. Our radio club had a Yagi-Uda type antenna with elements cut to +/- 0.5 mm, but which move around several hundred mm. It would probably have been okay for MEM (moon earth moon), as the much lower gravity and little atmosphere on the moon  would mean it could probably be supported on the moon. 😂😂😢 Buf on earth it was not useful. 

Dave
--
Dr. David Kirkby,
Kirkby Microwave Ltd,
drkirkby@...
https://www.kirkbymicrowave.co.uk/
Telephone 01621-680100./ +44 1621 680100

Registered in England & Wales, company number 08914892.
Registered office:
Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United Kingdom


Robin Szemeti - G1YFG
 

That reminds me .. is there an "accessible" antenna range in the UK where idiots such as myself can bring their construction and come away with a graph to laugh at?  I've thought about putting something together in the filed at the back of the house, but as always, it never quite happens.

On Fri, 17 Jul 2020 at 20:19, KENT BRITAIN <WA5VJB@...> wrote:
There has been some interesting work with very long, very narrow banded, log periodics
with excellent gain and low sidelobes.    But all the work I have seen was simulations,
looking forward to the day someone brings one to the antenna range.   Kent WA5VJB-2E0VAA

On Friday, July 17, 2020, 2:16:45 PM CDT, Dr. David Kirkby, Kirkby Microwave Ltd <drkirkby@...> wrote:


On Thu, 16 Jul 2020 at 21:57, Chris Bartram G4DGU <chris@...> wrote:
Hello Dave,

> what physics comes into play here? It's either something I never knew,
> or have forgotten.

It's all to do with diffraction. Optical telescope designers know all
about it, but I'd have to go back to a text to refresh my limited
understanding in order to say more.

73

Chris G4DGU

I’m not convinced that one can arrive at the same levels for a Yagi-Uda though, which I believe was mentioned earlier.  You can compute a capture area, which is itself not circular. 

I have thought about trying to reduce the first sidelobe on a Yagi-Uda type antenna, perhaps with a small bend on the ends of the elements. I personally find that sidelobe annoying as if one knows the general direction of a station one can rotate the beam and think one has the peak but doesn’t.

You can argue sidelobes are advantageous in a contest. 

Anyway, I will not take the argument too far from the dish. Yagis are a quite different beast. 

Dave
--
Dr. David Kirkby,
Kirkby Microwave Ltd,
drkirkby@...
https://www.kirkbymicrowave.co.uk/
Telephone 01621-680100./ +44 1621 680100

Registered in England & Wales, company number 08914892.
Registered office:
Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United Kingdom


KENT BRITAIN
 

Easier in my back yard.   I've got 100 Meters to work with.
A photo of my butt testing a 3 cm antenna is on the cover of the next CQ magazine.
Yes, lots of comments about showing my best side.

Kent

On Friday, July 17, 2020, 3:41:40 PM CDT, Robin Szemeti - G1YFG <robin@...> wrote:


That reminds me .. is there an "accessible" antenna range in the UK where idiots such as myself can bring their construction and come away with a graph to laugh at?  I've thought about putting something together in the filed at the back of the house, but as always, it never quite happens.

On Fri, 17 Jul 2020 at 20:19, KENT BRITAIN <WA5VJB@...> wrote:
There has been some interesting work with very long, very narrow banded, log periodics
with excellent gain and low sidelobes.    But all the work I have seen was simulations,
looking forward to the day someone brings one to the antenna range.   Kent WA5VJB-2E0VAA

On Friday, July 17, 2020, 2:16:45 PM CDT, Dr. David Kirkby, Kirkby Microwave Ltd <drkirkby@...> wrote:


On Thu, 16 Jul 2020 at 21:57, Chris Bartram G4DGU <chris@...> wrote:
Hello Dave,

> what physics comes into play here? It's either something I never knew,
> or have forgotten.

It's all to do with diffraction. Optical telescope designers know all
about it, but I'd have to go back to a text to refresh my limited
understanding in order to say more.

73

Chris G4DGU

I’m not convinced that one can arrive at the same levels for a Yagi-Uda though, which I believe was mentioned earlier.  You can compute a capture area, which is itself not circular. 

I have thought about trying to reduce the first sidelobe on a Yagi-Uda type antenna, perhaps with a small bend on the ends of the elements. I personally find that sidelobe annoying as if one knows the general direction of a station one can rotate the beam and think one has the peak but doesn’t.

You can argue sidelobes are advantageous in a contest. 

Anyway, I will not take the argument too far from the dish. Yagis are a quite different beast. 

Dave
--
Dr. David Kirkby,
Kirkby Microwave Ltd,
drkirkby@...
https://www.kirkbymicrowave.co.uk/
Telephone 01621-680100./ +44 1621 680100

Registered in England & Wales, company number 08914892.
Registered office:
Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United Kingdom