Oscillator phase noise will show up as noise sidebands on the signals
you're measuring. They'll limit your ability to use your spectrum
analyzer to judge the noise that's really coming from the signals. As a
very good and professional engineer who spent a number of years
designing some of the best commercial spectrum analyzers made, Wes most
certainly paid a great deal of attention to phase noise and dynamic
range, two of the most important potentially limiting factors of
spectrum analyzer performance.

Roy Lewallen, W7EL

Hans Summers wrote:

. . .
I suppose oscillator phase noise is also important. . .

Avery Fineman wrote:
In article , "Ian White, G3SEK"
writes:

A spectrum analyser is simply a frequency-swept receiver with a
dB-scaled output to a screen. To give meaningful results, that receiver
must have a very high dynamic range with very low spurious responses.

"Meaningful results" are subjective to the hobbyist.

I'm an amateur, not a professional, but have paid my dues in this area.
I have helped develop, and have built and used, two generations of HF (0
- 100MHz) spectrum analysers of the NE602 persuasion. Then a
basket-case Texscan 0-1GHz box came along which I rescued and used for
VHF work. That was followed by the HP-141 series analyser which I'm
currently using (and can now be bought for much less than the price of a
new rig). I also developed a 0-1.5GHz tracking generator for the 141
boxes. Through friends, I have some limited access to modern
professional equipment, and know how to use it.

My experience with the NE602 type is that it's great as a "fun"
analyser, but you can very quickly outgrow it. Yes, it will give you
general indications, like "Is that oscillator working?" or "Is that
final output stage generating harmonics?" Those indications even have
some numbers of MHz and dB attached.

NO analyzer means NO results.

Of course I'd rather have that much information than no analyser at all.
(Len, if you happen not to mention something that is completely obvious,
I still assume that you're completely aware of it. Kindly extend the
same courtesy to me.)


But once you have a "fun" analyser giving you basic indications, it
inevitably draws you into asking more searching questions, like "How
much of a harmonic problem do I really have?" That's where fun analysers
come unstuck, because you don't know how much of those indicated
harmonics are real, and how much is being generated in the analyser
itself.

An important practical crunch point for amateurs is "Does this
transmitter meet the FCC requirement for 60dB minimum harmonic
suppression?" Now you discover the big difference between 60-70dB
on-screen dynamic range, and 60-70dB *spurious-free* dynamic range. The
first is easy - the second is damned hard to guarantee in a HB design.

If you get to that point - and many people will - then you'll wish you'd
taken the extra trouble to build an analyser that you could trust a
little more. The NE602 type categorically will not hack that kind of
problem, because the equipment under test is probably much cleaner than
it is! The W7ZOI type may just do it, with care, and the same is true of
the 141 generation. Only the best modern analysers will ace the problem
without needing careful attention from the user.

The simple rule of thumb is:

1. If the performance of your test equipment is better - in whatever
ways are important for the measurement in hand - than the performance of
the equipment under test (EUT) then the results will be reliable enough
for amateur work.

2. If your test equipment is comparable with the EUT, you'll get useful
indications.

3. If your test equipment is worse than the EUT, what you're actually
testing is the testgear itself! You'll learn nothing useful about the
EUT, and may actually be misled.

So aim for zone 1, settle for zone 2 if you must, but stay out of zone
3.

and can you quantify it?

Very easily, in the same ways as you test a receiver for strong-signal
handling.... but most graphically by looking at the same spectra with
two analysers side-by-side. The one showing fewest signals is the one
you can rely on most.

Have you priced the used spectrum analyzers lately? Do you expect
others to have ready access to "another" spectrum analyzer?

That comment was mostly to make the point that the cleanest display is
the most real.

The absolute standard of comparison is the perfect spectrum analyser
with no spurious responses. That is an *idea* to which anyone has ready
access, any time.

Before I ever had a real spectrum analyser, I got a lot of good mileage
out of thinking "Now if I had a spectrum analyser, what would it
probably be showing me here?" That isn't empty dreaming - it's a highly
*practical* way of focusing your thinking on a problem in the
here-and-now.

When I built the first "fun" analyser (which had to wait until the
hardware collection had expanded to include a good oscilloscope) I found
it quite disappointing, because it still wasn't showing me reality. I
was still having to think "Now if I had a *good* spectrum analyser, what
would it probably be showing me here?"

Over the years, more and more of the mental testgear has been changed
into hardware, but the mental spectrum analyser still gets switched on
quite frequently - it has instant warmup and it also works very well in
the bath and, er, the other places where I do my best thinking!

Out in the workshop, the real hardware adds real-life numbers to the
ideas, and also checks that there isn't anything happening that I hadn't
thought of.



As with lots of beginner-level test equipment, it sometimes needs an
expert to understand it!

I disagree. There are any number of application notes free for the
downloading on the Internet, from Agilent they are copies of older
(two decades at least) FREE paper application notes. Agilent also
has free application notes on the basic building blocks within an
analyzer and much information on the characteristics of those blocks.

That's not the point. You are talking about the availability of
information, which is not in dispute. I am talking about understanding
and applying it.

By the time you've absorbed the available information - especially the
information about the *imperfections* of spectrum analysers, so you can
understand how much of what you see on-screen is real - you're not a
beginner any more. It will have been a useful learning experience... but
perhaps not the one you'd have chosen right then.

Being a beginner in any area is hard work. When you're ttrying to learn
what it's all about, the last thing you need is to have to fight poor
equipment as well. Musical instruments are very much the same -
so-called "beginner" instruments are actually quite hard work for anyone
to play; in contrast, a better-quality (read: expensive) instrument
feels almost like it's playing itself.

Unfortunately, budget reasons mean that beginners will almost always
have to learn on instruments (test or music) that are hard to use. We
have to accept that, but for heaven's sake let's not pretend to make a
virtue of it!

In this particular case, there does exist a very good, reliable,
second-level design (the W7ZOI). It's a little more expensive to build
than the entry-level stuff, but not significantly more complex. It's
very well designed, so while you're building it you'll learn some very
useful things from a top amateur (and former professional) RF designer.
It's also modular, so you can build it in stages, and upgrade various
parts as and when you want.


Bottom line: having played with the entry-level NE602 stuff myself, my
advice would be:

* If you only want a "fun" analyser, then go for it and have fun.

* But if you want the analyser as a tool to help you develop good radio
gear, go direct to the W7ZOI design.

For a VHF/UHF HB design - which is obviously going to be a lot more
complex, and really is an advanced project - search Google for: "S57MV
spectrum analyser" (without the quotes, and also look for the American
"analyzer" spelling).


--
73 from Ian G3SEK 'In Practice' columnist for RadCom (RSGB)
Editor, 'The VHF/UHF DX Book'
http://www.ifwtech.co.uk/g3sek

Avery Fineman wrote:
In article , "Ian White, G3SEK"
writes:

A spectrum analyser is simply a frequency-swept receiver with a
dB-scaled output to a screen. To give meaningful results, that receiver
must have a very high dynamic range with very low spurious responses.

"Meaningful results" are subjective to the hobbyist.

I'm an amateur, not a professional, but have paid my dues in this area.
I have helped develop, and have built and used, two generations of HF (0
- 100MHz) spectrum analysers of the NE602 persuasion. Then a
basket-case Texscan 0-1GHz box came along which I rescued and used for
VHF work. That was followed by the HP-141 series analyser which I'm
currently using (and can now be bought for much less than the price of a
new rig). I also developed a 0-1.5GHz tracking generator for the 141
boxes. Through friends, I have some limited access to modern
professional equipment, and know how to use it.

My experience with the NE602 type is that it's great as a "fun"
analyser, but you can very quickly outgrow it. Yes, it will give you
general indications, like "Is that oscillator working?" or "Is that
final output stage generating harmonics?" Those indications even have
some numbers of MHz and dB attached.

NO analyzer means NO results.

Of course I'd rather have that much information than no analyser at all.
(Len, if you happen not to mention something that is completely obvious,
I still assume that you're completely aware of it. Kindly extend the
same courtesy to me.)


But once you have a "fun" analyser giving you basic indications, it
inevitably draws you into asking more searching questions, like "How
much of a harmonic problem do I really have?" That's where fun analysers
come unstuck, because you don't know how much of those indicated
harmonics are real, and how much is being generated in the analyser
itself.

An important practical crunch point for amateurs is "Does this
transmitter meet the FCC requirement for 60dB minimum harmonic
suppression?" Now you discover the big difference between 60-70dB
on-screen dynamic range, and 60-70dB *spurious-free* dynamic range. The
first is easy - the second is damned hard to guarantee in a HB design.

If you get to that point - and many people will - then you'll wish you'd
taken the extra trouble to build an analyser that you could trust a
little more. The NE602 type categorically will not hack that kind of
problem, because the equipment under test is probably much cleaner than
it is! The W7ZOI type may just do it, with care, and the same is true of
the 141 generation. Only the best modern analysers will ace the problem
without needing careful attention from the user.

The simple rule of thumb is:

1. If the performance of your test equipment is better - in whatever
ways are important for the measurement in hand - than the performance of
the equipment under test (EUT) then the results will be reliable enough
for amateur work.

2. If your test equipment is comparable with the EUT, you'll get useful
indications.

3. If your test equipment is worse than the EUT, what you're actually
testing is the testgear itself! You'll learn nothing useful about the
EUT, and may actually be misled.

So aim for zone 1, settle for zone 2 if you must, but stay out of zone
3.

and can you quantify it?

Very easily, in the same ways as you test a receiver for strong-signal
handling.... but most graphically by looking at the same spectra with
two analysers side-by-side. The one showing fewest signals is the one
you can rely on most.

Have you priced the used spectrum analyzers lately? Do you expect
others to have ready access to "another" spectrum analyzer?

That comment was mostly to make the point that the cleanest display is
the most real.

The absolute standard of comparison is the perfect spectrum analyser
with no spurious responses. That is an *idea* to which anyone has ready
access, any time.

Before I ever had a real spectrum analyser, I got a lot of good mileage
out of thinking "Now if I had a spectrum analyser, what would it
probably be showing me here?" That isn't empty dreaming - it's a highly
*practical* way of focusing your thinking on a problem in the
here-and-now.

When I built the first "fun" analyser (which had to wait until the
hardware collection had expanded to include a good oscilloscope) I found
it quite disappointing, because it still wasn't showing me reality. I
was still having to think "Now if I had a *good* spectrum analyser, what
would it probably be showing me here?"

Over the years, more and more of the mental testgear has been changed
into hardware, but the mental spectrum analyser still gets switched on
quite frequently - it has instant warmup and it also works very well in
the bath and, er, the other places where I do my best thinking!

Out in the workshop, the real hardware adds real-life numbers to the
ideas, and also checks that there isn't anything happening that I hadn't
thought of.



As with lots of beginner-level test equipment, it sometimes needs an
expert to understand it!

I disagree. There are any number of application notes free for the
downloading on the Internet, from Agilent they are copies of older
(two decades at least) FREE paper application notes. Agilent also
has free application notes on the basic building blocks within an
analyzer and much information on the characteristics of those blocks.

That's not the point. You are talking about the availability of
information, which is not in dispute. I am talking about understanding
and applying it.

By the time you've absorbed the available information - especially the
information about the *imperfections* of spectrum analysers, so you can
understand how much of what you see on-screen is real - you're not a
beginner any more. It will have been a useful learning experience... but
perhaps not the one you'd have chosen right then.

Being a beginner in any area is hard work. When you're ttrying to learn
what it's all about, the last thing you need is to have to fight poor
equipment as well. Musical instruments are very much the same -
so-called "beginner" instruments are actually quite hard work for anyone
to play; in contrast, a better-quality (read: expensive) instrument
feels almost like it's playing itself.

Unfortunately, budget reasons mean that beginners will almost always
have to learn on instruments (test or music) that are hard to use. We
have to accept that, but for heaven's sake let's not pretend to make a
virtue of it!

In this particular case, there does exist a very good, reliable,
second-level design (the W7ZOI). It's a little more expensive to build
than the entry-level stuff, but not significantly more complex. It's
very well designed, so while you're building it you'll learn some very
useful things from a top amateur (and former professional) RF designer.
It's also modular, so you can build it in stages, and upgrade various
parts as and when you want.


Bottom line: having played with the entry-level NE602 stuff myself, my
advice would be:

* If you only want a "fun" analyser, then go for it and have fun.

* But if you want the analyser as a tool to help you develop good radio
gear, go direct to the W7ZOI design.

For a VHF/UHF HB design - which is obviously going to be a lot more
complex, and really is an advanced project - search Google for: "S57MV
spectrum analyser" (without the quotes, and also look for the American
"analyzer" spelling).


--
73 from Ian G3SEK 'In Practice' columnist for RadCom (RSGB)
Editor, 'The VHF/UHF DX Book'
http://www.ifwtech.co.uk/g3sek

"Ian White, G3SEK" wrote in message
...
Avery Fineman wrote:
In article , "Ian White, G3SEK"
writes:


[...]


But once you have a "fun" analyser giving you basic indications, it
inevitably draws you into asking more searching questions, like "How
much of a harmonic problem do I really have?" That's where fun analysers
come unstuck, because you don't know how much of those indicated
harmonics are real, and how much is being generated in the analyser
itself.

I think it does much better than just give general indications. It certainly
has opened my eyes to a whole new world inaccessible to me before having the
analyser. It's comparable to the change in my hobby life when I aquired my
first oscilloscope. You still enjoy the hobby with no oscilloscope, but when
you get one you feel like you'd been blind before, and can't imagine life
without it.

The other day I looked at the 39.5MHz IF in an old television tuned to a UK
TV station signal. I could see the video and sound subcarriers clearly
identifiable. Tuning the TV I could move the signals out of the Surface
Acoustic Wave (SAW) filter skirts and get a good idea of the shape of the
filter's response curve. Before I might have read about that stuff in
theory, now I can actually see it!

Under many cases it will be possible to identify any spurious signals
generated within the analyser itself. The input attenuator (in my case
0-63dB in 1dB steps) allows quite accurate measurement of the magnitude of
frequency components, and also allows reduction of strong signals. In many
cases this can be used to eliminate many of the spuri.

Or, if the signal being monitored is a VFO output for example, or another
variable frequency source, changing the frequency will often identify the
spurious signals. For example I might increase the VFO frequency, causing
the fundamental frequency peak to move left to right on the screen. If I see
some other peak move in the opposite direction, or if I see a peak moving
across the display very quickly, I know I've seen a spurious response.

Much of the spurious response problems come at the high end of the frequency
coverage range. For lower frequency work one could insert an additional
easily-constructed low pass filter and eliminate much of the problem.

So yes, no doubt the more effort you put in and the more complex the
spectrum analyser, the better the results obtainable. But even with a simple
analyser there's so much you can do.


An important practical crunch point for amateurs is "Does this
transmitter meet the FCC requirement for 60dB minimum harmonic
suppression?" Now you discover the big difference between 60-70dB
on-screen dynamic range, and 60-70dB *spurious-free* dynamic range. The
first is easy - the second is damned hard to guarantee in a HB design.

I could be wrong but I thought the FCC requirement was 40dB minimum harmonic
suppression, or 30dB for transmitters with output powers 5W or less? That's
well within the range of even a simple spectrum analyser. I doubt I'd have
any problem ensuring an HF QRP rig met FCC requirements with my analsyer.


If you get to that point - and many people will - then you'll wish you'd
taken the extra trouble to build an analyser that you could trust a
little more. The NE602 type categorically will not hack that kind of
problem, because the equipment under test is probably much cleaner than
it is! The W7ZOI type may just do it, with care, and the same is true of
the 141 generation. Only the best modern analysers will ace the problem
without needing careful attention from the user.

I'm already working on my Mk2 analyser. Not because I'm unhappy with the
measurements of the Mk1, but because I had so much fun with it, ending up
building it for its own sake, and not ready to stop having fun yet.


The simple rule of thumb is:

1. If the performance of your test equipment is better - in whatever
ways are important for the measurement in hand - than the performance of
the equipment under test (EUT) then the results will be reliable enough
for amateur work.

2. If your test equipment is comparable with the EUT, you'll get useful
indications.

3. If your test equipment is worse than the EUT, what you're actually
testing is the testgear itself! You'll learn nothing useful about the
EUT, and may actually be misled.

So aim for zone 1, settle for zone 2 if you must, but stay out of zone
3.

Not sure I agree completely - if the EUT is *supposed* to be in your zone 3,
but is not functioning properly, then even a simple spectrum analyser could
help you find the fault. It wouldn't help you make precise receiver
performance measurements on it when you corrected the fault, but would help
with diagnosis.

[...]


Bottom line: having played with the entry-level NE602 stuff myself, my
advice would be:

* If you only want a "fun" analyser, then go for it and have fun.

* But if you want the analyser as a tool to help you develop good radio
gear, go direct to the W7ZOI design.

Don't forget the W7ZOI analyser covers 0-70MHz only. You can double that
with a NE602 design. Nevertheless I liked the W7ZOI analyser a lot.


For a VHF/UHF HB design - which is obviously going to be a lot more
complex, and really is an advanced project - search Google for: "S57MV
spectrum analyser" (without the quotes, and also look for the American
"analyzer" spelling).

Sounded interesting, but a google search as described turned up 0 results.
In fact a search for S57MV on its own got 0 results. Do you have any other
references?

73

Hans G0UPL




--
73 from Ian G3SEK 'In Practice' columnist for RadCom (RSGB)
Editor, 'The VHF/UHF DX Book'
http://www.ifwtech.co.uk/g3sek

"Ian White, G3SEK" wrote in message
...
Avery Fineman wrote:
In article , "Ian White, G3SEK"
writes:


[...]


But once you have a "fun" analyser giving you basic indications, it
inevitably draws you into asking more searching questions, like "How
much of a harmonic problem do I really have?" That's where fun analysers
come unstuck, because you don't know how much of those indicated
harmonics are real, and how much is being generated in the analyser
itself.

I think it does much better than just give general indications. It certainly
has opened my eyes to a whole new world inaccessible to me before having the
analyser. It's comparable to the change in my hobby life when I aquired my
first oscilloscope. You still enjoy the hobby with no oscilloscope, but when
you get one you feel like you'd been blind before, and can't imagine life
without it.

The other day I looked at the 39.5MHz IF in an old television tuned to a UK
TV station signal. I could see the video and sound subcarriers clearly
identifiable. Tuning the TV I could move the signals out of the Surface
Acoustic Wave (SAW) filter skirts and get a good idea of the shape of the
filter's response curve. Before I might have read about that stuff in
theory, now I can actually see it!

Under many cases it will be possible to identify any spurious signals
generated within the analyser itself. The input attenuator (in my case
0-63dB in 1dB steps) allows quite accurate measurement of the magnitude of
frequency components, and also allows reduction of strong signals. In many
cases this can be used to eliminate many of the spuri.

Or, if the signal being monitored is a VFO output for example, or another
variable frequency source, changing the frequency will often identify the
spurious signals. For example I might increase the VFO frequency, causing
the fundamental frequency peak to move left to right on the screen. If I see
some other peak move in the opposite direction, or if I see a peak moving
across the display very quickly, I know I've seen a spurious response.

Much of the spurious response problems come at the high end of the frequency
coverage range. For lower frequency work one could insert an additional
easily-constructed low pass filter and eliminate much of the problem.

So yes, no doubt the more effort you put in and the more complex the
spectrum analyser, the better the results obtainable. But even with a simple
analyser there's so much you can do.


An important practical crunch point for amateurs is "Does this
transmitter meet the FCC requirement for 60dB minimum harmonic
suppression?" Now you discover the big difference between 60-70dB
on-screen dynamic range, and 60-70dB *spurious-free* dynamic range. The
first is easy - the second is damned hard to guarantee in a HB design.

I could be wrong but I thought the FCC requirement was 40dB minimum harmonic
suppression, or 30dB for transmitters with output powers 5W or less? That's
well within the range of even a simple spectrum analyser. I doubt I'd have
any problem ensuring an HF QRP rig met FCC requirements with my analsyer.


If you get to that point - and many people will - then you'll wish you'd
taken the extra trouble to build an analyser that you could trust a
little more. The NE602 type categorically will not hack that kind of
problem, because the equipment under test is probably much cleaner than
it is! The W7ZOI type may just do it, with care, and the same is true of
the 141 generation. Only the best modern analysers will ace the problem
without needing careful attention from the user.

I'm already working on my Mk2 analyser. Not because I'm unhappy with the
measurements of the Mk1, but because I had so much fun with it, ending up
building it for its own sake, and not ready to stop having fun yet.


The simple rule of thumb is:

1. If the performance of your test equipment is better - in whatever
ways are important for the measurement in hand - than the performance of
the equipment under test (EUT) then the results will be reliable enough
for amateur work.

2. If your test equipment is comparable with the EUT, you'll get useful
indications.

3. If your test equipment is worse than the EUT, what you're actually
testing is the testgear itself! You'll learn nothing useful about the
EUT, and may actually be misled.

So aim for zone 1, settle for zone 2 if you must, but stay out of zone
3.

Not sure I agree completely - if the EUT is *supposed* to be in your zone 3,
but is not functioning properly, then even a simple spectrum analyser could
help you find the fault. It wouldn't help you make precise receiver
performance measurements on it when you corrected the fault, but would help
with diagnosis.

[...]


Bottom line: having played with the entry-level NE602 stuff myself, my
advice would be:

* If you only want a "fun" analyser, then go for it and have fun.

* But if you want the analyser as a tool to help you develop good radio
gear, go direct to the W7ZOI design.

Don't forget the W7ZOI analyser covers 0-70MHz only. You can double that
with a NE602 design. Nevertheless I liked the W7ZOI analyser a lot.


For a VHF/UHF HB design - which is obviously going to be a lot more
complex, and really is an advanced project - search Google for: "S57MV
spectrum analyser" (without the quotes, and also look for the American
"analyzer" spelling).

Sounded interesting, but a google search as described turned up 0 results.
In fact a search for S57MV on its own got 0 results. Do you have any other
references?

73

Hans G0UPL




--
73 from Ian G3SEK 'In Practice' columnist for RadCom (RSGB)
Editor, 'The VHF/UHF DX Book'
http://www.ifwtech.co.uk/g3sek

Ian White, G3SEK wrote:

For a VHF/UHF HB design - which is obviously going to be a lot more
complex, and really is an advanced project - search Google for: "S57MV
spectrum analyser" (without the quotes, and also look for the American
"analyzer" spelling).

I just tried the suggested Google search, and came up dry -- using
both spellings: analyzer/analyser. Since I cut-and-pasted "S57MV"
from your post, I doubt that it was a spelling error.

Any suggestions?

--
I don't need speed-reading,
I need speed-bookcase-building.
(with thanks to Nancy Lebovitz)

Ian White, G3SEK wrote:

For a VHF/UHF HB design - which is obviously going to be a lot more
complex, and really is an advanced project - search Google for: "S57MV
spectrum analyser" (without the quotes, and also look for the American
"analyzer" spelling).

I just tried the suggested Google search, and came up dry -- using
both spellings: analyzer/analyser. Since I cut-and-pasted "S57MV"
from your post, I doubt that it was a spelling error.

Any suggestions?

--
I don't need speed-reading,
I need speed-bookcase-building.
(with thanks to Nancy Lebovitz)

Hans Summers wrote:

[snipped much on which we either agree, or can amicably agree to differ]

Certainly, build something rather than nothing - as Hans says, it will
really open your eyes!


For a VHF/UHF HB design - which is obviously going to be a lot more
complex, and really is an advanced project - search Google for: "S57MV
spectrum analyser" (without the quotes, and also look for the American
"analyzer" spelling).

Sounded interesting, but a google search as described turned up 0
results. In fact a search for S57MV on its own got 0 results. Do you
have any other references?

Sorry, it was S53MV. A lot of the older information has disappeared off
the web, and is now only to be found on the site of VHF Communications
magazine where SM3MV's original article appeared.

http://www.vhfcomm.co.uk/articles.htm

Since the article is said to be too big for the web site (and the web
articles have no diagrams anyway) it seems that you cannot learn
anything useful about this excellent project without buying a back-issue
of the magazine :-(


--
73 from Ian G3SEK 'In Practice' columnist for RadCom (RSGB)
Editor, 'The VHF/UHF DX Book'
http://www.ifwtech.co.uk/g3sek

Hans Summers wrote:

[snipped much on which we either agree, or can amicably agree to differ]

Certainly, build something rather than nothing - as Hans says, it will
really open your eyes!


For a VHF/UHF HB design - which is obviously going to be a lot more
complex, and really is an advanced project - search Google for: "S57MV
spectrum analyser" (without the quotes, and also look for the American
"analyzer" spelling).

Sounded interesting, but a google search as described turned up 0
results. In fact a search for S57MV on its own got 0 results. Do you
have any other references?

Sorry, it was S53MV. A lot of the older information has disappeared off
the web, and is now only to be found on the site of VHF Communications
magazine where SM3MV's original article appeared.

http://www.vhfcomm.co.uk/articles.htm

Since the article is said to be too big for the web site (and the web
articles have no diagrams anyway) it seems that you cannot learn
anything useful about this excellent project without buying a back-issue
of the magazine :-(


--
73 from Ian G3SEK 'In Practice' columnist for RadCom (RSGB)
Editor, 'The VHF/UHF DX Book'
http://www.ifwtech.co.uk/g3sek

In article , "Ian White, G3SEK"
writes:

Avery Fineman wrote:
In article , "Ian White, G3SEK"
writes:

A spectrum analyser is simply a frequency-swept receiver with a
dB-scaled output to a screen. To give meaningful results, that receiver
must have a very high dynamic range with very low spurious responses.

"Meaningful results" are subjective to the hobbyist.

I'm an amateur, not a professional, but have paid my dues in this area.

Okay. Then allow me to state that I've been a professional in radio
and electronics for 51 years AND an electronics hobbyist in my
free time for longer. I've not "paid any dues" in the latter area except
a lot of hours spent having fun in my workshop..

My experience with the NE602 type is that it's great as a "fun"
analyser, but you can very quickly outgrow it. Yes, it will give you
general indications, like "Is that oscillator working?" or "Is that
final output stage generating harmonics?" Those indications even have
some numbers of MHz and dB attached.

? What is wrong with that?

An NE602 or SA602 is just a means to an end. It's a nice little IC
that allows some balanced mixing AND as the active device for the
local oscillator. A single IC, dual function. Used within its limitations
it is AS ACCURATE AS ANY OTHER DEVICE.

NO analyzer means NO results.

Of course I'd rather have that much information than no analyser at all.
(Len, if you happen not to mention something that is completely obvious,
I still assume that you're completely aware of it. Kindly extend the
same courtesy to me.)

I have NO idea of what is "completely obvious" to anyone. Everyone's
mileage varies.

But once you have a "fun" analyser giving you basic indications, it
inevitably draws you into asking more searching questions, like "How
much of a harmonic problem do I really have?" That's where fun analysers
come unstuck, because you don't know how much of those indicated
harmonics are real, and how much is being generated in the analyser
itself.

Let's not lump all the "non-pro" analyzers into the "fun-toy" category,
okay?

Sorry, but I can quickly categorize/calibrate/check ANY analyzer as
to whether or not whatever it shows is "real" or internal. There's lots
of written information available on how to do that...and a bit of thinking
will reveal the very same thing.

An important practical crunch point for amateurs is "Does this
transmitter meet the FCC requirement for 60dB minimum harmonic
suppression?" Now you discover the big difference between 60-70dB
on-screen dynamic range, and 60-70dB *spurious-free* dynamic range. The
first is easy - the second is damned hard to guarantee in a HB design.

Sorry, you've drifted outside of specific spectrum analyzer use. The
harmonic content of ALL transmitters can be measured WITHOUT
any spectrum analyzer, as accurately (in most cases moreso) as with
an analyzer. A spectrum analyzer is much much faster in that
application, but harmonic content measurement is not dependent on
its use.

Good operation and good specifications in homebrew designs are as
good as the designer, as good as the other equipment for calibration
and characterisation that is used. It's all together. It's only "damned
hard to guarantee" if there is nothing available to ascertain operation
and calibration. That can exist equally well for an HP 141 plug-in or
one that you term a "fun" NE602.

If you get to that point - and many people will - then you'll wish you'd
taken the extra trouble to build an analyser that you could trust a
little more. The NE602 type categorically will not hack that kind of
problem, because the equipment under test is probably much cleaner than
it is!

Good grief! There's plenty of instances in metrology where actual
measurements are done to a higher level than what each piece of
test equipment can do. Time and frequency measurement is a good
area for examples in that, other areas' explanations will take more
time to describe than I have at the moment.

EVERY single piece of test equipment MUST be used within its
limitations. That INCLUDES your "HP 141" racks, Anritsu analyzers,
etc., and whatever else is there.

The W7ZOI type may just do it, with care, and the same is true of
the 141 generation. Only the best modern analysers will ace the problem
without needing careful attention from the user.

Nonsense. Each and every "user" had damn well pay close attention
to using whatever they have within its equipment's limitations. That
holds true for the best of Rhode & Schwarz wares AND for the "fun"
SA602/NE602 analyzers.


Very easily, in the same ways as you test a receiver for strong-signal
handling.... but most graphically by looking at the same spectra with
two analysers side-by-side. The one showing fewest signals is the one
you can rely on most.

Have you priced the used spectrum analyzers lately? Do you expect
others to have ready access to "another" spectrum analyzer?

That comment was mostly to make the point that the cleanest display is
the most real.

You're going to have to explain that better..."clean" is highly subjective
and subjective judgement in any metrology effort is a no-no.

The absolute standard of comparison is the perfect spectrum analyser
with no spurious responses.

Okay, show me one. They must be in storage with bottles of the
"universal solvent" and other fine ideas...


When I built the first "fun" analyser (which had to wait until the
hardware collection had expanded to include a good oscilloscope) I found
it quite disappointing, because it still wasn't showing me reality. I
was still having to think "Now if I had a *good* spectrum analyser, what
would it probably be showing me here?"

That's nice. But you are drifting towards extreme categorization of
equipment. Simple, low parts count home projects are "fun" (in your
terms, like toys or something for recreation. The only "GOOD" gear
is semi-pro, expensive, etc., etc., etc.


Out in the workshop, the real hardware adds real-life numbers to the
ideas, and also checks that there isn't anything happening that I hadn't
thought of.

In MY workshop, nearly everything I do and use is pure FUN. If I use
them within their limitations, then they give me REAL numbers as
valid as anything I use from a small pro lab's $400K capital collection.


In this particular case, there does exist a very good, reliable,
second-level design (the W7ZOI). It's a little more expensive to build
than the entry-level stuff, but not significantly more complex. It's
very well designed, so while you're building it you'll learn some very
useful things from a top amateur (and former professional) RF designer.
It's also modular, so you can build it in stages, and upgrade various
parts as and when you want.

Bottom line: having played with the entry-level NE602 stuff myself, my
advice would be:

* If you only want a "fun" analyser, then go for it and have fun.

* But if you want the analyser as a tool to help you develop good radio
gear, go direct to the W7ZOI design.

So...bottom line is that spectrum analyzers with NE602s in them are
kiddie toys useable only for entertainment and recreation?!?

Not a good characterization, that. Used within its limitations, the
"kiddie" instrument is as accurate as anything else. Input level
displayed linearity is a direct function of the successive detection
logarithmic chip used in the back end...has very little to do with the
front end. Resolution is again a back-end concern and is a direct
function of the filtering of the final IF in the analyzer...and the sweep
rate and sweep span (all related). If you want linearity in the
horizontal as to frequency, then you have to pay a lot of attention to
VCO (or equivalent) control and that has little to do with messing
about with the front end mixer.

The front end of any spectrum analyzer is the place where your
spurious responses are generated, if any at all. If you are going for
the very best in higher-level inputs then you absolutely need a high-
IM-rated mixer, which also probably needs a high-level LO. Both of
those cost a great deal of money...and the higher input level free of
spurious responses may or may not be worth it. It may not be
worth it when one of those "fun" analyzers don't exhibit any false
signals or spurii below a specific input level.

Len Anderson
retired (from regular hours) electronic engineer person
formerly an Associate Editor with HAM RADIO Magazine

PS: I once or twice used an HP 141. It always had "ordinary"
oscilloscope plug-ins in it to make it overall an oscilloscope. Shrug.