Low cost spectrum analyzer <100MHz?

[doug]

It sure does. However, most of the units available today are equipped
with the 8554B plug-in that goes up to 1.3GHz. I am used to the 8553B
(to 110MHz max). Do you remember how well the 8554B does when most of
the stuff you have to measure is between 10MHz and 100MHz? I know it
covers that range but it'll be at the bottom 10% of the dial range.

Both the 8553 and 8554 work pretty well. For the lower frequency stuff,
the dials are more convenient on the 8553 since it has a 11MHz and a
110MHz range. There should be lots of 8553 plugins available. I gave
away four or five of them a couple of months ago. The limiting factor
is the 8552 IF unit. The B is far superior to the A. For the 8556
low frequency analyzer, the 8552a is useless as the drift is too high.
The signals are converted to a higher IF and then downconverted and
detected. The if filters are in the IF unit and the B has narrower
filters and a more stable oscillator. The 141 mainframes have various
sets of problems but are probably tolerable. The storage does not
work very well. I believe that some of the units were equipted to
work with a 8750 storage normalizer, a digital storage and display
adapter.

You might also look at the 8569 analyzer. It goes to 20GHZ but the
basic analyzer is a 1.7GHz unit. They work pretty well and the display
is digital. They are weak below 10MHz though.

The 8562 is quite nice and goes from about 50khz to 22GHz (all numbers
from memory and not guaranteed). These go for a few thousand on ebay.
The trouble is that fixing them is a difficult job. The if filters are
varacter tuned for the different widths. Mine has an issue with the
100khz filter.

 

[Joerg]

Both the 8553 and 8554 work pretty well. For the lower frequency stuff,
the dials are more convenient on the 8553 since it has a 11MHz and a
110MHz range. There should be lots of 8553 plugins available. I gave
away four or five of them a couple of months ago. The limiting factor
is the 8552 IF unit. The B is far superior to the A. For the 8556
low frequency analyzer, the 8552a is useless as the drift is too high.
The signals are converted to a higher IF and then downconverted and
detected. The if filters are in the IF unit and the B has narrower
filters and a more stable oscillator. The 141 mainframes have various
sets of problems but are probably tolerable. The storage does not
work very well. I believe that some of the units were equipted to
work with a 8750 storage normalizer, a digital storage and display
adapter.

Thanks, Doug. It's good to know that we can also use the 8554B even if
we never exceed 100MHz. I remember that storage was already an issue
when the units were fairly knew. But I never relied on it anyway and we
don't have to here. All we need is noise floor measurements and even
there the NF of the analyzer input doesn't matter because I've put
preamps in the client's system.

You might also look at the 8569 analyzer. It goes to 20GHZ but the
basic analyzer is a 1.7GHz unit. They work pretty well and the display
is digital. They are weak below 10MHz though.

The 8562 is quite nice and goes from about 50khz to 22GHz (all numbers
from memory and not guaranteed). These go for a few thousand on ebay.
The trouble is that fixing them is a difficult job. The if filters are
varacter tuned for the different widths. Mine has an issue with the
100khz filter.

Haven't seen any 8569 analyzers for sale lately. But 20GHz would be
total overkill here anyway. Those should get into the hands of people
who really need that.

The amazing thing is that when I used that 141T about 25 years ago the
reason many others didn't want it and didn't pester me to "get done with
my stuff" was that it doesn't offer push button control. Everyone wanted
to be able to punch in start, stop, RBW, VBW and so on. Not use dials
and rotary switches. Like when they see my car. "Oh, a stick shift?
Yech, I wouldn't want to drive that."

 

[Joerg]

I have a great love of setting my own parameters and have hated the auto
setup things. When people use autosettings, they tend to stop thinking.

The dynamic range on the 8553/8554 is as good as any you are going to
find especially when you have a 8552B if unit and the narrow resolution
bandwidth. Because they did not have a lot of computer correction, they
had to make the basics work well.

Yep, and they sure did work well. I was very happy with that 141T back
then. Especially because it wasn't gone when I came in the next morning
like most other gear.

The 8569 goes to 20GHz on harmonic mixing. The basic unit is 1.7GHz so
it is much like a 8554 with a digital display. They show up on ebay for
$1000 or so on occasion. If you do not need a narrow resolution
bandwidth, the 8559 in an 853 digital mainframe is nice and they go for
less than $1000. For my home lab I have a 8562 with a couple of 8569s
and a 8559 for backup. After I got the 8562, I got rid of about a dozen
sets of 140/141/8553/8664/8555/8556 sets. A few months ago I could have
sent you several sets. ...

Sigh. I should have known...

... Time never works the way you want. It is the
same when you throw out old parts you have not seen in years and then
find a need for them the next day.

Tell my wife. She is also in charge of things when my lab becomes "too
full"

On ebay 141/8553/8552 systems are selling for $200 to $400. You can
afford to get a couple to use for spares at that price.

Today there was a 141/8553/8552 that went for $400 despite the fact that
the seller stated that he couldn't get a trace onto the screen. Blew me
away. That can mean anything. Either a loose switch which is fixed in a
few minutes or a linear regulator could have gone lala and fried
anything downstream of it.

 

[doug]

Thanks, Doug. It's good to know that we can also use the 8554B even if
we never exceed 100MHz. I remember that storage was already an issue
when the units were fairly knew. But I never relied on it anyway and we
don't have to here. All we need is noise floor measurements and even
there the NF of the analyzer input doesn't matter because I've put
preamps in the client's system.



Haven't seen any 8569 analyzers for sale lately. But 20GHz would be
total overkill here anyway. Those should get into the hands of people
who really need that.

The amazing thing is that when I used that 141T about 25 years ago the
reason many others didn't want it and didn't pester me to "get done with
my stuff" was that it doesn't offer push button control. Everyone wanted
to be able to punch in start, stop, RBW, VBW and so on. Not use dials
and rotary switches. Like when they see my car. "Oh, a stick shift?
Yech, I wouldn't want to drive that."

I have a great love of setting my own parameters and have hated the auto
setup things. When people use autosettings, they tend to stop thinking.

The dynamic range on the 8553/8554 is as good as any you are going to
find especially when you have a 8552B if unit and the narrow resolution
bandwidth. Because they did not have a lot of computer correction, they
had to make the basics work well.

The 8569 goes to 20GHz on harmonic mixing. The basic unit is 1.7GHz so
it is much like a 8554 with a digital display. They show up on ebay for
$1000 or so on occasion. If you do not need a narrow resolution
bandwidth, the 8559 in an 853 digital mainframe is nice and they go for
less than $1000. For my home lab I have a 8562 with a couple of 8569s
and a 8559 for backup. After I got the 8562, I got rid of about a dozen
sets of 140/141/8553/8664/8555/8556 sets. A few months ago I could have
sent you several sets. Time never works the way you want. It is the
same when you throw out old parts you have not seen in years and then
find a need for them the next day.

On ebay 141/8553/8552 systems are selling for $200 to $400. You can
afford to get a couple to use for spares at that price.

 

[doug]

Yep, and they sure did work well. I was very happy with that 141T back
then. Especially because it wasn't gone when I came in the next morning
like most other gear.





Sigh. I should have known...



Tell my wife. She is also in charge of things when my lab becomes "too
full"



Today there was a 141/8553/8552 that went for $400 despite the fact that
the seller stated that he couldn't get a trace onto the screen. Blew me
away. That can mean anything. Either a loose switch which is fixed in a
few minutes or a linear regulator could have gone lala and fried
anything downstream of it.

I have asked the givee of the analyzers how attached he is to one of
them. I will let you know what he says. The 8553 were in good shape
and the 8554 was in a cal lab where no one knew how to use it so it
is in almost new shape. There also was a lovely 432/478 power meter
that probably has an hour on it but it has no resale value these days.

 

[Joerg]

I have asked the givee of the analyzers how attached he is to one of
them. I will let you know what he says. The 8553 were in good shape
and the 8554 was in a cal lab where no one knew how to use it so it
is in almost new shape. There also was a lovely 432/478 power meter
that probably has an hour on it but it has no resale value these days.

Yes, please let me know.

Analog power meters can be great. I have one from Rhode&Schwarz and it's
the only thing that really works when noise is totally random. That can
freak out analyzers. This meter has that traditional old mirror display.
When I bought it the thing looked like it was never used. Not a scratch
yet they almost gave it away. It was a company closure, one of the best
opportunities to buy equipment. AFAIR it was a silent auction. Good
bargains but it can be boring, I really like it Texas style ;-)

 

[RST Engineering \(jw\)]

Good God, man, the next time you throw a generous fit, would you keep those
of us scraping by with old systron-donners in mind?

Jim




After I got the 8562, I got rid of about a dozen

 

[Oliver Betz]

Question: Is there a low cost spectrum analyzer that allows low level
noise measurements? Roughly like this:

Check http://www.hameg.com/16.0.html - misses "less than 9kHz
bandwidth" but has a "video filter" to reduce displayed noise.

Oliver

 

[Don Lancaster]

Joerg wrote:




Check http://www.hameg.com/16.0.html - misses "less than 9kHz
bandwidth" but has a "video filter" to reduce displayed noise.

Oliver

An ordinary UHF tv tuner can be converted into a low cost spectrum analyzer.

Construction projects appeared a decade ago, possibly in RE.

--
Many thanks,

Don Lancaster voice phone: (928)428-4073
Synergetics 3860 West First Street Box 809 Thatcher, AZ 85552
rss: http://www.tinaja.com/whtnu.xml email: [email protected]

Please visit my GURU's LAIR web site at http://www.tinaja.com

 

[Joerg]

Joerg wrote:




Check http://www.hameg.com/16.0.html - misses "less than 9kHz
bandwidth" but has a "video filter" to reduce displayed noise.

We've bought and used it in another company. But it's really more like
an EMC pre-compliance machine. Nice quality. Well, after all, it's
Hameg. My very first scope from them is around 30 years old and works
like new.

VBW can't really make up for missing RBW if there are carriers present
in the band. My client's site has some big transmitters next door :-(

 

[Joerg]

An ordinary UHF tv tuner can be converted into a low cost spectrum
analyzer.

Construction projects appeared a decade ago, possibly in RE.

Problem is, the dynamic range of the average TV tuner ain't nothing to
write home about.

 

[Robert Latest]

["Followup-To:" header set to sci.electronics.design.]

Nice quality. Well, after all, it's Hameg.

I think Hamegs are great. I've recently bought some Teks from the "cheap"
end of their lightweight, 4-channel LCD ones, but I prefer the old Hamegs
analog/digital combo scopes.

robert

 

[Joerg]

["Followup-To:" header set to sci.electronics.design.]
Joerg wrote:

Nice quality. Well, after all, it's Hameg.

I think Hamegs are great. I've recently bought some Teks from the "cheap"
end of their lightweight, 4-channel LCD ones, but I prefer the old Hamegs
analog/digital combo scopes.

Same here. Somehow those classy green/blue CRTs let you see more nuances
in a signal. Try to detect a wee modulation on an LCD.