wrote in message
...

If you have log before ADC I think you need a superhet conversion rather
than direct conversion to baseband, for the reasons I mentioned earlier.
If
you just ADC'ed the lot there'd be some chance of sorting the confusion
out
in software, though it'd take a brave heart to try it.

So use 2 18-bit ADC's (sound card maybe 16-bit but you won't get 90dB
range from
it, you'll find a fair bit of noise from your average sound card - well
the
first 3 bits will be noisy), one with the I and the other with the Q (from
the
DC output) and do it that way ?

I still think you stand a fair chance of problems, when essentially you're
doing your signal strength measurement in baseband and sweeping at rates in
the same frequency range. The aquisition time for the signal strength
measurement needs to be much shorter than the sweep, or the signal strength
will change while you're attmpting to measure it.

I guess it's a bit like the problem of designing audio-derived AGC in direct
conversion receivers. It takes too many cycles to measure the audio volume
at low frequencies to get a correction signal to control gain. Similar
problem. There might be a way of disentangling it in software, I don't know
the mathematics but I can intuitively imagine that it could work.

But I'm no expert on this, just a clown with soldering iron...

Hans G0UPL
http://www.hanssummers.com

hans is right about the PC being a kludge when compared to an
oscilloscope. I live in India and second-hand oscilloscopes are a
rarity. The new ones cost an engineer's whole year's salary. I have
purchased a tektronix 454 last month on ebay for $300, it is still on
its way to India. So, while i can imagine that in a number of more
developed countries oscilloscopes are not really a problem, they do
remain a problem in many parts of the world.

That puts another angle on it. In that case I'd say you could produce a nice
instrument by using a complete analogue analyser such as my design
http://www.hanssummers.com/electroni...yser/index.htm
or similar, and feeding the sweep (X-axis) and logarithmic output (Y-axis)
into the PC via ADC's on the parallel port. 8-bit ADC's would probably be
marginally good enough, 10 bit for sure.


More often than not, an amateur already has a PC. Adding a simple
hardware to be able to do quick and dirty spectrum analysis might be
an interesting option. As Wes writes in SSD and EMRFD, the purpose of
test instruments is to help with the projects on hand, rather than be
projects themselves (then he went on to homebrewing a spectrum
analyser, hehe).

It became a project in its own right for me. So much that it spawned a
second project
http://www.hanssummers.com/electroni...ser2/index.htm.
But not wasted time - I learnt so much making it and got my first experience
at VHF work etc, so it was time well spent.

Hans G0UPL
http://www.hanssummers.com

hans is right about the PC being a kludge when compared to an
oscilloscope. I live in India and second-hand oscilloscopes are a
rarity. The new ones cost an engineer's whole year's salary. I have
purchased a tektronix 454 last month on ebay for $300, it is still on
its way to India. So, while i can imagine that in a number of more
developed countries oscilloscopes are not really a problem, they do
remain a problem in many parts of the world.

That puts another angle on it. In that case I'd say you could produce a nice
instrument by using a complete analogue analyser such as my design
http://www.hanssummers.com/electroni...yser/index.htm
or similar, and feeding the sweep (X-axis) and logarithmic output (Y-axis)
into the PC via ADC's on the parallel port. 8-bit ADC's would probably be
marginally good enough, 10 bit for sure.


More often than not, an amateur already has a PC. Adding a simple
hardware to be able to do quick and dirty spectrum analysis might be
an interesting option. As Wes writes in SSD and EMRFD, the purpose of
test instruments is to help with the projects on hand, rather than be
projects themselves (then he went on to homebrewing a spectrum
analyser, hehe).

It became a project in its own right for me. So much that it spawned a
second project
http://www.hanssummers.com/electroni...ser2/index.htm.
But not wasted time - I learnt so much making it and got my first experience
at VHF work etc, so it was time well spent.

Hans G0UPL
http://www.hanssummers.com

Granted, one problem is how fast the data transfer is through the
parallel
port, that would limit how many samples/second the system could process.

I think a parallel port is easily fast enough for this application


Almost forgot... WIN XP and WIN2000 make getting to the parallel port
for
something like this MUCH more difficult then it was under WIN 98.

True enough. I had this problem with a different project. There's a
shareware driver available which helps access the parallel port under XP
etc. See http://www.hanssummers.com/electroni...trix/index.htm for
details of how it works.

Hans G0UPL
http://www.hanssummers.com

Granted, one problem is how fast the data transfer is through the
parallel
port, that would limit how many samples/second the system could process.

I think a parallel port is easily fast enough for this application


Almost forgot... WIN XP and WIN2000 make getting to the parallel port
for
something like this MUCH more difficult then it was under WIN 98.

True enough. I had this problem with a different project. There's a
shareware driver available which helps access the parallel port under XP
etc. See http://www.hanssummers.com/electroni...trix/index.htm for
details of how it works.

Hans G0UPL
http://www.hanssummers.com

I still think you stand a fair chance of problems, when essentially you're
doing your signal strength measurement in baseband and sweeping at rates in
the same frequency range. The aquisition time for the signal strength
measurement needs to be much shorter than the sweep, or the signal strength
will change while you're attmpting to measure it.

Yes Sweep rate can be a problem, but it's all down the baseband bandwidth and
what kind of update rate you want/need.

If you want to sweep across a 100MHz within say a second, assuming your using
44100 sampling rate, that's 44100 samples you'll get across the 100MHz sweep in
that one second - 2.26Khz wide freq segments (non-overlapping), but you'd need a
baseband bandwidth of 1KHz to overcome the anti-alias problem.

Doesn't sound to bad really. A one second scan across the entire 100MHz is fine
really (depending on what your doing) - the PC is a perfect storage scope.

To convert a linear I/Q baseband sample to a log scale is no problem at all in
software.

The software cud easily deal with any variation in VCO/Mixer level differences
across the whole band.

Best to have the PC control the vco though, then as you say, you can zoom in on
a desired freq range etc.

Clive

I still think you stand a fair chance of problems, when essentially you're
doing your signal strength measurement in baseband and sweeping at rates in
the same frequency range. The aquisition time for the signal strength
measurement needs to be much shorter than the sweep, or the signal strength
will change while you're attmpting to measure it.

Yes Sweep rate can be a problem, but it's all down the baseband bandwidth and
what kind of update rate you want/need.

If you want to sweep across a 100MHz within say a second, assuming your using
44100 sampling rate, that's 44100 samples you'll get across the 100MHz sweep in
that one second - 2.26Khz wide freq segments (non-overlapping), but you'd need a
baseband bandwidth of 1KHz to overcome the anti-alias problem.

Doesn't sound to bad really. A one second scan across the entire 100MHz is fine
really (depending on what your doing) - the PC is a perfect storage scope.

To convert a linear I/Q baseband sample to a log scale is no problem at all in
software.

The software cud easily deal with any variation in VCO/Mixer level differences
across the whole band.

Best to have the PC control the vco though, then as you say, you can zoom in on
a desired freq range etc.

Clive

wrote in message
news

I still think you stand a fair chance of problems, when essentially
you're
doing your signal strength measurement in baseband and sweeping at rates
in
the same frequency range. The aquisition time for the signal strength
measurement needs to be much shorter than the sweep, or the signal
strength
will change while you're attmpting to measure it.

Yes Sweep rate can be a problem, but it's all down the baseband bandwidth
and
what kind of update rate you want/need.

If you want to sweep across a 100MHz within say a second, assuming your
using
44100 sampling rate, that's 44100 samples you'll get across the 100MHz
sweep in
that one second - 2.26Khz wide freq segments (non-overlapping), but you'd
need a
baseband bandwidth of 1KHz to overcome the anti-alias problem.

The problem I'm having seeing how it can work is this. If you take 44100
measurements per second, that's one measurement every 27uS. But you've low
pass filtered your baseband at 1KHz. It would take at least one of those
cycles to measure the envelope amplitude with any degree of accuracy but
you're allowing just 2.7% of a single 1KHz sample, how does that work? Or am
I looking at it too simplistically?

I certainly agree that the PC can make a nice storage and display device,
specially if 'scopes aren't so easily available. Seems to me a question of
how much of the signal chain is implemented in analogue and how much in
digital. I Just think it saves an awful lot of hassle to add that little
extra analogue stage before you go digital, i.e. 2nd IF and detector. Log
could be done on the digital side no problem if desired provided enough ADC
resolution was available. VCO/Mixer level differences could still be
compensated in software, and the PC control the VCO.

Hans G0UPL
http://www.hanssummers.com

wrote in message
news

I still think you stand a fair chance of problems, when essentially
you're
doing your signal strength measurement in baseband and sweeping at rates
in
the same frequency range. The aquisition time for the signal strength
measurement needs to be much shorter than the sweep, or the signal
strength
will change while you're attmpting to measure it.

Yes Sweep rate can be a problem, but it's all down the baseband bandwidth
and
what kind of update rate you want/need.

If you want to sweep across a 100MHz within say a second, assuming your
using
44100 sampling rate, that's 44100 samples you'll get across the 100MHz
sweep in
that one second - 2.26Khz wide freq segments (non-overlapping), but you'd
need a
baseband bandwidth of 1KHz to overcome the anti-alias problem.

The problem I'm having seeing how it can work is this. If you take 44100
measurements per second, that's one measurement every 27uS. But you've low
pass filtered your baseband at 1KHz. It would take at least one of those
cycles to measure the envelope amplitude with any degree of accuracy but
you're allowing just 2.7% of a single 1KHz sample, how does that work? Or am
I looking at it too simplistically?

I certainly agree that the PC can make a nice storage and display device,
specially if 'scopes aren't so easily available. Seems to me a question of
how much of the signal chain is implemented in analogue and how much in
digital. I Just think it saves an awful lot of hassle to add that little
extra analogue stage before you go digital, i.e. 2nd IF and detector. Log
could be done on the digital side no problem if desired provided enough ADC
resolution was available. VCO/Mixer level differences could still be
compensated in software, and the PC control the VCO.

Hans G0UPL
http://www.hanssummers.com

This is a very interesting discussion
I have thought about doing this in the past, but have never been brave
enough!
What about a scheme like:
Input - attenuator - LO/mixer-LPF/if amp -Direct coversionLO/Mixer
- switchable LPF to say 150KHz - Log det (broadband)- ADC - software
/PC

For a 100MHz sweep you would probably need 300KHz bandwidth max which
would be achieved by a 150KHz DSB receiver.
You could go higher than 100MHz as the first IF - use ring diode mixer
to a helical filter as the roofing filter
I would be interested in a cooperative project

Richard



Ashhar Farhan wrote:
"Hans Summers" wrote in message news:bmj291


Interesting ideas Len. I guess the idea of an all-digital spectrum analyser
is similar to that of an all-digital HF amateur radio tranceiver. It can be
done but at the current state of the art, it's a difficult proposal for the
hobbyist and certainly difficult to obtain the same level of performance as
the equivalent analogue device for the same amount of cost and/or effort.


I have a PC-based oscilloscope that does something pretty close. It
digitizes at the rate of 4ns per sample (taking 8bit samples
unfortunately) and generates an FFT display of magnitude/power
spectrum/power density).


I think it is feasible to use an analogue RF front end under computer
control, the a PC controlling the VCO and sampling the logarithmic output.
In essence just replacing the oscilloscope as the display system.


I guess, it is not necessary to PC control the VCO. Sweep generators
are easy to come by. probably, if there is a way to feed the sweep
into the PC to generate the X-axis, then that might be a better
alternative.

- farhan