Fw: Re: [Test-Equipment] Spectrum Analyzer BW measurement
Gary Schafer
[email protected]
Fri, 30 Jan 2004 11:38:01 -0500
Hi Kees,
Ok I think I understand now what you are asking.
Look at it this way: If you use your receiver with a 2 khz filter in it
and tune across an ssb signal it appears to be a certain number of khz
wide. If you then switch to an AM filter that is 6 khz wide and tune
across that same ssb signal it appears to be a much wider signal.
The same thing is happening in the spectrum analyzer.
Take that same receiver and tune across a high speed data signal that is
exactly 10 khz wide. When your 2 khz filter is centered at the lower
edge of the signal your dial is reading the lower frequency edge of the
signal. But your filter has some width (2 khz) so you can tune the dial
1 khz lower still and the lower edge of the signal is still (just) in
the band pass of the filter. You can do the same thing at the upper edge
of the signal. Center the filter on the upper edge of the signal. You
can then tune an additional 1 khz above the edge and still have the
upper edge of the signal in the band pass.
Now if you do the same exercise with a 5 khz filter you will still find
the lower edge of the signal where the center of the band pass of the
filter is (if you are able to determine where the center of the band
pass is). You will now be able to tune an additional 2 khz lower in
frequency and still have the lower edge of the signal in the band pass.
Same for the high side of the signal. You can tune an additional 2 khz
above and still have the signal in the band pass.
So if you tune so that the signal is at the very edge of the band pass
you can see that your dial will show that the signal appears to be wider
with the wider filter than with the narrower one.
With the 2 khz filter the band width of the signal appears to be 12 khz
wide. With the 5 khz band width filter that same signal appears to be 15
khz wide.
If you subtract the width of the filter from the observed width of the
signal that will give you the true width of the signal. 1/2 of the
filter width affects one edge and the other half of the filter affects
the other edge of the signal.
Using the narrowest filter possible gives a truer signal width without
subtracting filter width.
Of course the filters are not brick wall filters and do not have perfect
shape factor so if you want to really get accurate you will need to take
the slope of the filter into account too.
To see the shape of the filter in the analyzer just tune into an
unmodulated carrier or the zero frequency signal (local oscillator) on
the analyzer. With a 10 khz filter switched in you will see a signal
that looks like it is 10 khz wide. Switch to a 500 hz filter and the
signal now looks to be 500 hz wide (at the -3 db points).
A steady unmodulated signal has zero bandwidth (unless it is noisy) so
any apparent width that you see is the filter width.
Good way to check your internal filters!
73
Gary K4FMX
[email protected] wrote:
> Thanks Gary, best answer yet although I have not dug through the HP info
> suggested by several.
>
> What I'm trying to do is measure the BW of a radio control transmitter.
> The FCC specifications are: For a 20Khz signal frequency spacing,
> various BWs at -25dB, -45dB and -55dB. Lets take the -55dB at 20Khz using
> a 3Khz RBW as specified by the AMA. My SA has automatic filter selection
> of 5Hz, 50Hz, 500Hz, 5Khz, 50Khz. For this measurement I select a scale
> of 5Khz /Div which gives a "normal" -3dB RBW of 500Hz. Sounds good to me
> and gives a great 2.5% RBW to BW ratio. The SA also allows selection of
> 10x (5Khz) or 1/10x (50Hz) RBW. 50Hz is, of course, way too slow, but 10x
> gives a MUCH WIDER BW reading (2x ???). So I'm wondering if by using the
> 500Hz RBW and if the test spec used a 3Khz RBW are we supposed to measure
> the same BW ? I don't think so.
>
> Question:
> So how do I correlate between the two BWs, never mind that I think I have
> the more accurate BW representation using the 500Hz RBW.
>
> All BW measurements are made in "peak detect" mode.
>
> --------- Forwarded message ----------
> From: Gary Schafer <[email protected]>
> To: [email protected]
> Cc: [email protected]
> Date: Thu, 29 Jan 2004 21:49:14 -0500
> Subject: Re: [Test-Equipment] Spectrum Analyzer BW measurement
> Message-ID: <[email protected]>
> References: <[email protected]>
>
> Kees,
>
> Not sure exactly what you are trying to measure?
> I think what you are reading is related to the minimum bandwidth filter
> that can be used when looking at a given spectrum width. It has to do
> with how fast you are moving the filter across the signal.
> The more spectrum you try to see with a given bandwidth filter the
> slower you have to sweep it.
>
> Think of it as how long the particular signal remains in the filter band
> pass. The wider the spectrum you try to look at the faster you need to
> sweep it so the display does not blink from the sweep. If you use a very
> narrow filter when sweeping a lot of spectrum then a particular signal
> does not spend much time in the filter band pass. Not enough energy gets
> through. Using a wider filter will allow the signal more time in the
> band pass and thus more energy to get through.
>
> Using manual modes it is easy to see. Look at a signal with a rather
> narrow filter. Then speed up the sweep. When the sweep speed starts to
> get too fast the amplitude of the signal will start to drop.
> If you slow the sweep down to the point just before the signal starts to
> drop in amplitude and switch in a narrower filter you will again see the
> signal amplitude drop. You would then need to slow down the sweep speed
> further to bring the amplitude back to normal.
>
> Also narrowing the sweep width (look at a narrower part of the spectrum)
> will allow a narrower filter to be used. Again, as you look at a smaller
> part of the spectrum that allows more time for the signal to remain in
> the filter band pass.
>
> Most analyzers automatically switch filters and sweep speed as you
> change sweep widths. The wider the sweep width you select the wider the
> filter that is switched in. They do this so that the minimum filter
> width is used. If you go below that minimum width you will not get a
> true representation on the screen. You will not see the full amplitude
> that is really there.
>
> Most have a manual mode that allows any filter at any sweep width and
> any sweep speed as I described above. You have to determine how far you
> can go in the manual mode and still have a true picture.
>
> Hope this is what you are referring to.
>
> 73
> Gary K4FMX
>
>
> [email protected] wrote:
>
>>I'm trying to measure some 10-30Khz bandwidths. As I understand, the
>>correct method is to use the instrument's -3dB resolution BW (RBW) of
>>approx 1% ?? of what you are trying to measure. This means a RBW of
>>approx
>>0.1-0.3Khz. This instrument has and RBW of 5Hz, 50Hz, 500Hz, 5Khz, and
>>50Khz so I use 500Hz. If I use 5Khz the BW shown nearly doubles. I
>
> expect
>
>>it to get wider but why twice ?. And if it's twice, it seems like I'd
>
> get
>
>>widely variant BW measurements depending on if I'm able to select
>
> 500Hz,
>
>>1Khz or 3Khz or ? bandwidths. Or the BW would always have to be
>
> specified
>
>>with a RBW ...which begs the question "What do you do if the spec is at
>>3Khz RBW and the closest you can get is 500Hz and 5Khz ? Do you assume
>>the narrower RBW (500Hz) is more accurate anyway ?
>>
>>Please send the resonse to me directly since I'm set to receive daily
>>summary notes.
>>
>>73 Kees K5BCQ
>>_______________________________________________
>>Test-Equipment mailing list
>>[email protected]
>>http://mailman.qth.net/mailman/listinfo/test-equipment
>>
>
> _______________________________________________
> Test-Equipment mailing list
> [email protected]
> http://mailman.qth.net/mailman/listinfo/test-equipment
>