EMRFD Message Archive 8448

Message Date From Subject
8448 2013-04-01 19:34:27 Jerry Haigwood Spectrum Analyzer
Hi All,

I have noticed that on the auction site there are some low cost spectrum
analyzers from China. These all seem to be made by a company named ATTEN.
Their lowest cost model is the AT5005. It is a 150KHz to 500 MHz model with
a digital readout for center frequency and marker frequency. The amplitude
input is from -100 dbm to +13 dbm. The cost is $357 + $114 shipping. Has
anyone on this list purchased one? Do you know of anyone who has purchased
one? What do you think of it, etc.?

Jerry W5JH

"building something without experimenting is just solder practice"





[Non-text portions of this message have been removed]
8449 2013-04-01 19:58:55 Ray Re: Spectrum Analyzer
Contact Todd, VE7BPO at his website (http://www.qrp.pops.net/blog.asp.) He has the AT6011 and doesn't think much of it. He suggests the DSA-815 which was reviewed in QST a couple of months back. I am seriously considering the DSA-815. However, it is quite a bit more money - $1495 with the tracking generator option.

Ray
AB7HE

8450 2013-04-01 21:34:11 Kerry Re: Spectrum Analyzer
I really like my DSA-815 (with TG), bought about 6 months ago; I sold my HP141 systems for A$500 to help fund it.

At its price I think it's a true bargain.

The only area in which it's not as good as expensive HP etc SAs is phase noise; it's still good though and is fine for amateur/experimenter use.

Kerry VK2TIL.
8457 2013-04-04 13:58:28 michael taylor Re: Spectrum Analyzer
8458 2013-04-04 22:27:57 Kerry Re: Spectrum Analyzer
The GSP-730 has a minimum RBW of 30 kHz.

The strength of the DSA-815 is in its RBW which goes down to 100 Hz; whilst this is not as good as the 1 Hz minimum RBW found in high-end SAs (which can cost as much as a good car) it's a very good spec for the price and makes the DSA very useful for the average experimenter.

Here it is displaying two signals about 500 Hz apart;

http://i46.tinypic.com/n5kmxs.jpg

[The displayed frequencies are interesting; they do not agree with the digital readouts of the two generators used. Since two of my instruments (HP5345A counter and HP 8657B generator) have the high-stability reference oscillator option I think it's time I built a distribution amplifier so that all these instruments can be singing the same song. :) ]

The DSA should be OK for two-tone SSB testing; I have done amplifier IP3 tests with good results;

http://i48.tinypic.com/20a29tc.jpg

My decision on a SA came-down to the Rigol or the Signal Hound; I decided that I prefer a stand-alone instrument that operates conventionally so I chose the Rigol.

The Signal Hound ran a close second though; a good instrument I think.


Kerry VK2TIL.
8459 2013-04-05 08:17:04 Brooke Clarke Re: Spectrum Analyzer
Hi Kerry:

Most HP equipment has both an input and output for the reference frequency so you can just daisy chain them.
If one instrument has the high stability option use it as the source for the others.
Another option is to use a GPS Disciplined Oscillator as the source:
http://www.prc68.com/I/ThunderBolt.shtml
http://www.prc68.com/I/PRS10.shtml

The NRD-545 receiver can be used as a spectrum analyzer with the addition of a computer and NRD free software.
10 Hz IF bandwidth because of DSP If processing.
http://www.prc68.com/I/NRD545.shtml

The ultimate SA for me is the HP 4395 Network - Spectrum - Impedance Analyzer with true 1 Hz RBW and true RMS detection
option.
Covers audio to 500 MHz.
http://www.prc68.com/I/4395A.shtml

The HP 70000 series SA may be faster than the 4395A and has a higher upper frequency limit.
There's also the possibility of using a tracking generator.
http://www.prc68.com/I/HP71100C.html

Have Fun,

Brooke Clarke, N6GCE
http://www.PRC68.com
http://www.end2partygovernment.com/2012Issues.html
8460 2013-04-05 13:56:27 Kerry Re: Spectrum Analyzer
G'day Brooke.

Yes; the HP 4395 is nice. There is one for sale at Test Equipment Depot for only $12,395. I shudder at the thought of the shipping cost to Australia.

My DSA-815, N2PK VNA and DG8SAQ VNWA will have to do me for my modest needs for the moment. :)

I have four instruments that should be fed from one master oscillator; Rohde & Schwarz SMS, HP8657B, HP5345A and Rigol DSA-815.

Both the 8657 and the 5345 have the HP10544A high-stability oscillator option so it would make sense to use one of those to clock all four instruments.

The one in the 5345 runs constantly; I don't know if the one in the 8657 also runs 24/7.

I will look at the instruments to see if daisy-chaining is possible; if not, a distribution amp will do the job.

Kerry VK2TIL.
8463 2013-04-05 19:50:10 ik0ixi@ik0ixi.it Re: Spectrum Analyzer
HI guys.
Last January I purchased the ATTEN AT5011 model (1 Ghz) with built-in tracking generator.
Simple instrument, low price, good for ham radio but nothing compared to other stuff as old used HP exc.
A bargain if a simple spectrum analisys is you goal.

73 Fabio, IK0IXI - KF1B

Da: emrfd@yahoogroups.com
A: emrfd@yahoogroups.com
Cc:
Data: Tue, 02 Apr 2013 02:58:54 -0000
Oggetto: [emrfd] Re: Spectrum Analyzer





















>  


>

>

>



>
Contact Todd, VE7BPO at his website (http://www.qrp.pops.net/blog.asp.) He has the AT6011 and doesn't think much of it. He suggests the DSA-815 which was reviewed in QST a couple of months back. I am seriously considering the DSA-815. However, it is quite a bit more money - $1495 with the tracking generator option.
>
Ray>
AB7HE
>
8464 2013-04-05 19:50:10 Leonard Re: Spectrum Analyzer
> RBW Range : 30kHz, 100kHz, 300kHz, 1MHz

My problem with this is the 30khz rbw isn't low enough for me. I like to be able to separate out the 2 sidebands and the carrier on a ham rig. Take a look at http://golddredgervideo.com/kc0wox/bitxver3/version3balmod.wmv The modulation frequency was 1khz and the rbw on the HP was set to 300hz. I love my 8552/8553. I have it in a regular 140 main frame as I didn't have the 141 but it does what I want and I even have a spare 140 mainframe.

Leonard
8465 2013-04-05 19:50:10 Ronald RiemVis Re: Spectrum Analyzer
Jerry,

I live in China and maybe can find a way to see that equipment.
If you are serious I can try it

Greetings,

Ronald


On 2 April 2013 10:34, Jerry Haigwood <jerry@w5jh.net> wrote:

> **
>
>
> Hi All,
>
> I have noticed that on the auction site there are some low cost spectrum
> analyzers from China. These all seem to be made by a company named ATTEN.
> Their lowest cost model is the AT5005. It is a 150KHz to 500 MHz model with
> a digital readout for center frequency and marker frequency. The amplitude
> input is from -100 dbm to +13 dbm. The cost is $357 + $114 shipping. Has
> anyone on this list purchased one? Do you know of anyone who has purchased
> one? What do you think of it, etc.?
>
> Jerry W5JH
>
> "building something without experimenting is just solder practice"
>
> [Non-text portions of this message have been removed]
>
>
>


[Non-text portions of this message have been removed]
8467 2013-04-05 19:56:04 Jerry Haigwood Re: Spectrum Analyzer
Hi Ronald,

I think I have enough information. The big killer is the RBW or lack thereof. The smallest bandwidth is 20KHz which is not good.

Jerry W5JH

"building something without experimenting is just solder practice"







8468 2013-04-05 20:01:39 Jerry Haigwood Re: Spectrum Analyzer
The problem I see with the Signal Hound is the sweep width. I believe it is
limited to something fairly small - at least in the upper range. They try
to compensate by giving you a "harmonic function" which helps but does not
cure the problem. If you just want to sweep a large segment of spectrum
looking for some odd mixing product or spur, I don't think it can be done.
I still think the Rigol is a much better instrument.

Jerry W5JH

"building something without experimenting is just solder practice"







8469 2013-04-05 23:48:18 Kerry Re: Spectrum Analyzer
> The problem I see with the Signal Hound is the sweep width.

***************************

As far as I know the Signal Hound can be set to any sweep width.

It seems to sweep much more slowly than the DSA-815; for instance the SH spec for a span of 10 MHz and a RBW of 100 kHz is 270 ms in Fast Sweep mode.

http://www.signalhound.com/old_sh/sweeptime.htm

Just for interest I set my Rigol to the same span & RBW and it sweeps in 10 ms.

There's not much real difference between 270 ms and 10 ms in use but sweep times get much longer as you "home-in" for more detail; if the difference remains in proportion it might be 270 seconds -v- 10 seconds and that is substantial.

I joined the SH Yahoo Group when I was shopping; one question I asked was why the Harmonic mode existed. The answer was to speed-up the sweep time.

I imagine that, in this mode, the SH only samples at or near integer multiples of the fundamental, thus saving a lot of time by not sampling data between the harmonics.

There are plenty of SHs in use and people like them; my decision for the Rigol was just a personal one.

Kerry VK2TIL.
8470 2013-04-06 02:38:47 Roelof Bakker Re: Spectrum Analyzer
Hello Kerry,

I own a Rigol as well and I am very pleased with it.
Here in Europe a used decent HP SA costs about the same and is 20 year old.
No warranty at all whilst, the Rigol comes with a three year warranty.

It is not cheap, but for general HF work it is truly excellent.
You can measure narrow CW filters with it, but there are other ways to do
so.
E.g. PC based FFT analyzers.

Designing high, low or bandpass filters and evaluating the results with
the Rigol is very educating.
If one is looking for true professional standards, be prepared to pay a
professional price!

73,
Roelof Bakker, pa0rdt
8471 2013-04-06 03:25:06 Lasse Moell Re: Spectrum Analyzer
Just want to chime in... as an owner of the Signal Hound SA44B.
The Rigol was not available when I made my decision so it was down to
old HP or new SH :)

One of the neat features of the SH is it is powered through USB and has
a very small foot print! I am amazed on how good the performance is vs.
size. And frequency coveragae is 0-4.4 GHz which may come in handy.

Sweep time is due to how they designed the analyzer, it has a 250 kHz
band with and then step this. So if you need close up sweeps it is
faster, and may be anoyingly slow with high resolution and larger
sweeps. It do some internal band switching around 150 MHz which can add
to sweep time. Looking at sweep times for the Rigol it sure is no speed
deamon either, when going for narrow resolution bandwith.

Still both instruments are neat, and I am sure they both can do
excellent in any ham lab!
Now the Signal Hound has a game changer the BB60A real time spectrum
analyzer, it can record 20 MHz bandwith and has a scanning of 24 GHz/sec
with rbw >10kHz, expensive at 2,5k yes but hopefully this shows what can
be done in near future by others (read Asia)

Still I would love a modern light weight version of the HP 8566/8, same
face but 1/10 in size and weight!

As for the cheaper ATTEN spectrum analyzers they all seems to be nicely
packed but simple in performance. Unlocked VCO and using the sweep
voltage to indicate center freqency on display. Using cheap ceramic
filters to acheive rbw. Like have an expensive W7ZOI analyzer but with
built in CRT and a bit wider in coverage.


On 6 apr 2013 08:48 "Kerry" <planningpower@iprimus.com.au> wrote:

>
>
>
>
> > The problem I see with the Signal Hound is the sweep width.
>
> ***************************
>
> As far as I know the Signal Hound can be set to any sweep width.
>
> It seems to sweep much more slowly than the DSA-815; for instance the
> SH spec for a span of 10 MHz and a RBW of 100 kHz is 270 ms in Fast
> Sweep mode.
>
> <http://www.signalhound.com/old_sh/sweeptime.htm>
>
> Just for interest I set my Rigol to the same span & RBW and it sweeps
> in 10 ms.
>
> There's not much real difference between 270 ms and 10 ms in use but
> sweep times get much longer as you "home-in" for more detail; if the
> difference remains in proportion it might be 270 seconds -v- 10
> seconds and that is substantial.
>
> I joined the SH Yahoo Group when I was shopping; one question I asked
> was why the Harmonic mode existed. The answer was to speed-up the
> sweep time.
>
> I imagine that, in this mode, the SH only samples at or near integer
> multiples of the fundamental, thus saving a lot of time by not
> sampling data between the harmonics.
>
> There are plenty of SHs in use and people like them; my decision for
> the Rigol was just a personal one.
>
> Kerry VK2TIL.
>
>
>
>
>

[Non-text portions of this message have been removed]
8472 2013-04-06 03:29:20 Lasse Moell Re: Spectrum Analyzer
Just want to add.. if you need a very very high performance spectrum
analyzer, up to say 40 MHz get a SDR. My Perseus has an amazing dynamic
range and resolution and speed! Way better than any HP I have access to!
So if I need to measure phase noise or look at close in stuff, SDR is
the way to go!!

/Lasse SM5GLC

>

[Non-text portions of this message have been removed]
8473 2013-04-06 06:33:35 Roelof Bakker Re: Spectrum Analyzer
Hello Lasse,

I can only second what you said about the PERSEUS SDR.
It is more a piece of accurate measurement gear than a receiver.
It can also be used as a (very) selective level meter.

Using a broad band noise source, it can even be used for measuring filter
performance.
So you don't need a tracking generator.

I certainly am biased as I have owned one now for 5 years.
It was bought sheer for curiosity of how direct sampling SDR technology
worked out in practice.

73,
Roelof bakker, pa0rdt
8474 2013-04-06 06:52:04 Lasse Moell Re: Spectrum Analyzer
One thing that is easily forgotten is how poor phase noise these cheap
spectrum analyzers have!
That may be the limiting factor for resolution....

The SDR on the other hand uses a crystal oscillator (often of high
qualtiy too), and this gives an enormous advantage!

One thing I forgot to add about the Signal Hound is that each instrument
is calibrated against NIST tracable standards, and you download your
correction for your specific uint. I find the level measurement of very
high qualitiy.

/Lasse SM5GLC

On 6 apr 2013 15:33 "Roelof Bakker" <roelof@ndb.demon.nl> wrote:

> Hello Lasse,
>
> I can only second what you said about the PERSEUS SDR.
> It is more a piece of accurate measurement gear than a receiver.
> It can also be used as a (very) selective level meter.
>
> Using a broad band noise source, it can even be used for measuring
> filter
> performance.
> So you don't need a tracking generator.
>
> I certainly am biased as I have owned one now for 5 years.
> It was bought sheer for curiosity of how direct sampling SDR
> technology
> worked out in practice.
>
> 73,
> Roelof bakker, pa0rdt
>
>

[Non-text portions of this message have been removed]
8475 2013-04-06 10:16:08 Alberto I2PHD Re: Spectrum Analyzer
8476 2013-04-06 14:32:54 Jerry Haigwood Re: Spectrum Analyzer
Hi Kerry,

I am probably wrong (I usually am) but here is where I got the
information from. According to this person, the RBW suffers when doing a
large scan. Here are the comments:



"I think it really only has a maximum span of 200 kHz. Beyond 200 kHz, it
goes into a special "fast scan" mode that disables the actual I/Q receiver
and uses just an IF power detector. This is good up to 200 MHz (with a 1.5 s
scan time...). As you go past 200 MHz, the RBW becomes very poor (5 MHz) and
the scan time increases more.

I bet the RBW and/or noise floor is poor with a scan range anything beyond
200 kHz.

I didn't see any specs for the broadband sweeps, which makes me wonder what
they are..."



This comment was by KC6F. So, maybe he is wrong? I don't know and I
haven't used a Signal Hound (but I used to play one on TV). ;-)

Jerry W5JH

"building something without experimenting is just solder practice"







8477 2013-04-06 14:42:13 Kerry Re: Spectrum Analyzer
The Perseus looks interesting; I hadn't heard of those before now.

I certainly didn't intend to "knock" the Signal Hound or any other instrument; as I said earlier, the SH ran a close second to the Rigol when I was making my choice.

My decision was influenced by the fact that I already had an N2PK VNA and a DG8SAQ VNWA; with those and the kind of things that I do I don't actually use a spectrum analyser a lot.

If I hadn't had those my decision might (or might not, I don't know) have been different.

Phase noise of the Rigol is certainly higher than the SH or the older HP/Agilents as I mentioned earlier and that was a consideration; I decided that the "good" Rigol spec as distinct from the "excellent" or even "outstanding" spec of the SH or HP) would be fine for me and so it has transpired.

At the end my limited budget meant that I was faced with a decision similar to that described by Lasse; old or new.

I'm not good at repairs, preferring to build things from scratch, and I wanted an instrument to use, not to repair, so "new" won with me; others may be skilled in repair so their decision might be different.

This is an excellent discussion; I've learned a lot from it.

Kerry VK2TIL.
8478 2013-04-06 15:45:20 Lasse Moell Re: Spectrum Analyzer
Jerry,
I do have the hardware (signal hound SA44B)... if you tell me what
settings I should try I would be happy to do so! And I can do screen
captures too if you want me to!

Currently using minimum available RBW:
250 MHz span and 50 kHz RBW with 1.8 sec sweep
100 MHz span and 25 kHz RBW with 842 msec sweep.
I cannot set any less RBW without reducing span...

50 MHz span gives 13 kHz RBW at 515 msec
10 MHz span gives 6.5 kHz RBW at 218 msec
2 MHz span gives 400 Hz RBW at 1.9 sec with noise floor less than -100
dBm with no preamp on.
500 kHz span gives 100 Hz RBW and 640 msec
200 kHz span gives 50 Hz RBW and 280 msec
100 kHz span gives 25 Hz RBW and 422 msec

Not having the Rigol is I would love to see some numbers for an actual
user :) But it seems some of these settings are substansially slower
than the Rigol.

For wide band sweep you can use the 5 MHz RBW and bypass the I/Q
receiver. Span >200 MHz allows for 5 MHz RBW and sweep is 90 ms for 250
MHz span. There is a limitation that start frequency need to be ~170MHz
or lager.

So far the only true limiting factor for me has been the b/w of the I/Q
receiver of 250 kHz, this is no good for wide band signals >250 kHz

Still the Signal Hound SA44B is not state of the art in any way, but for
its size, and low power, a really fine instrument and measurement
receiver!!

What would worry me is the mediocre phase noise on the Rigol, does
anyone have trustworthy numbers??

I found this on EEVBLOG fourm posts:
-----------snip-----------------------

If you set the frequency to go from 0Hz to 150MHz, or 1GHz to 1.15GHz
(or any 150MHz range) the values given are:
RBW=100Hz, Time=1500s (Highest possible -- should be 15000s)
RBW=1KHz, Time=150s
RBW=10KHz, Time=1.5s
RBW=100KHz, Time=15ms
RBW=1MHz, Time=10ms (Lowest possible -- only needs 5ms)

and -87dBc/Hz at 10kHz offset at the lower end of the analyser range.
Another plot shows the phase noise to be about -105dBc/Hz at 100kHz
offset.

-------------end-snip------------------
8479 2013-04-06 16:08:28 Lasse Moell Re: Spectrum Analyzer
The problem for a potential buyer is to get hard facts from real users,
not guestimates from those who never seen the instrument!

I can easiliy see a Rigol 815 Mk2 here in my lab in the future, once
they have got lower phase noise and a few other things fixed :)

/Lasse SM5GLC

On 6 apr 2013 23:42 "Kerry" <planningpower@iprimus.com.au> wrote:

>
>
> The Perseus looks interesting; I hadn't heard of those before now.
>
> I certainly didn't intend to "knock" the Signal Hound or any other
> instrument; as I said earlier, the SH ran a close second to the Rigol
> when I was making my choice.
>
> My decision was influenced by the fact that I already had an N2PK VNA
> and a DG8SAQ VNWA; with those and the kind of things that I do I don't
> actually use a spectrum analyser a lot.
>
> If I hadn't had those my decision might (or might not, I don't know)
> have been different.
>
> Phase noise of the Rigol is certainly higher than the SH or the older
> HP/Agilents as I mentioned earlier and that was a consideration; I
> decided that the "good" Rigol spec as distinct from the "excellent" or
> even "outstanding" spec of the SH or HP) would be fine for me and so
> it has transpired.
>
> At the end my limited budget meant that I was faced with a decision
> similar to that described by Lasse; old or new.
>
> I'm not good at repairs, preferring to build things from scratch, and
> I wanted an instrument to use, not to repair, so "new" won with me;
> others may be skilled in repair so their decision might be different.
>
> This is an excellent discussion; I've learned a lot from it.
>
>

[Non-text portions of this message have been removed]
8480 2013-04-06 17:07:38 Jerry Haigwood Re: Spectrum Analyzer
Hi Lasse,

Well right now I still haven't found a spectrum analyzer that I like and
that I can afford. I'll just have to keep my eyes open. I agree with Kerry
that the Rigol is probably the deal for the features it has.

Jerry W5JH

"building something without experimenting is just solder practice"







8482 2013-04-07 03:18:59 Roelof Bakker Re: Spectrum Analyzer
Hello Lasse,

I have been looking at the sweep time of he Rigol TS815-TG.
The RBW can be selected in a 1-3-10 sequence, so it is not possible to
repeat your settings exactly.
I have taken the nearest RBW.
Note that the minimum RBW for the Rigol is 100 Hz.
The Video Band Width is set to a 1:1 ratio in regard to the RBW, which
gives the same values of course.

Full Span, 0 - 1.5 GHz and 30 kHz RBW with 1.6 sec sweep.

SH > 250 MHz span and 50 kHz RBW with 1.8 sec sweep.
Rigol > 250 MHz span and 30 kHz RBW with 277 msec sweep.

SH > 100 MHz span and 25 kHz RBW with 842 msec sweep.
Rigol > 100 MHz span and 30 kHz RBW woth 111 msec sweep.

SH > 50 MHz span gives 13 kHz RBW at 515 msec.
Rigol > 50 MHz span gives 10 kHz RBW at 500 msec.

SH > 10 MHz span gives 6.5 kHz RBW at 218 msec.
Rigol > 10 MHz span gives 10 kHz RBW at 100 msec.

SH > 500 kHz span gives 100 Hz RBW and 640 msec.
Rigol > 500 kHz span gives 100 Hz RBW and 50 sec.

SH > 200 kHz span gives 50 Hz RBW and 280 msec.
Rigol > 200 kHz span gives 100 Hz RBW and 20 sec.

SH > 100 kHz span gives 25 Hz RBW and 422 msec
Rigol > 100 kHz span gives 100 z RBW and 10 sec.


It is clear that the SH is slower at wider spans and has a restricted Span
for these.
The Rigol is much slower at a low RBW.

However, sweep time is just one thing.
I have made amplifier IMD measurements with the Rigol and for third order
IMD's up to + 40 dBm, I am more than happy to use long sweep times to
resolve weak IMD products.

I have learned not to look for the best possible instrument, but rather
what is good enough for my needs.
After all, it is just a hobby!

73,
Roelof, pa0rdt
8483 2013-04-07 03:45:43 Lasse Moell Re: Spectrum Analyzer
Hi Roelof,
thank you for taking the time doing the measurements... This should aid
people evaluating the instruments.

But as you say there is a lot more to look for. Must say I am impressed
that you manage to evaluate up to 40 dBm IP... I know how hard this can
be :)

It's true of most of us that this is just a hobby and the main thing is
to have fun and learn somthing too!

73
Lasse SM5GLC

On 7 apr 2013 12:18 "Roelof Bakker" <roelof@ndb.demon.nl> wrote:

> Hello Lasse,
>
> I have been looking at the sweep time of he Rigol TS815-TG.
> The RBW can be selected in a 1-3-10 sequence, so it is not possible to
> repeat your settings exactly.
> I have taken the nearest RBW.
> Note that the minimum RBW for the Rigol is 100 Hz.
> The Video Band Width is set to a 1:1 ratio in regard to the RBW, which
> gives the same values of course.
>
> Full Span, 0 - 1.5 GHz and 30 kHz RBW with 1.6 sec sweep.
>
> SH > 250 MHz span and 50 kHz RBW with 1.8 sec sweep.
> Rigol > 250 MHz span and 30 kHz RBW with 277 msec sweep.
>
> SH > 100 MHz span and 25 kHz RBW with 842 msec sweep.
> Rigol > 100 MHz span and 30 kHz RBW woth 111 msec sweep.
>
> SH > 50 MHz span gives 13 kHz RBW at 515 msec.
> Rigol > 50 MHz span gives 10 kHz RBW at 500 msec.
>
> SH > 10 MHz span gives 6.5 kHz RBW at 218 msec.
> Rigol > 10 MHz span gives 10 kHz RBW at 100 msec.
>
> SH > 500 kHz span gives 100 Hz RBW and 640 msec.
> Rigol > 500 kHz span gives 100 Hz RBW and 50 sec.
>
> SH > 200 kHz span gives 50 Hz RBW and 280 msec.
> Rigol > 200 kHz span gives 100 Hz RBW and 20 sec.
>
> SH > 100 kHz span gives 25 Hz RBW and 422 msec
> Rigol > 100 kHz span gives 100 z RBW and 10 sec.
>
>
> It is clear that the SH is slower at wider spans and has a restricted
> Span
> for these.
> The Rigol is much slower at a low RBW.
>
> However, sweep time is just one thing.
> I have made amplifier IMD measurements with the Rigol and for third
> order
> IMD's up to + 40 dBm, I am more than happy to use long sweep times to
> resolve weak IMD products.
>
> I have learned not to look for the best possible instrument, but
> rather
> what is good enough for my needs.
> After all, it is just a hobby!
>
> 73,
> Roelof, pa0rdt
>
>

[Non-text portions of this message have been removed]
8484 2013-04-07 17:38:58 Kerry Re: Spectrum Analyzer
I did similar tests to Roelof's and got similar results; the Rigol, like any swept SA, is slow at narrow RBW/VBW settings.

The SH wins there.

As far as phase noise is concerned, the Rigol is specified at -89 dBc/Hz at 10 kHz.

I have never measured phase noise so I had to do some research; it was a good learning experience. :)

I used a crystal oscillator (my G3UUR test set) to inject a signal into the Rigol; it's assumed that the crystal oscillator has less phase noise than the SA so that the noise displayed is analyser noise.

I measured -87 dBc/Hz at 10 kHz.

The HP 8590 series, which would have been in my ca.$2000 price range, are specified at -90 dBc/Hz at 10 kHz.

The Rohde & Schwarz FSV is specified at -110 dBc/Hz at 10 kHz; it's a mere $30 000 second-hand at TED.

Kerry VK2TIL.
8485 2013-04-07 19:28:26 Ashhar Farhan Re: Spectrum Analyzer
i have been struggling with homebrewing a spectrum analyzer for two years
now, having tried out various configurations. I realized that what i was
really looking for is a computer controlled measurement receiver coupled
with a low noise DDS or Si570 based oscillator.

this is what i have on the bench now (it is a development of the kopski
style sweep generator):

a single conversion measurement receiver with an IF of 40 MHz can be built
easily. 40 MHz fundamental mode crystals are available for cheap, I managed
to built several with a nose of 2 KHz. There is a bit of a ripple, but that
is OK for measurement receiver. I use a high dynamic range mixer like
Chris' KISS mixer, followed by an 10 db attenuator feeds to the IF filter.
A 30 db IF amp with strong performance brings the level up again to feed to
an AD8703 power meter. The local oscillator is an Si570.

The power meter is read by Arduino which also controls the Si570.

The rest of the magic is run on a PC. The sweeps are pretty good at this
point. The total project cost is less than $100 dollars as I solder my own
Arduino controllers, the Ugly way). The scans take about 2 seconds to
complete, the RBW is 2 KHz. I haven't managed to scan anything critical
(just harmonics of an oscillator, etc.).

Those wishing to go all analog could use a Kopski variation sweep generator
for LO as well as sweep generator. Probably, opening up the bandwidth of
the 40 MHz IF filter will make it workable with analog scope's P1
persistence.

At present, I do not use any low pass filtering in the front-end, but I
suppose that some 'weeds' in the grass could be due the IF image.

I'll post the design and the source code as soon as it is stable enough.

- farhan