Spectrum Analyzer - Rigol DSA815

[dpenev]

Hi Orange,

Thank you for your comment!

I have experience with HP4195A which is network/spectrum analyzer and it has log frequency scale.
BTW in my work I found pretty often pure spectrum plots in the the log frequency scale.
For example the EMI certification is based on the log-x scaled spectral masks.

DSA815TG is indeed spectrum analyzer but it has tracking generator which allows scalar network measurements.
I am not sure about RF filters but cable or lumped network filters have roll off looking better in the log freq scale.
I guess it will be the same for the RF filters?
 
In MHO log frequency scale option will be a plus for DSA815TG.
Is this only me thinking like that?

Dimitar

[jahonen]

Log frequency scale is useful if sweeping a wide bandwidth and one wants to maintain constant relative frequency resolution, i.e. see difference of 100 and 500 kHz at same time than 100 and 500 MHz. I use it quite often with my R&S FSV7. For example, a measurement of a SMPS output having some spurious resonance problems:



There one can clearly see the switching fundamental and harmonics, and a severe resonance at about 30 MHz.

Regards,
Janne

[G0HZU]

Hi Guys,

I have already spend 1 day with my DSA8150TG.
Is this only me who miss a logarithmic frequency scale?
I think this is really important, especial for the filter measurements.   

I hope I have just missed to find this option.
Dimitar

Well. This is an RF spectrum analyser. The old school analysers I know don't have a log scale for the frequency. And to be quite honest. If you are adjusting an RF-filter you want to have a linear scale, else you might be fooled about the passband shape.
I don't recall that I have ever seen a log freq scale on an RF analyser.

My Advantest TR4172 is an old school RF spectrum analyser and it has the ability to show a log display on frequency.

It has a tracking gen as standard and can also display phase and group delay. With the optional external bridge it can also display impedance.

It also has various EMC features such as quasi peak detection.  A very versatile instrument but I think it was one of the most expensive RF spectrum analysers in the world about 25 years ago

I think the HP8568B can also have a log frequency display but you have to send it the program code information over GPIB to draw the display on the screen. But it can be done.

[EMC]

I totally agree regards log frequency axis.   Can live with log or dB y axis.   The easiest way I can see would be via remote interface.  i.e. make an application to download the trace and display in a graph that you construct.   I am still looking for the best environment for a hobbiest to do this, anyone made an application for the DSA815TG?

[dpenev]

Hi Gents,

There is a PC based software Ultra Spectrum which is suppose to connect to DSA815.
I was not able to install on the Windows 7 64bit. 
Probably it adds some extra features like the log frequency axis?
Can someone share his opinion about it?

Dimitar

[TomThomas]

Hi Gents,

There is a PC based software Ultra Spectrum which is suppose to connect to DSA815.
I was not able to install on the Windows 7 64bit. 
Probably it adds some extra features like the log frequency axis?
Can someone share his opinion about it?

Dimitar

Hi Dimitar,

to be able to use Ultra Spectrum you need to buy a license. Without license you have 15 days trial time.
Installing under Win7 worked on my Computer fine. Have you installed Ultra Sigma first?
Also take care about regional Settings (dot and comma) It has to be "US style"

br
Tom

[EMC]

Dimitar,

I trialed ultra spectrum for 14 days & reached the conclusion above.  It was stable & operated correctly on my 64 bit windows 7 notebook.    It may have installed correctly on your machine & you are not invoking it.   Thr Rigol instrument software has a 'unique' structure .   First connect the USB cable & allow driver dection time to initialise.  Then invoke Ultra Sigma, Rigol instrument control.   At this point the DSA815 should be there and an *IDN? followed by a read operation will verify communications.   Right click DSA815 if Ultra Spectrum in installed select & invoke.

Go to advanved for more functioality including waterfall, but all I could use it is good for would be an electromagnetic ambient site survey.  No good for EMI testing.

Steve

[fqahmad66]

http://peter.dreisiebner.at/rigol-dsa815-lan/

[EMC]

Peter,

Thanks for that program.   I have to set up a reliable network connection to evaluate  , not done yet.   The progam is very small; what enviroment did you use to create it?

Steve

[Wall-E]

Source:  anonymous

                    Enhanced Specifications for DSA815
                  with 10 Hz Resolution Bandwidth Option

     Resolution             Displayed Average                SSB
     Bandwidth            Noise Level (DANL)           Phase Noise     
    -------------           -----------------------          --------------

 10 Hz to 1 MHz,               -145dBm                   -85 dBc/Hz
1-3-10 sequence                                            @10 kHz offset
   
                                                                               Re.  Option 03  June 2013

[Wall-E]

The DSA815 10 Hz RBW option was initially released to sell in the U.S. for $499, but was discontinued prior to any sales. It is currently planed for incorporation into the up coming DSA830 3 GHz Spectrum Analyzer. 

[TomThomas]

The DSA815 10 Hz RBW option was initially released to sell in the U.S. for $499, but was discontinued prior to any sales. It is currently planed for incorporation into the up coming DSA830 3 GHz Spectrum Analyzer.

Who gave you this kind of information? I'm wondering why Rigol should add another 3GHz Spectrum Analyzer? DSA830 + DSA1030 + DSA1030A for me this step makes no sense.

Rgds
Tom

[Wall-E]

Re: DSA800 10 Hz RBW Option 03 
Source: Rigol anonymous   
Why: Newer technology design with lower cost, reduced weight and warmup time, using DSA800 series firmware base, etc. 

[fqahmad66]

Peter just made his application unavailable..

[dr.diesel]

Peter just made his application unavailable..

Yeah, curious.  He adds a nice screen capture utility though, works great on my Linux box.

[olsenn]

This is only the current statement; if the device is physically capable of operating at a 3GHz bandwidth, and there is a business need/desire to operate the device at a 3GHz bandwidth, the by god the device is gonna operate at a 3GHz bandwidth!

[auato]

Hi there.
I wish to get your attention for a strange behaviour about my Rigol DSA815TG just to know if my instrument have a trouble or I am forgotting something.
Yesterday I tried to check the response curve of a simple 145MHz cavity filter (first picture) by two different ways:

1) by my Rigol DSA815TG scalar network integrated system;
2) by HP8595E spectrum analyzer + Wiltron 6621B programmable sweep generator (two instruments run independently so they are no syncronized; Wiltron generator runs faster at 10ms of sweep and HP8959E has a slow sweep at 20 seconds - see the second picture).









The curves of the same cavity with the same cables are different. Two spectrum analyzers are set in the same mode:
-Center frequency 145MHz
-Span 200MHz
-Ref. level 0dBm
-ampl. scale 10dB/div.


Wiltron:
-start freq. 10MHz to 12.5GHz
-level 0dBm

Of course I have just normalized the DSA815's tracking generator before.
As you can see both show the peak at 150MHz but the Rigol doesn't show the negative peak afterwards (red arrow on the third picture)

What do you think about?

Rgds
auato

[Mr Simpleton]

This is surely a measurement error! 

Try to find the noise floor of the Rigol with no input signal, and then adjust the TG amplitud so the peak is ~50 dB above, you should be able to see the dip.

Reducing noise floor (less RBW) and slower sweep may be a better idea!

[Clint]

Have you tried using the sweep gen and measuring the output with a T into both SA's the RIGOL in normal SA mode ?

[G0HZU]

I think the Rigol defaults to the peak detector algorithm in the DSP so the noise floor will look very high on the Rigol unless you choose a different detector in the menus. Many analysers offer a 'normal' detector, a positive peak detector, negative peak detector and a sample detector.

Most analysers will have a noise figure of around 30dB with 0dB attenuation so the noise floor with a 1MHz RBW and 10dB atten selected should be about -74dBm (i.e. -134dBm/Hz on a 1Hz noise marker) on a typical analyser. The analysers I have here typically show a noise floor a few dB lower than this with the 10dB attenuator selected. The Rigol response looks to be about 20dB worse than this with the peak detector selected but you aren't supposed to use the peak detector to look at noise.

Most analysers would default to the normal detector as an all round compromise so maybe try again with this selected. You could also try the sample detector (best detector for noise) but on many analysers this can cause some subtle types of signal information to be lost if it is close to the noise floor. Also try turning down the RBW a bit as already suggested.

Also, I think you need to select a much narrower span if you want to see the (narrow) cavity response properly and this applies to both analysers.
Try looking at it on a 10MHz span with a 10kHz RBW on both analysers and you should get a better idea of its true insertion loss.

[johnny_canuck]

Auato, try slowing down the '815's sweep time. 20 Seconds (HP) vs 10 milliSeconds (Rigol) isn't exactly a level playfield.

Ken, VE3FIT

[auato]

Thanks all, guys!

Here following some suggested hints:

[Mr Simpleton]

I find it a bit odd that the noise floor with no input is so high 
Should be 20 db lower... still all pics I've seen shows the traces around the middle of the display.

Maybe we could have someone  test and verify that the instrument do meet the specs wrt. noise floor and sensitivity vs. RBW, especillaly with wider filters

[auato]

I find it a bit odd that the noise floor with no input is so high 
Should be 20 db lower... still all pics I've seen shows the traces around the middle of the display.

Maybe we could have someone  test and verify that the instrument do meet the specs wrt. noise floor and sensitivity vs. RBW, especillaly with wider filters

Yes, you are right. Big difference of noise floor between them (-53dBm vs -65dBm).
But for the DSA815 it sounds like normal

[olsenn]

Yes, you are right. Big difference of noise floor between them (-53dBm vs -65dBm).
But for the DSA815 it sounds like normal

Try turning the input attenuation off and enabling the pre-amplifier