Low Cost PCB's Low Cost Components

Author Topic: Rigol DSA 815 900 MHz spur - funny looking thing...  (Read 2264 times)

0 Members and 1 Guest are viewing this topic.

Offline KD0RC

  • Regular Contributor
  • *
  • Posts: 113
  • Country: us
Rigol DSA 815 900 MHz spur - funny looking thing...
« on: March 04, 2014, 02:40:43 PM »
Ever since I got my DSA 815, I have noticed a little discontinuity in the full span trace right at 900 MHz.  I also see it on other folk's YouTube videos, so I am pretty sure it is not some fault unique to me.  Curiosity got the better of me, so I zoomed in on it and it looks funny :-DD.  Well, not that kind of funny...

The first pic is the full span with the little jog at 900 MHz (I did not realize at the time that it is at EXACTLY 900 MHz).

The second pic is zoomed in to a span of 100 Hz, a RBW of 10 Hz and a VBW of 1 Hz.  I also turned on video averaging.  This is a really small spur (I captured this with a 50 ohm load on the input so I am pretyty sure it is internal) so I am not worried about it disturbing anything.  The funny thing about it is the little 6.42 dB jump in amplitude at about 6 Hz from the peak.  The jump is really vertical, so I think it must be as the result of something switching in or out.

Anyone have any ideas as to why this is not just a plain peak?  Could it be at the point where a different filter or oscillator or something kicks in or out?  Is this the kind of thing you would see on any SA, or is this a low end equipment kind of thing?  (I love this thing, but I know it is not a $30,000 unit...)
 

Offline 1design

  • Regular Contributor
  • *
  • Posts: 108
Re: Rigol DSA 815 900 MHz spur - funny looking thing...
« Reply #1 on: March 04, 2014, 06:27:55 PM »
I quess the VCOs switch at that frequency, from the picutres of the inside of this unit I would say that there should be 3 stages.

BR.,
Miha
 

Online georges80

  • Frequent Contributor
  • **
  • Posts: 680
  • Country: us
Re: Rigol DSA 815 900 MHz spur - funny looking thing...
« Reply #2 on: March 05, 2014, 03:43:49 AM »
Just checked mine. Similar shape.
cheers,
george.
« Last Edit: March 05, 2014, 03:45:57 AM by georges80 »
My 'job': www.taskled.com
 

Offline G0HZU

  • Super Contributor
  • ***
  • Posts: 2228
  • Country: gb
Re: Rigol DSA 815 900 MHz spur - funny looking thing...
« Reply #3 on: March 05, 2014, 04:17:59 AM »
Quote
The jump is really vertical, so I think it must be as the result of something switching in or out.
It certainly looks odd...

Your second plot on the narrow span represents a significant issue with respect to the measurement uncertainty of the instrument at 900MHz. It shouldn't look like that on a narrow span because there is a 4 or 5dB step and also a similar step change in the noise floor. So it begs the question "Which side of the step is correct in terms of calibration?"

Obviously something happens at 900MHz. eg some kind of range change or calibration step that affects the display level (post DSP) and the instrument clearly isn't managing this change very well on narrow spans.

Obviously, the RBW filter is done in DSP so it can't create this step in its shape on its own...Something has shifted or sliced the trace level at this point. So it looks like a bug to me coupled along with a 4 dB calibration table issue at 900MHz.

« Last Edit: March 05, 2014, 04:32:39 AM by G0HZU »
 

Online georges80

  • Frequent Contributor
  • **
  • Posts: 680
  • Country: us
Re: Rigol DSA 815 900 MHz spur - funny looking thing...
« Reply #4 on: March 05, 2014, 04:31:21 AM »
Yeah, looks like a software hickup, though to see the blip at all really requires hunting it down, 10hz rbw and 1 hz vbw. Bump the vbw to 10hz and it looks 'smooth' without the discontinuity. Increase the rbw and the whole peak disappears 'into the noise'...

I won't lose much sleep over it :)

cheers,
george.
My 'job': www.taskled.com
 

Offline tom66

  • Super Contributor
  • ***
  • Posts: 3268
  • Country: gb
Re: Rigol DSA 815 900 MHz spur - funny looking thing...
« Reply #5 on: March 05, 2014, 05:22:34 AM »
It's not a common external interference source, like 900MHz GSM?
 

Offline G0HZU

  • Super Contributor
  • ***
  • Posts: 2228
  • Country: gb
Re: Rigol DSA 815 900 MHz spur - funny looking thing...
« Reply #6 on: March 05, 2014, 06:57:37 AM »
It looks like an internal spurious term. At -107dBm or -114dBm it's quite low and insignificant really. I would expect there are quite a few more internal spurious signals in that analyser. I'd be more concerned about the sudden step change in the level and the noise at 900MHz.

I've noticed on these analysers that the general noise level seems very high on the full sweep even with the peak detector option selected and there's quite an upward slope in the noise floor across the range.

On my old HP analysers the noise floor on similar RBW settings but with the default Rosenfell detector selected the noise floor is about -80dBm with 10dB attenuation selected and 1MHz RBW (noise marker shows -140dBm/Hz on sample mode). About 25dB lower? i.e the noise figure of the old HP analyser is about 34dB with 10dB attenuation or 24dB with 0dB attenuation which is fairly typical and nothing special.


« Last Edit: March 05, 2014, 07:22:03 AM by G0HZU »
 

Offline KD0RC

  • Regular Contributor
  • *
  • Posts: 113
  • Country: us
Re: Rigol DSA 815 900 MHz spur - funny looking thing...
« Reply #7 on: March 05, 2014, 01:14:53 PM »
I put a pair of delta markers on the jump and another pair on the noise to get an idea of the magnitude of this thing.  Last night after my initial post, I was measuring 6.42 dB, and I did not think to measure the noise floor higher and lower than the spur.  When I got home from work tonight, I fired it up and made the measurements and got 10 - 11 dB on the jump and 2 - 3 dB on the noise floor.  I did turn the input attenuator down to 0 dB which lowered the overall floor to around -140 dBm but made the delta jump to more than 10 dB.  The pic below is with the attenuator at 0 dB.

Adding 10 dB of attenuation got me back to the 6.4 ish dB number and setting it to max at 30 dBm attenuation changed it to around 5 dB.  It is interesting that the jump is sensitive to the amount of attenuation, but the noise floor delta is more constant at around 2 - 3 dB

I think I forgot to mention that all of this is with the preamp on and the sweep set to 'Accuracy'.

I am with George - I will not lose sleep over this, but it does show that this is not an engineering lab quality instrument!  That being said, in the short time I have had it, this thing has helped me learn quite a bit and for the price, I am very happy with it.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf