Author Topic: Rigol DS2072 averaging gets clearer then goes fuzzy.  (Read 396 times)

0 Members and 1 Guest are viewing this topic.

Offline Circlotron

  • Super Contributor
  • ***
  • Posts: 1212
  • Country: au
Rigol DS2072 averaging gets clearer then goes fuzzy.
« on: September 19, 2018, 02:21:54 pm »
Every other DSO I've played with that has an averaging function would show the trace starting off somewhat fuzzy then progressively get sharper and sharper as more and more traces accumulated. With my DS2072 though, while getting sharper it gets to a point where something seems to overflow and it goes back to being less sharp. Still sharper than when it began but not as sharp as it got to before it rolled over or whatever. Always been like this. See video.

 

Offline tautech

  • Super Contributor
  • ***
  • Posts: 12808
  • Country: nz
  • NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: Rigol DS2072 averaging gets clearer then goes fuzzy.
« Reply #1 on: September 19, 2018, 02:26:44 pm »
A = fundamental frequency jitter. More averages needed.
Avid Rabid Hobbyist
 

Offline Circlotron

  • Super Contributor
  • ***
  • Posts: 1212
  • Country: au
Re: Rigol DS2072 averaging gets clearer then goes fuzzy.
« Reply #2 on: September 19, 2018, 02:49:52 pm »
Video was 128 averages. Tried it again, this time 8192 averages. Took what felt like about 10 minutes but basically the same result. It's almost as if during the accumulating of traces the dither frequency between an upper and lower pixel is relatively low, giving just a single non-returning step during a sloping part of the waveform. When the average count has finished the dithering rate seems to increase greatly, giving the appearance of a blurry trace where it gradually progresses in a sort of PWM fashion from one horizontal row of pixels to the next one down. It might be technically correct but I don't like it!
« Last Edit: September 19, 2018, 02:54:21 pm by Circlotron »
 

Offline tautech

  • Super Contributor
  • ***
  • Posts: 12808
  • Country: nz
  • NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: Rigol DS2072 averaging gets clearer then goes fuzzy.
« Reply #3 on: September 19, 2018, 03:04:01 pm »
Video was 128 averages. Tried it again, this time 8192 averages. Took what felt like about 10 minutes but basically the same result. It's almost as if during the accumulating of traces the dither frequency between an upper and lower pixel is relatively low, giving just a single non-returning step during a sloping part of the waveform. When the average count has finished the dithering rate seems to increase greatly, giving the appearance of a blurry trace where it gradually progresses in a sort of PWM fashion from one horizontal row of pixels to the next one down. It might be technically correct but I don't like it!
Is this your first DSO ?
Looks like normal behavior to me.
Won't make any differences to usage or graticule measurement.
Avid Rabid Hobbyist
 

Offline Circlotron

  • Super Contributor
  • ***
  • Posts: 1212
  • Country: au
Re: Rigol DS2072 averaging gets clearer then goes fuzzy.
« Reply #4 on: September 19, 2018, 03:10:56 pm »
Is this your first DSO ?
Looks like normal behavior to me.
Yep, but I've had it for 5 years and used it fairly often so I'm not totally unfamiliar with DSOs, just not as familiar with them as some other people here.
 

Offline Circlotron

  • Super Contributor
  • ***
  • Posts: 1212
  • Country: au
Re: Rigol DS2072 averaging gets clearer then goes fuzzy.
« Reply #5 on: September 19, 2018, 03:20:40 pm »
Bumped the frequency up to 10kHz so the 8192 samples would be quicker. Pics show trace just before it goes "bad" and about 10 seconds after.

 
 

Offline Circlotron

  • Super Contributor
  • ***
  • Posts: 1212
  • Country: au
Re: Rigol DS2072 averaging gets clearer then goes fuzzy.
« Reply #6 on: September 20, 2018, 08:13:21 am »
Okay.... after sleeping on it I think I know what is going on.
In games there is often diagonal lines and when displayed on a monitor with a finite amount of pixels this line, depending on the angle, will have something of a staircase appearance. To reduce the visibility of that artefact the software (or maybe the video card driver?) applies antialiasing to the line to blur the stepped transitions from one line of pixels to the next. I’d say that is exactly what is happening here. When the total amount of traces has been accumulated and averaged the scope in its wisdom decides it will smudge and blur the line in an effort to make it look pretty. That might be useful if the screen had ten times the resolution that it does now, but especially if the vertical resolution is only eight bits then this antialiasing just makes a mess of things AFAIAC.
 

Online Bud

  • Super Contributor
  • ***
  • Posts: 2681
  • Country: ca
Re: Rigol DS2072 averaging gets clearer then goes fuzzy.
« Reply #7 on: September 20, 2018, 09:22:54 am »
You are of too high opinion of the Chinese. This may be as simple as some small vertical offset drift. Try slightly change the vertical offset using Y control while the scope is averaging and see if that produces similar result
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 8580
  • Country: 00
Re: Rigol DS2072 averaging gets clearer then goes fuzzy.
« Reply #8 on: September 20, 2018, 12:25:45 pm »
There's nothing "bad" about it, it's just the signal landing in between two pixels.

Try using the knob to move the trace up by half a pixel.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 7591
  • Country: us
  • DavidH
Re: Rigol DS2072 averaging gets clearer then goes fuzzy.
« Reply #9 on: September 21, 2018, 07:17:43 am »
Looks like normal behavior to me.

It is normal behavior if the processing record is only 8 bits and the display does not have higher resolution or antialiasing.  The Rigol DS1000 series works this way and it looks like the DS2000 series does also.

Quote
Won't make any differences to usage or graticule measurement.

It can make a difference.  A user might expect that averaging 8 bit acquisitions will produce a higher resolution measurement but this is not the case if the result is still 8 bits and only noise was removed.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf