Author Topic: EEVblog #1213 - The Oscilloscope Interpolation Trap  (Read 10834 times)

0 Members and 1 Guest are viewing this topic.

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 7563
  • Country: au
Re: EEVblog #1213 - The Oscilloscope Interpolation Trap
« Reply #25 on: May 22, 2019, 08:59:31 am »
In the interpolated signal there is a dead give away that the ringing is due to some kind of digital filtering / interpolation: there was ringing before the transition. Ringing from a probe or poor signal quality has ringing only after the transition.

In the case of testing a limited bandwidth system, like analog TV, the signal I mentioned in my previous posting (a "T"or "2T" "sine squared pulse)* will often show both "pre shoot" & "post shoot", the amplitude of which are primarily related to the bandwidth & Group Delay of the system.

Unfortunately, the artefact at 2:40 in Dave's video looks like a typical result of such testing
Another "trap for old players!"

*In the early days of TV, a rectangular pulse with sharp rise & fall times was used, but massive ringing due to the limited system bandwidth caused difficulties in measurement, so the " sine squared pulse"was devised.
 

Offline wilfred

  • Super Contributor
  • ***
  • Posts: 1247
  • Country: au
Re: EEVblog #1213 - The Oscilloscope Interpolation Trap
« Reply #26 on: May 22, 2019, 09:38:22 am »
Hm, is is just me or does that red sinc drawing look a bit "pubertal"? I mean I would totally expect things like that in a school toilet but I guess some people (who don't even bother to look into the video) could be offended by this on Youtube.

Look at the bright side. Fewer people will fall for the interpolation trap as a result of the increased popularity of the video. I have recently become aware that catchy thumbnails is one way of fertilising the seeds of knowledge.

 

Offline EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 37662
  • Country: au
    • EEVblog
Re: EEVblog #1213 - The Oscilloscope Interpolation Trap
« Reply #27 on: May 22, 2019, 01:29:11 pm »
As Dave said, it's "the 10 year old" in him that saw that.  I didn't see it myself until after watching the video and hearing that comment from Dave.
Personally, if people want to see that, then I think they are looking for it and complaining would be their childish tantrum way of making themselves feel good.  Edit:  I was not referring to you, 0xdeadbeef!
I'd have been a little happier if it hadn't been pointed out - but, hey, Dave's channel ... Dave's call.  I can live with it.

Obviously the 40+ yo in me saw it too.
I put out a public call on twitter the other week for the first person to get a manufacturer to silkscreen a dick'n'balls onto their chip wins The Internet.
The world has gone to politically correct hell, the "easily offended" need to be ridiculed, not pandered to. Dick'n'Balls on everything.
 

Offline schmitt trigger

  • Super Contributor
  • ***
  • Posts: 2205
  • Country: mx
Re: EEVblog #1213 - The Oscilloscope Interpolation Trap
« Reply #28 on: May 22, 2019, 05:17:50 pm »
I've seen Mickey Mouse and other goofy stuff etched on chips, a politically incorrect Viva Cuba Libre, would be a quantum step improvement.
« Last Edit: May 22, 2019, 05:19:26 pm by schmitt trigger »
 

Offline Bud

  • Super Contributor
  • ***
  • Posts: 6877
  • Country: ca
Re: EEVblog #1213 - The Oscilloscope Interpolation Trap
« Reply #29 on: May 22, 2019, 06:56:04 pm »
Dick'n'Balls on everything.

...and IN everything, like firmware  ;)
Facebook-free life and Rigol-free shack.
 

Offline rhodges

  • Frequent Contributor
  • **
  • Posts: 300
  • Country: us
  • Available for embedded projects.
    • My public libraries, code samples, and projects for STM8.
Re: EEVblog #1213 - The Oscilloscope Interpolation Trap
« Reply #30 on: May 22, 2019, 08:22:12 pm »
Look at the Broadcom logo...
Currently developing STM8 and STM32. Past includes 6809, Z80, 8086, PIC, MIPS, PNX1302, and some 8748 and 6805. Check out my public code on github. https://github.com/unfrozen
 
The following users thanked this post: senso

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 7563
  • Country: au
Re: EEVblog #1213 - The Oscilloscope Interpolation Trap
« Reply #31 on: May 23, 2019, 02:32:22 am »
Dick'n'Balls on everything.

...and IN everything, like firmware  ;)

Well they dick around with,the firmware & balls it up! ;D
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16547
  • Country: us
  • DavidH
Re: EEVblog #1213 - The Oscilloscope Interpolation Trap
« Reply #32 on: May 23, 2019, 03:23:28 am »
Equivalent time sampling scopes don't have the same trouble with aliasing provided the signal you are observing is repetitive.

That is true but DSOs that support equivalent time sampling usually do not at slower sweep speeds (1) where the real time sample rate will completely fill the acquisition record in one acquisition and many DSOs back then did not support equivalent time sampling at all.  For a while it was a premium feature.

So detecting possible aliasing by comparing the processed trigger and sample rate would still be useful even on a DSO which supports equivalent time sampling.  It would be useful even with the huge record lengths available today.

Quote
Unfortunately Tek has dropped this feature on its newer scopes in favour of interpolation with all the issues of aliasing. It's a pity because they did it extremely well on the scopes that offered it.

Most manufacturers dropped it and the reason is simple enough; the performance of integrated ADCs and memory reached a point where real time sampling rates were sufficient to support digital triggering at a cost lower than the dedicated hardware needed to measure the trigger to sampling clock delay for equivalent time sampling.

Support for equivalent time sampling is still found in higher end DSOs were the cost of supporting a real time sample rate high enough for increased bandwidth is prohibitive.

(1) DPO style DSOs do not care a whit about this.  They effectively always operate in equivalent time mode.
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 26755
  • Country: nl
    • NCT Developments
Re: EEVblog #1213 - The Oscilloscope Interpolation Trap
« Reply #33 on: January 08, 2020, 06:33:52 pm »
Late to the game but since this video came up in another thread: the sin x/x is broken on the SDS1202X-E which is shown in the video. On any decent scope sin x/x will show a sine wave up to less than 2.5 fs (that is less than 2.5 points per period). And the frequency response on moderns DSOs doesn't roll off Gaussian. The filters are much steeper to avoid aliasing.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16547
  • Country: us
  • DavidH
Re: EEVblog #1213 - The Oscilloscope Interpolation Trap
« Reply #34 on: January 09, 2020, 02:32:38 am »
And the frequency response on moderns DSOs doesn't roll off Gaussian. The filters are much steeper to avoid aliasing.

Is the Rigol DS1000Z series considered modern?  Various reported tests here show that it has something close to a 6dB/octave Gaussian or Bessel response.

https://www.eevblog.com/forum/testgear/rigol-ds1074z-inside-picture/msg337710/#msg337710

I would only expect a steeper roll-off in a much higher bandwidth oscilloscope where a Gaussian or Bessel response is not practical.
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 26755
  • Country: nl
    • NCT Developments
Re: EEVblog #1213 - The Oscilloscope Interpolation Trap
« Reply #35 on: January 09, 2020, 09:43:42 am »
And the frequency response on moderns DSOs doesn't roll off Gaussian. The filters are much steeper to avoid aliasing.

Is the Rigol DS1000Z series considered modern?  Various reported tests here show that it has something close to a 6dB/octave Gaussian or Bessel response.

https://www.eevblog.com/forum/testgear/rigol-ds1074z-inside-picture/msg337710/#msg337710

I would only expect a steeper roll-off in a much higher bandwidth oscilloscope where a Gaussian or Bessel response is not practical.
It is more likely Rigol just didn't care to implement a proper anti-aliasing filter. After all it adds more parts and thus costs.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline thinkfat

  • Supporter
  • ****
  • Posts: 2150
  • Country: de
  • This is just a hobby I spend too much time on.
    • Matthias' Hackerstübchen
Re: EEVblog #1213 - The Oscilloscope Interpolation Trap
« Reply #36 on: January 09, 2020, 09:58:11 am »
It is more likely Rigol just didn't care to implement a proper anti-aliasing filter. After all it adds more parts and thus costs.

And it's not justified. After all, your anti-aliasing filter doesn't need to be perfect, you just need the aliasing products to be drowned in the noise of the instrument. Dynamic range is a factor, too, and and 8 bit DAC doesn't have a lot of that, either.
Everybody likes gadgets. Until they try to make them.
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 26755
  • Country: nl
    • NCT Developments
Re: EEVblog #1213 - The Oscilloscope Interpolation Trap
« Reply #37 on: January 09, 2020, 10:47:59 am »
It is more likely Rigol just didn't care to implement a proper anti-aliasing filter. After all it adds more parts and thus costs.

And it's not justified. After all, your anti-aliasing filter doesn't need to be perfect, you just need the aliasing products to be drowned in the noise of the instrument. Dynamic range is a factor, too, and and 8 bit DAC doesn't have a lot of that, either.
In the frequency domain the roll-off of the (analog) anti-aliasing filter mirrors around Fs/2 so you don't need full attenuation at Fs/2. Still a simple RC filter isn't enough; a 2nd order filter is about the minimum.
Look at these graphs I made on the R&S RTM3004:


and the GW Instek GDS-2204E:

There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf