Author Topic: New Rigol DS1054Z oscilloscope  (Read 2042423 times)

0 Members and 3 Guests are viewing this topic.

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: New Rigol DS1054Z oscilloscope
« Reply #50 on: September 26, 2014, 10:39:58 am »
On my scope, the Sin(x)/x key is indeed greyed out if less than three channels are turned on, but the text reads "ON". It can only be turned off when three or four channels are in use.

Thanks for clarifying; it's just as I originally proposed (i.e. the opposite).
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: New Rigol DS1054Z oscilloscope
« Reply #51 on: September 26, 2014, 10:42:33 am »
On my scope, the Sin(x)/x key is indeed greyed out if less than three channels are turned on, but the text reads "ON". It can only be turned off when three or four channels are in use.

Can you check what it reads when you have just 1 or 2 channels enabled with a slower sample rate (e.g. something <=250MSa/s)?
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 16560
  • Country: 00
Re: New Rigol DS1054Z oscilloscope
« Reply #52 on: September 26, 2014, 10:48:49 am »
On my scope, the Sin(x)/x key is indeed greyed out if less than three channels are turned on, but the text reads "ON". It can only be turned off when three or four channels are in use.

Ah, ok.... PEBOAC.

Can you check what it reads when you have just 1 or 2 channels enabled with a slower sample rate (e.g. something <=250MSa/s)?

I just went down to 25Msa/s. It never enables the setting.

(And that agrees with what the manual says...)

« Last Edit: September 26, 2014, 10:53:28 am by Fungus »
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: New Rigol DS1054Z oscilloscope
« Reply #53 on: September 26, 2014, 10:51:43 am »
I just went down to 25Msa/s. It never enables the setting.

(And that agrees with what the manual says...)

I don't want to know if it enables the setting - I already know it doesn't. I want to know if it says ON or OFF at sample rates <= 250MSa/s. My theory is that the DSO functions like the DS2000 when running 1 or 2 channels - automatically switching OFF sin(x)/x when the sampling rate is 1/4 or less than the maximum.
« Last Edit: September 26, 2014, 10:53:31 am by marmad »
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 16560
  • Country: 00
Re: New Rigol DS1054Z oscilloscope
« Reply #54 on: September 26, 2014, 10:57:48 am »
I don't want to know if it enables the setting - I already know it doesn't. I want to know if it says ON or OFF at sample rates <= 250MSa/s. My theory is that the DSO functions like the DS2000 when running 1 or 2 channels - automatically switching OFF sin(x)/x when the sampling rate is 1/4 or less than the maximum.

It never says "OFF", always "ON".
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: New Rigol DS1054Z oscilloscope
« Reply #55 on: September 26, 2014, 10:59:27 am »
It never says "OFF", always "ON".

Another bug then.  :)  There is no way sin(x)/x interpolation is running at very slow sample rates. I guess Rigol didn't feel the code necessary to change the display to ON or OFF was necessary if it wasn't enabled.  ;D

The menu name should ACTUALLY be something like "AUTO sin(x)/x".
« Last Edit: September 26, 2014, 11:02:55 am by marmad »
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 16560
  • Country: 00
Re: New Rigol DS1054Z oscilloscope
« Reply #56 on: September 26, 2014, 11:13:55 am »
The menu name should ACTUALLY be something like "AUTO sin(x)/x".

Or maybe the key should display "AUTO" instead of "ON".

 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: New Rigol DS1054Z oscilloscope
« Reply #57 on: September 26, 2014, 11:33:44 am »
It never says "OFF", always "ON".

BTW, if you want to see if/when the DSO is using sin(x)/x versus linear interpolation: set the memory depth to the minimum, set the the time base to make the sampling rate 1/2 or 1/4 of the maximum,  then look at a zoomed signal and check for 'smoothing'. Exit zoom, turn the time base one setting slower, then zoom in again. Keep repeating until the smoothing vanishes and the signal becomes a collection of straight-line vectors.

Here are 2 images of the same signal which show the difference between AUTO sin(x)/x - linear switching on the DS2000.
« Last Edit: September 26, 2014, 11:39:14 am by marmad »
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16544
  • Country: us
  • DavidH
Re: New Rigol DS1054Z oscilloscope
« Reply #58 on: September 26, 2014, 11:59:36 am »
Weird. Mine only lets me switch it on when running 3 or 4 channels.

That is either a bug - or working the opposite of what it's actually displaying. It makes no sense that way.

If the digitizer suffers from excessive amplitude or phase jitter when interleaving is used which happens when only 1 or 2 channels are active, then Rigol might force (x)/x interpolation off because it makes aliasing generated in the digitizer through intermodulation more apparent.  This occurs whether the input signal has components above the Nyquist frequency or not and no anti-aliasing filter can improve the situation.
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: New Rigol DS1054Z oscilloscope
« Reply #59 on: September 26, 2014, 12:15:31 pm »
Weird. Mine only lets me switch it on when running 3 or 4 channels.

That is either a bug - or working the opposite of what it's actually displaying. It makes no sense that way.

If the digitizer suffers from excessive amplitude or phase jitter when interleaving is used which happens when only 1 or 2 channels are active, then Rigol might force (x)/x interpolation off because it makes aliasing generated in the digitizer through intermodulation more apparent.  This occurs whether the input signal has components above the Nyquist frequency or not and no anti-aliasing filter can improve the situation.

As already indicated by subsequent posts, this is not the case - and in fact, Fungus had it backwards, just as I wrote - the DSO only allows you to turn it OFF when 3 or 4 channels are enabled. Besides, IMO, your theory presumes a bad design to begin with - and I doubt there's a modern DSO, inexpensive or otherwise - that does what you propose.
« Last Edit: September 26, 2014, 12:48:22 pm by marmad »
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16544
  • Country: us
  • DavidH
Re: New Rigol DS1054Z oscilloscope
« Reply #60 on: September 26, 2014, 01:16:30 pm »
Weird. Mine only lets me switch it on when running 3 or 4 channels.

That is either a bug - or working the opposite of what it's actually displaying. It makes no sense that way.

If the digitizer suffers from excessive amplitude or phase jitter when interleaving is used which happens when only 1 or 2 channels are active, then Rigol might force (x)/x interpolation off because it makes aliasing generated in the digitizer through intermodulation more apparent.  This occurs whether the input signal has components above the Nyquist frequency or not and no anti-aliasing filter can improve the situation.

As already indicated by subsequent posts, this is not the case - and in fact, Fungus had it backwards, just as I wrote - the DSO only allows you to turn it OFF when 3 or 4 channels are enabled. Besides, IMO, your theory presumes a bad design to begin with - and I doubt there's a modern DSO, inexpensive or otherwise - that does what you propose.

Well, the above discussion about what the Rigol is doing is clear as mud. :)

The old DSOs which I have used that support sin(x)/x interpolation only forced it off when using equivalent time sampling where it is useless.  In all other cases and especially at low sample rate to bandwidth ratios, it was controlled by the user.

Some pretty expensive modern oscilloscopes which use interleaving visibly suffer from this problem to one extent or another.  I have yet to see anybody test for it on Rigol's oscilloscopes or any other inexpensive DSO.

This Agilent application note discusses the interleaving issue:

http://literature.cdn.keysight.com/litweb/pdf/5989-5732EN.pdf

And this one shows on page 5 what happens to a fast transition time signal when this occurs on a real-time DSO although it does not discuss this specific cause:

http://cp.literature.agilent.com/litweb/pdf/5988-8008EN.pdf
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: New Rigol DS1054Z oscilloscope
« Reply #61 on: September 26, 2014, 01:39:59 pm »
Well, the above discussion about what the Rigol is doing is clear as mud. :)

I think rolycat's post makes it crystal clear that sin(x)/x is either ON all the time with 1/2 channels *or* (more likely) is automatically switched ON/OFF based on sample rate (as it's bigger brother, the DS2000 does).

Quote
Some pretty expensive modern oscilloscopes which use interleaving visibly suffer from this problem to one extent or another.  I have yet to see anybody test for it on Rigol's oscilloscopes or any other inexpensive DSO.

I understand the problem with interleaving - but to propose that the interleaving happening inside a single chip - which has been designed specifically to have it's multiple ADC's interleaved - is bad enough to warrant switching off sin(x)/x seems a stretch to me.  :)

In any case - as I showed with my posted images - it's quite simple for any DS1000Z owner to find out when and where the DSO is using sin(x)/x interpolation.
« Last Edit: September 26, 2014, 01:44:09 pm by marmad »
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16544
  • Country: us
  • DavidH
Re: New Rigol DS1054Z oscilloscope
« Reply #62 on: September 26, 2014, 02:39:53 pm »
Well, the above discussion about what the Rigol is doing is clear as mud. :)

I think rolycat's post makes it crystal clear that sin(x)/x is either ON all the time with 1/2 channels *or* (more likely) is automatically switched ON/OFF based on sample rate (as it's bigger brother, the DS2000 does).

This is not the first time that I have found Rigol's documentation to be less than unambiguous.

Quote
Quote
Some pretty expensive modern oscilloscopes which use interleaving visibly suffer from this problem to one extent or another.  I have yet to see anybody test for it on Rigol's oscilloscopes or any other inexpensive DSO.

I understand the problem with interleaving - but to propose that the interleaving happening inside a single chip - which has been designed specifically to have it's multiple ADC's interleaved - is bad enough to warrant switching off sin(x)/x seems a stretch to me.  :)

This is why I wish someone would test this.  From what I remember, the ADC does some type of internal self-calibration to minimize distortion caused by interleaving but it is still at the mercy of its external clock source.

Quote
In any case - as I showed with my posted images - it's quite simple for any DS1000Z owner to find out when and where the DSO is using sin(x)/x interpolation.

It would be easier if the oscilloscope would just report its true state.
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 16560
  • Country: 00
Re: New Rigol DS1054Z oscilloscope
« Reply #63 on: September 26, 2014, 03:55:32 pm »
Well, the above discussion about what the Rigol is doing is clear as mud. :)

Mea culpa.

This is why I wish someone would test this.  From what I remember, the ADC does some type of internal self-calibration to minimize distortion caused by interleaving but it is still at the mercy of its external clock source.

You're saying Rigol don't know how to make a stable clock source?

 

Offline Mark_O

  • Frequent Contributor
  • **
  • Posts: 939
  • Country: us
Re: New Rigol DS1054Z oscilloscope
« Reply #64 on: September 26, 2014, 04:25:24 pm »
Why does the DS1000Z have a switch to turn on and off sin(x)/x? Because it should be switched off when running 3 or 4 channels

Weird. Mine only lets me switch it on when running 3 or 4 channels.

I'm just going to hazard a guess, since I don't have a 1000z, but sinx/x may default On (or always be on) for 1 or 2 channels.  And defaults Off with 3 or 4 enabled (with the option to turn it back on, as Fungus noted, IF you know the input signal characteristics would allow it).
 

Offline rolycatTopic starter

  • Super Contributor
  • ***
  • Posts: 1101
  • Country: gb
Re: New Rigol DS1054Z oscilloscope
« Reply #65 on: September 26, 2014, 04:49:07 pm »
Why does the DS1000Z have a switch to turn on and off sin(x)/x? Because it should be switched off when running 3 or 4 channels

Weird. Mine only lets me switch it on when running 3 or 4 channels.

I'm just going to hazard a guess, since I don't have a 1000z, but sinx/x may default On (or always be on) for 1 or 2 channels.  And defaults Off with 3 or 4 enabled (with the option to turn it back on, as Fungus noted, IF you know the input signal characteristics would allow it).
This question has already been resolved - see my subsequent post.

Fungus, maybe you could edit your posts to avoid misunderstandings by readers skimming the thread?

 

Offline Lightages

  • Supporter
  • ****
  • Posts: 4313
  • Country: ca
  • Canadian po
Re: New Rigol DS1054Z oscilloscope
« Reply #66 on: September 26, 2014, 04:52:37 pm »
Bandwidth aside, is there anything you're lacking with your existing scope that a newer budget DSO could provide (i.e. existing scope is analog, so some DSO features might be nice to have)?

I have a modified DS1052E. What is missing is the update speed, the much bigger memory, and protocol decoding. These could all be helpful to me. I probably will get one. I don't need much more than what this scope offers so a DS2072A is much more money for not much more of what I could use.
 

Offline nanofrog

  • Super Contributor
  • ***
  • Posts: 5446
  • Country: us
Re: New Rigol DS1054Z oscilloscope
« Reply #67 on: September 26, 2014, 05:40:40 pm »
Bandwidth aside, is there anything you're lacking with your existing scope that a newer budget DSO could provide (i.e. existing scope is analog, so some DSO features might be nice to have)?

I have a modified DS1052E. What is missing is the update speed, the much bigger memory, and protocol decoding. These could all be helpful to me. I probably will get one. I don't need much more than what this scope offers so a DS2072A is much more money for not much more of what I could use.
Makes sense.  :)

I'm still only using a CRO (Tek 2445B, so not exactly a hardship  :P), and some of the features of a DSO would be nice (screenshots for example).
 

Offline 0xfede

  • Regular Contributor
  • *
  • Posts: 201
  • Country: it
Re: New Rigol DS1054Z oscilloscope
« Reply #68 on: September 26, 2014, 05:44:27 pm »
Hi all.

This morning a friend came in my lab with a new little toy (hacked DS1054z) and asked me to have some measurement done; since I was busy I made just a couple of test of it.

Measured bandwidth -3db @ 116MHZ with my old (and trusty) Agilent 8657B.
Also attached you will find the best rise and fall time measured.

Still I can't believe that this thing costs only 299€.

Semel in anno licet insanire.
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16544
  • Country: us
  • DavidH
Re: New Rigol DS1054Z oscilloscope
« Reply #69 on: September 26, 2014, 06:28:11 pm »
This is why I wish someone would test this.  From what I remember, the ADC does some type of internal self-calibration to minimize distortion caused by interleaving but it is still at the mercy of its external clock source.

You're saying Rigol don't know how to make a stable clock source?

I am suggesting that Rigol is playing games with the sin(x)/x reconstruction filter because it makes their digitizer linearity or sampling jitter look bad but maybe Occam's Razor or Hanlon's Razor applies.

All someone has to do is measure multiple single shot responses on either fast edges or clean sine waves close below the Nyquist frequency and compare them with and without interleaving which is easy enough to do by enabling the other channels.  If there is a problem with clock source jitter, ADC linearity (presumably made worse by interleaving), or ADC sampling jitter (also presumably made worse with interleaving), it will show up with or without the sin(x)/x reconstruction but it is easier to see with it.
 

Offline leppie

  • Frequent Contributor
  • **
  • Posts: 269
  • Country: za
Re: New Rigol DS1054Z oscilloscope
« Reply #70 on: September 26, 2014, 06:47:47 pm »
Hi all.

This morning a friend came in my lab with a new little toy (hacked DS1054z) and asked me to have some measurement done; since I was busy I made just a couple of test of it.

Measured bandwidth -3db @ 116MHZ with my old (and trusty) Agilent 8657B.
Also attached you will find the best rise and fall time measured.

Still I can't believe that this thing costs only 299€.

That is awesome, thanks  :-+
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: New Rigol DS1054Z oscilloscope
« Reply #71 on: September 26, 2014, 07:21:48 pm »
I am suggesting that Rigol is playing games with the sin(x)/x reconstruction filter because it makes their digitizer linearity or sampling jitter look bad but maybe Occam's Razor or Hanlon's Razor applies.

Honestly, your "suggestion" is based on pure speculation - and, IMO, contradictory to both posted information from users and common sense. To me it seems as if you're just spreading misinformation.

It's already been reported that sin(x)/x is displayed as being automatically ON with 1 or 2 channels enabled - and having manual selection available ONLY when the highest sampling rate is 250MSa/s makes perfect sense - both from the point of view of avoiding possible aliasing in the sin(x)/x interpolation, plus that of porting code from the other UltraVision models.

Instead of continually expounding this theory, perhaps you can post a link to one single document from ANY scope manufacturer that describes turning OFF sin(x)/x at the FASTEST real-time sampling rates to avoid the problem you're imagining.
« Last Edit: September 28, 2014, 06:16:48 pm by marmad »
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 16560
  • Country: 00
Re: New Rigol DS1054Z oscilloscope
« Reply #72 on: September 26, 2014, 08:42:04 pm »
Still I can't believe that this thing costs only 299€.

If Rigol wanted to set the market on fire it looks like they succeeded.

They sell themselves. I just showed mine to somebody and he's like, "I'm gonna get one...".
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16544
  • Country: us
  • DavidH
Re: New Rigol DS1054Z oscilloscope
« Reply #73 on: September 26, 2014, 10:02:29 pm »
I am suggesting that Rigol is playing games with the sin(x)/x reconstruction filter because it makes their digitizer linearity or sampling jitter look bad but maybe Occam's Razor or Hanlon's Razor applies.

Honestly, your "suggestion" is based on pure speculation - and, IMO, contradictory to both posted information from users and common sense. To me it seems as if you're just spreading misinformation.

I do not expect common sense from a company that repeatedly confused peak detection with envelope detection in an apparent attempt to mislead customers into thinking their DSOs had the former when they actually only had the later.  I ran across this years ago when I was evaluating DSOs (before the Rigol Z series was even announced) and it led me to disqualify Rigol.  Ever since then, I have been suspicious of their motives.  Of course this may have been common sense for them from a marketing point of view.

Quote
It's already been reported that sin(x)/x is displayed as being automatically ON with 1 or 2 channels enabled - and having manual selection available ONLY when the highest sampling rate is 250MSa/s makes perfect sense - both from the point of view of avoiding aliasing, plus that of porting code from the other UltraVision models.

It does NOT make sense for it to be disabled (or enabled) to avoid aliasing because sin(x)/x reconstruction neither causes nor increase aliasing.  It merely makes it more apparent.

Quote
Instead of continually expounding this theory, perhaps you can post a link to one single document from ANY scope manufacturer that describes turning OFF sin(x)/x at the FASTEST real-time sampling rates to avoid the problem you're imagining.

I would love to link to a set of screen shots or videos showing if the aliasing problem exists or not in a Rigol DSO but I do not have one to test.  I can show it on other (old) DSOs and in Agilent's application notes but that is not very helpful except to show that the problem exists in a general sense.  Agilent pointed the problem out to distinguish themselves from Tektronix.

No company in their right mind is going to include as a reason for turning off sin(x)/x interpolation that it is to conceal aliasing made worse by interleaving done to increase the real-time sample rate.
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: New Rigol DS1054Z oscilloscope
« Reply #74 on: September 26, 2014, 10:37:16 pm »
I do not expect common sense from a company that repeatedly confused peak detection with envelope detection in an apparent attempt to mislead customers into thinking their DSOs had the former when they actually only had the later.  I ran across this years ago when I was evaluating DSOs (before the Rigol Z series was even announced) and it led me to disqualify Rigol.  Ever since then, I have been suspicious of their motives.  Of course this may have been common sense for them from a marketing point of view.

Your problems with them from the past are meaningless to this discussion, IMO - virtually every single company has made mistakes with some product at some point or another. The point is rather that you started this theory of yours in response to a MISTAKE that was posted in this thread by an owner (Fungus) about the way the DSO dealt with with sin(x)/x.

Quote
It does NOT make sense for it to be disabled (or enabled) to avoid aliasing because sin(x)/x reconstruction neither causes nor increase aliasing.  It merely makes it more apparent.

As has been mentioned before: for sin(x)/x interpolation to be accurate, you have to have an analog input signal that has no frequency content above the Nyquist frequency - which, when 3 or 4 channels are on, is 125MHz. The normal frequency response of the DS1000Z does not roll-off fast enough to minimize aliasing for sin(x)/x interpolation - i.e. LINEAR interpolation should be used  - or-  to put it another way, there exists a good reason for being able to manually keep sin(x)/x turned OFF when 3 or 4 channels are on, if you need to. On the other hand, if you enable the 20MHz bandwidth limiter for each channel that's turned on, you can use sin(x)/x interpolation without problems, regardless of the number of channels on.

Quote
I would love to link to a set of screen shots or videos showing if the aliasing problem exists or not in a Rigol DSO but I do not have one to test.  I can show it on other (old) DSOs and in Agilent's application notes but that is not very helpful except to show that the problem exists in a general sense.  Agilent pointed the problem out to distinguish themselves from Tektronix.

Again, where are these application notes? I want to see a document describing turning off sin(x)/x interpolation because of interleaving problems at the fastest real-time sampling rates. I can link to reams of literature about the problem of aliases in sin(x)/x interpolation, if you like.

Quote
No company in their right mind is going to include as a reason for turning off sin(x)/x interpolation that it is to conceal aliasing made worse by interleaving done to increase the real-time sample rate.

Do you mean that no company will have published literature about this made-up theory of yours?  ;)
« Last Edit: September 26, 2014, 11:24:23 pm by marmad »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf