Author Topic: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing  (Read 39126 times)

0 Members and 1 Guest are viewing this topic.

Offline EEVblog

  • Administrator
  • *****
  • Posts: 30135
  • Country: au
    • EEVblog
EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« on: January 03, 2015, 07:58:20 am »
Dave tries out the new DS1054Z firmware from Rigol that is supposed to fix the 5us delay jitter issue, and the AC Trigger Coupling jitter issues.
Have they fixed it, or is there still a problem?
What caused the 5us jitter issues to begin with, Dave also investigates the ADC PLL based sample clock with an e-Field probe and the Rigol DSA815 spectrum analyser.
Did Rigol also change the PLL coefficients?

ADF4360 PLL datasheet: http://www.analog.com/static/imported-files/data_sheets/ADF4360-7.pdf

 

Offline allikat

  • Contributor
  • Posts: 30
  • Country: gb
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #1 on: January 03, 2015, 08:40:46 am »
Nice to see it fixed Dave.  You're absolutely right, that is a pretty horrible clock waveform.  Maybe they got a few out of spec parts sneak through the QC checks, or someone in management decided to order based on cost not spec.  You never know how these things happen.  Your previous teardown video did rather imply the design was cost focused, and that can mean some pretty tight requirements for parts when you're effectively trying to do things the hard way for cost reasons.
Any engineer can readily identify 3 smells:
1: Coffee, 2: Escaped magic smoke, 3: Bullshit
(from an original post by John Coloccia)
 

Offline SeanB

  • Super Contributor
  • ***
  • Posts: 15169
  • Country: za
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #2 on: January 03, 2015, 08:55:42 am »
Better performance for sure, though I doubt that the original filter is correct for this application. You are right that it likely will need "upgraded" component values, likely done in production as "after serial xxxxxxx". Looks like the PLL can do better, though I guess that that is only really obtainable at specific combinations of input and output frequencies with optimal layout. I would also guess that you would need close tolerance capacitors in that filter as well, and likely they might have been substituted with lower cost ones in production. Might be worth it to get the values in a NPO type and replace the on board versions and see if it helps the jitter, as it being temperature dependant seems to point to the capacitors being cheaper types.
 

Offline rob77

  • Super Contributor
  • ***
  • Posts: 1848
  • Country: sk
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #3 on: January 03, 2015, 09:57:00 am »
it's nice to see that Rigol do care about their customers and they're able to fix issues  :-+  :-+  :-+ planning to buy a 2nd scope, and i think i made the decision now ;) (and i like the looks of 1054Z and the big screen anyways :D )
 

Offline matt303

  • Newbie
  • Posts: 4
  • Country: gb
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #4 on: January 03, 2015, 10:23:20 am »
Great video, nicely showed the PLL issue, does look like them making the best of a bad job but if it works and works reliably it's a fix. Looking forward to your review video now, looks a great scope for the money especially if the waveform update rate is as good as the spec sheet says.
 

Offline Chipguy

  • Supporter
  • ****
  • Posts: 303
  • Country: de
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #5 on: January 03, 2015, 10:42:32 am »
Nicely done!
Where is that smoke coming from?
 

Offline JonR

  • Newbie
  • Posts: 2
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #6 on: January 03, 2015, 11:28:43 am »
Would be interesting to measure the clock on the 2000 series too (and even the old 1054E!).

 

Offline maor

  • Contributor
  • Posts: 28
  • Country: il
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #7 on: January 03, 2015, 02:44:36 pm »
The spectrum after the update is fine, the span around a central frequency generated from a PLL is referred to as phase noise, and the division goes as follows:
0-10Khz from offset : the quality of the of the input frequency.
10-100khz from offset : quality of the loop filter.
100khz-1Mhz from offset : issues with the PFD.
seeing that before the fix, the 100Khz offset was only -6.5 dBc, it would make sense how easily it could effect the signal and cause the jitter, going -37 dBc is more than enough  of a spectral clarity for these applications.
 

Offline nixfu

  • Supporter
  • ****
  • Posts: 344
  • Country: us
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #8 on: January 03, 2015, 02:56:56 pm »
With the amplitude of the extra components down 40db, that is probably good enough to not cause the problem anymore.  I am guessing you could use the old version and cycle through all the multipliers and see at which point it stopped having issues, eg its -6db at 5us, -12db at 15us, etc.. at some point you will see where the clock signals are down low enough to not affect it.  If someone has an unpatched scope, you could try and see at which multiple the jitter goes away, and that would allow you to calculate at what signal level (-db) the extra clock harmonic does not affect the scope anymore even before this fix.

And I am guessing that there is NO way to actually get rid of these components totally unless you add extra bandpass filtering, and that would require more LC filters or something and not just a simple ac coupling cap on the PLL output.

More than likely the scope math works fine as long as the level of those extra clock products are down below a decent threshold.  I mean -40db is 1/10,000th  less signal level than the main peak.  That is pretty damn good actually compared the original output where the first product was only -6db down, which means the first product was fully at 25% of signal level of the peak.

So you have gone from the first product being .25 of the primary to only having the power of 0.0001 of the primary signal. 

That would be considered extremely good oscillator output quality improvement, in the world of RF where we deal with oscillators, and multiplier products all the time.

I say good on Rigol, that is much better results than I would have expected to achieve without physically adding extra physical bandpass filtering.  At those levels, the software should now have zero trouble detecting the primary clock output and rejecting the spurious harmonics.


So, Dave, now is it time to get out that first hour of the review you already have shot?
« Last Edit: January 03, 2015, 04:29:45 pm by nixfu »
 

Offline dmg

  • Regular Contributor
  • *
  • Posts: 101
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #9 on: January 03, 2015, 03:44:24 pm »
Well, they are saved by the fact that the ADC is only 8 bit so the performance degradation due to jitter doesn't affect that much, as the limiting factor for performance is still the quantization noise up to above the required bandwidth of the scope. In any case that clock looks kinda crappy and the PLL seems to have locking issues. It seems that they use a too narrow loop filter for the application, and that means that cutoff frequency for VCO phase noise is too low so it lets most of it go through and prevents correct lock. In the previous firmware the loop was completely unstable and now it's at most borderline stable. For these applications when your reference is good you usually use a relatively wide (10's to 100's of kHz) loop filter because that minimizes overall phase noise and jitter.

Anyway, this issue could be related to Rigol using Analog Devices' driver for the PLL. They offer linux and os-independent drivers for most of their synthesizers but they are pretty buggy and use an approach to calculate divider values that usually gives suboptimal results, at least for ADF4350/51 which is pretty similar to the ADF4360 they use but with integrated multiband VCO. We used them in a project and we had lots of trouble because of lots of annoying bugs, the worst one being that the documentation for device tree configuration parameters was not up to date and new versions of the driver used totally different parameter names, so it wasn't reading them properly so everything was being misconfigured. Even when that was solved we got even more issues from calculations that led to fractional operation when the frequency requested could be synthesized in less noisy integer mode, obsession with adjusting reference divider to 7 which led to a weird PFD frequency and made it fail to calculate values to generate exact 1 GHz from 25 MHz reference and things like those. So the PLL didn't properly lock at certain frequencies or the output wasn't exact and had lots of spurs.

We finally ended up ditching the driver and writing our own based on the code from the desktop application that analog devices gives to interact with the evaluation board, which was written in a much more sensible and failproof way (by different people), and everything worked well. They should look into the issue because it's unacceptable to have a bad clock on a high performance ADC, even when it "works".
 

Offline Chryseus

  • Regular Contributor
  • *
  • Posts: 87
  • Country: gb
    • My Website
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #10 on: January 03, 2015, 04:27:13 pm »
It has not really fixed my jitter issue, although I had almost none before the update, perhaps even a little less it's hard to tell.


Isn't really that bad so I can live with it.
 

Offline rob77

  • Super Contributor
  • ***
  • Posts: 1848
  • Country: sk
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #11 on: January 03, 2015, 07:54:11 pm »
It has not really fixed my jitter issue, although I had almost none before the update, perhaps even a little less it's hard to tell.


Isn't really that bad so I can live with it.

and what would you expect from a $500 scope ? according to the pictures you attached, your scope was and is just fine.
 

Offline pickle9000

  • Super Contributor
  • ***
  • Posts: 2132
  • Country: ca
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #12 on: January 03, 2015, 07:54:37 pm »
That video turned out well. Short, useful, educational.

The update fixed my scope, it did have issues. It did not kill any of the key "hacks".

I bet Rigol will now include some more thorough production line testing. This particular fault should be a thing of the past. Siglent, Hantek are watching as well and probably changing their production testing as well. Although many may say how can this happen I always see it as beneficial. Rigol will certainly have a better product and more reliable scope because of it.

I have no doubt they will change the hardware slightly, that's usual. You can bet that many will be checking the hardware on this scope over and over. This scope is under a microscope (literally) and will be until another one breaks it's hold on it's place in the market.

I was asked by a friend the other day about where this scope belongs on the bench. I replied it's a good "beater" scope. Inexpensive, massive features, good for daily general purpose measurements.   

 
« Last Edit: January 03, 2015, 07:59:01 pm by pickle9000 »
 

Offline Howardlong

  • Super Contributor
  • ***
  • Posts: 4788
  • Country: gb
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #13 on: January 03, 2015, 08:07:43 pm »
The video was shaking about all over the place at the start. Should've done it at 60fps 4k, isn't that what Hollywood uses these days? *








(*=joke)
 

Offline Chryseus

  • Regular Contributor
  • *
  • Posts: 87
  • Country: gb
    • My Website
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #14 on: January 03, 2015, 09:34:02 pm »
and what would you expect from a $500 scope ? according to the pictures you attached, your scope was and is just fine.

The output of the PLL is pretty awful, for $500 I'd certainly expect it to be better than that in a piece of test equipment, while the actual effect on my scope can be pretty much ignored it still doesn't excuse the failure on Rigol's part to properly check the PLL and the associated components.  :palm:
 

Offline nixfu

  • Supporter
  • ****
  • Posts: 344
  • Country: us
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #15 on: January 03, 2015, 09:47:20 pm »

Isn't really that bad so I can live with it.

Are you sure that is jitter and not just persistance? Make sure persistence levels is turned to none.

 

Offline rob77

  • Super Contributor
  • ***
  • Posts: 1848
  • Country: sk
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #16 on: January 03, 2015, 09:58:50 pm »
and what would you expect from a $500 scope ? according to the pictures you attached, your scope was and is just fine.

The output of the PLL is pretty awful, for $500 I'd certainly expect it to be better than that in a piece of test equipment, while the actual effect on my scope can be pretty much ignored it still doesn't excuse the failure on Rigol's part to properly check the PLL and the associated components.  :palm:

my statement was related exactly to your very case, not the overall situation around the clock of Rigol scopes ;) you were complaining about the patch not fixing your jitter issue - while you got jitter (if one could call that jitter) adequate to a $500 scope.
 

Offline Lightages

  • Supporter
  • ****
  • Posts: 4295
  • Country: ca
  • Canadian po
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #17 on: January 03, 2015, 10:35:35 pm »
Are you sure that is jitter and not just persistance? Make sure persistence levels is turned to none.

There is no "none" for persistence on the DS1004Z series. This I hate. There is always persistence, minimum and some intermediate settings, and infinite. If you want to see the raw signal you need to set to infinite.
 

Offline rob77

  • Super Contributor
  • ***
  • Posts: 1848
  • Country: sk
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #18 on: January 04, 2015, 07:34:49 am »
Are you sure that is jitter and not just persistance? Make sure persistence levels is turned to none.

There is no "none" for persistence on the DS1004Z series. This I hate. There is always persistence, minimum and some intermediate settings, and infinite. If you want to see the raw signal you need to set to infinite.

and what would "none" do ? redraw the waveform many times a second with 100% intensity ? that would yield a flickering thick line => no added value in that ;) (just my very opinion)
 

Offline michalK

  • Contributor
  • Posts: 7
  • Country: pl
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #19 on: January 04, 2015, 11:00:59 am »
Yesterday I watched new Dave's video about software upgrade for DS1054Z.
As soon as I saw that it worked with Dave's unit I decided to write a firmware request to Rigol. Today morning I got a reply from Application Engineer from Rigol - do those people never sleep, it's Sunday  ;D

First thing first, so I would like to show you how it worked with my scope.
I had the same problem with it as probably most of its users. The signal jitter on 5us trigger shift and AC trigger coupling.
Below you can see some screenshots BEFORE firmware upgrade. My firmware version was 00.04.01.SP2.

0 us


5 us


AC trigger coupling



AFTER firmware upgrade

0 us


5 us


AC trigger coupling


Also, I would like to note that at first (as suggested by Rigol engineer) I have upgraded the bootloader of my scope, then I upgraded firmware to version 00.04.02 SP3 and finally to 00.04.02 SP4.
I'm not sure if such procedure is necessary (Dave didn't do that on video) but that is what I got from Rigol and I have no reason not to listen to their advices.
At last I can enjoy my DSO as it is written on the screen after successful upgrade precedure  :scared:
« Last Edit: January 04, 2015, 11:03:27 am by michalK »
 

Online Monkeh

  • Super Contributor
  • ***
  • Posts: 6026
  • Country: gb
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #20 on: January 04, 2015, 12:28:30 pm »
and what would you expect from a $500 scope ?

I don't know, what would you expect from a $1200 scope?
 

Offline radiogeek97

  • Regular Contributor
  • *
  • Posts: 95
  • Country: us
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #21 on: January 04, 2015, 12:53:55 pm »
Pickle and others
   Thanks for the post update info.  My scope has all of the hacks installed, would you folks that do have these "updatese" and have done th FW upgrade mind posting your hardware info.  I do have the jitter problem but would be more effected by loosing the "updates" than I would with the jitter problem.   I will post my hardware info when I get to my bench. 
 

Offline DanielS

  • Frequent Contributor
  • **
  • Posts: 798
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #22 on: January 04, 2015, 01:36:22 pm »
and what would you expect from a $500 scope ?
I would expect a PLL that performs close to its reference design capabilities through the application-specific operating range out of any properly designed product at any price point since there is practically no material cost difference between a good and bad PLL circuit based on the same crystal and PLL chip.
 

Offline tggzzz

  • Super Contributor
  • ***
  • Posts: 10301
  • Country: gb
    • Having fun doing more, with less
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #23 on: January 04, 2015, 02:22:03 pm »
and what would you expect from a $500 scope ?

I would expect that they had checked that their phase locked loops were actually locked! If not just stuff an RC oscillaroe in there!

Ditto for a $200 scope.

If their engineering appears to be, um, "slapdash" in one area then it probably is in others too. Motto with software: the number of remaining bugs is proportional to the number of bugs already found.
There are lies, damned lies, statistics - and ADC/DAC specs.
Glider pilot's aphorism: "there is no substitute for span". Retort: "There is a substitute: skill+imagination. But you can buy span".
Having fun doing more, with less
 

Offline Lightages

  • Supporter
  • ****
  • Posts: 4295
  • Country: ca
  • Canadian po
Re: EEVblog #699 - Rigol DS1054Z Oscilloscope Jitter Fix Testing
« Reply #24 on: January 04, 2015, 03:02:42 pm »
There is no "none" for persistence on the DS1004Z series. This I hate. There is always persistence, minimum and some intermediate settings, and infinite. If you want to see the raw signal you need to set to infinite.

and what would "none" do ? redraw the waveform many times a second with 100% intensity ? that would yield a flickering thick line => no added value in that ;) (just my very opinion)

There is sometimes a bit of detail hidden in the displayed signal when there is any persistence. Try it. Look at a waveform and then turn persistence to infinite and you will see some hidden detail. This is only a partial solution as there is garbage on the screen from the infinite persistence. There should be a no persistence option, period. I don't want my scope hiding information from me if I want to see it.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf