EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems

[jkrichards]

Rigol has created updated firmware for the MSO1000Z and DS1000Z series of scopes in immediate response to the issues found by Dave and the EEVBlog community.

   Firmware has passed the engineering and applications tests and is proceeding to the full suite of testing to go into production. We expect the final firmware for the MSO1000Z and DS1000Z series to be released by December 19th with the MSO2000A, DS2000A, and DS2000 firmware to follow and to be available by toward the end of December...

So Chris, is Rigol still going to release the final firmware for the DS1000Z series today (19 December)?
If not, when do you think there will be a solution available for those of us who are experiencing 70ns or more jitter?
Maybe a chance to at least reload the previous firmware which was not so bad?

I hope to hear back from you on this soon and with some good news.

They are still working on it... The firmware will not be ready till maybe the first half of January    This is what I have been told by Rigol yesterday.  It would be a safe bet to say it won't be ready till February / March.  Hope I'm wrong    Rigol gave us new owners an extension for return for a full refund if this doesn't pan out.

[edavid]

Rigol has created updated firmware for the MSO1000Z and DS1000Z series of scopes in immediate response to the issues found by Dave and the EEVBlog community.

   Firmware has passed the engineering and applications tests and is proceeding to the full suite of testing to go into production. We expect the final firmware for the MSO1000Z and DS1000Z series to be released by December 19th with the MSO2000A, DS2000A, and DS2000 firmware to follow and to be available by toward the end of December...

So Chris, is Rigol still going to release the final firmware for the DS1000Z series today (19 December)?
If not, when do you think there will be a solution available for those of us who are experiencing 70ns or more jitter?
Maybe a chance to at least reload the previous firmware which was not so bad?

I hope to hear back from you on this soon and with some good news.

They are still working on it... The firmware will not be ready till maybe the first half of January    This is what I have been told by Rigol yesterday.  It would be a safe bet to say it won't be ready till February / March.  Hope I'm wrong    Rigol gave us new owners an extension for return for a full refund if this doesn't pan out.

This person in the other thread was told the new firmware is already shipping in new units:  https://www.eevblog.com/forum/testgear/new-rigol-ds1054z-oscilloscope/msg571343/#msg571343

Typical Rigol 

[marmad]

This person in the other thread was told the new firmware is already shibeypping in new units:  https://www.eevblog.com/forum/testgear/new-rigol-ds1054z-oscilloscope/msg571343/#msg571343

Typical Rigol 

The poor communication that exists between Rigol and it's customer base seems to also be prevalent between Rigol and it's various distribution/support arms (USA, EU, etc) as well as it's dealers. But you'd be wrong to think this is restricted to Rigol. From my experience, many (if not all) of the Chinese T&M equipment manufacturers suffer from the same problem to some degree or another ((although Rigol might be the worst). It seems to be tied to long-standing Chinese cultural differences towards business and responsibility.

[kerrsmith]

Unfortunately my scope did not arrive as planned even though the courier emailed me with a definite delivery day (today).  The couriers tracking shows that it is sat in their delivery depot and has not moved all day.

Looks like I will have to wait until Monday for it to arrive (I am really excited about getting it so this will be a long weekend)...

It will also be interesting to see what firmware is on it as well as is sounds like people have been hearing different things from difference people.

[RoTTe]

Hi there!

I take the pyRigolWFM (https://github.com/mabl/pyRigolWFM) and reverse the wfm files from DS1054Z -almost complete-. Right now I have the active channels, Y scale, trigger offset, and etc -most relevant values to represent the wave file on graph o csv with real values-

It's just a WIP, and need a lot of work. But for two evenings it works. I try to document the header file a little when I've some time.

Note: The FFT scale is wrong.
Note 2 -curiosity-: The wfm files can crash the rigol horribly. I just load two modded files to evaluate some values and the Rigol just freeze. Some changes on WFM files affect to the work of the DS1054Z on runtime. You can upload one file that enable channels 1 and 4 but with a sample rate of 500M/samples (1+4 only works at 250M/samples). The DS1054Z works, but all things are wrong. Scales, timming base, measures. It's really funny.

[Lightages]

I received my scope a couple of days ago and have finally gotten to testing the jitter performance. I cannot reproduce any 5 micro second jitter issue. I can reproduce the AC trigger coupling problem but it is very small. I am only able to supply a 2MHz signal at this time.

[matt303]

Thanks for this informative thread  , although slightly frustrating reading for someone that was just about to order one before this issue came to light. The mixed messages coming from different parts of Rigol or their distributors isn't giving me a lot of confidence in claims made about a software fix for the jitter. Looks like I'll be holding off until we have some solid evidence and user feedback that it was a firmware bug or revised hardware is shipping.

[leppie]

Thanks for this informative thread  , although slightly frustrating reading for someone that was just about to order one before this issue came to light. The mixed messages coming from different parts of Rigol or their distributors isn't giving me a lot of confidence in claims made about a software fix for the jitter. Looks like I'll be holding off until we have some solid evidence and user feedback that it was a firmware bug or revised hardware is shipping.

So far on this forum, I only know of 2 (maybe 3) users that the issue got worse with the beta FW. It should not put you off buying this scope. And yes, mine is one of the scopes that got worse, and I still say that. Hopefully, my unit is not faulty on a hardware level.

[RoTTe]

Thanks for this informative thread  , although slightly frustrating reading for someone that was just about to order one before this issue came to light. The mixed messages coming from different parts of Rigol or their distributors isn't giving me a lot of confidence in claims made about a software fix for the jitter. Looks like I'll be holding off until we have some solid evidence and user feedback that it was a firmware bug or revised hardware is shipping.

I've some mixed feelings about this scope, is it really that bad ? I don't know, but the signals we'are looking at with this tool are not exactly state of art. Yes, it's a bug (hardware or software or both) that need addresing at some point. I can list about a dozen programs with worst bugs and people still buying it -hi sony movie!-.

the bad thing: the scope had some things wrong. The good thing: Rigol has acknowledged. If the worst happens, you can always throw fire, sulfur and the scope on the same box to Rigol and ask for a refund or RMA. I just take apart and replace whatever it need.

[matt303]

Given that it looks like it'll be mid January before the next batch hits the UK I can sit back and see how things develop. At the end of the day it's a lot of scope for under £300, probably better than the kit I used when I did electronics as my day job in the 90s.

[poida_pie]

Rigol has created updated firmware for the MSO1000Z and DS1000Z series of scopes in immediate response to the issues found by Dave and the EEVBlog community.

   Firmware has passed the engineering and applications tests and is proceeding to the full suite of testing to go into production. We expect the final firmware for the MSO1000Z and DS1000Z series to be released by December 19th with the MSO2000A, DS2000A, and DS2000 firmware to follow and to be available by toward the end of December.
..

So Chris, is Rigol still going to release the final firmware for the DS1000Z series today (19 December)?
If not, when do you think there will be a solution available for those of us who are experiencing 70ns or more jitter?
Maybe a chance to at least reload the previous firmware which was not so bad?

I hope to hear back from you on this soon and with some good news.

so far I have heard this..
<crickets>

In the meantime I contacted the EU representative for Rigol, asking for the latest firmware.
I saw here in this thread that some people have got scopes with later f/w than the beta.
My reply from the EU rep:
"Dear poida_pie

Normally Update requests from Australia are handled directly from China,

But you already have the latest available firmware installed.
so I took over to tell you that.

Best regards
Thomas Rottach
Application Engineer
Rigol Technologies EU GmbH
"

So, the latest f/w exists
https://www.eevblog.com/forum/testgear/new-rigol-ds1054z-oscilloscope/msg571343/#msg571343
but evidently it's not available for distributors to offer to existing customers.

But not to worry, it's a cultural thing, according to Marmad
https://www.eevblog.com/forum/blog/eevblog-683-rigol-ds1000z-ds2000-oscilloscope-jitter-problems/msg571463/#msg571463
so sit back, relax and shut the fuck up Mr Poida_pie.

By the way, would anyone here consider buying my DS1054Z, and if so, at what discount from new price.
I am wondering just how little it is worth, with it's 70ns jitter at about 35us horizontal delay.

Or time to go animal with it (RoTTe will agree)
Time to read up on the ADF4360-7 and it's loop filter. It's not as if Rigol design guys felt the need to.
First and likely most ignorant guess would be to re-implement the filter with a trim cap set to the existing value of C2. Fiddle with that
while watching the pll lock output on pin 20 (or whatever it was, this is from memory).

[orin]

So new firmware exists in new scopes.  Has anyone opened one up and checked the PLL filter components?  If they changed both filter components and firmware, what makes anyone think that the new firmware would help the problem on older scopes?  Investigation here implies that they 1. got the PLL filter wrong and 2. didn't program the PLL correctly.  A correct fix would appear to require both a hardware and firmware change.  So, as yet, the existence of a new firmware in very recently shipped scopes doesn't really tell us anything.

[Bud]

  A correct fix would appear to require both a hardware and firmware change. 

This is what I think.

[AlessandroAU]

I have come across a 1074 that does not experience the 5us jitter issue when looking at the waveform,

it does however still have surplus frequencies in an FFT.

[Michal_S]

Hi,
If anyone wants to play with it, here's my version of pyRigolWfm for DS1054Z and maybe other scopes from the series, unfortunately it doesn't calculate the voltages correctly (I failed to find all the necessary data in the WFM files), but can be used to dump raw data to CSV files, get some info about the WFM file, plot the data vs time and FFT.
https://github.com/michal-szkutnik/pyRigolWfm1000Z

To plot data vs time plus FFT:

Code: [Select]

wfmutil.py plot -rf filename.wfm
To export raw data to CSV with time in first column:

Code: [Select]

wfmutil.py csv -r input.wfm output.csv
To export raw data to CSV without time (sample values only):

Code: [Select]

wfmutil.py csv -rt input.wfm output.csv

[Solder_Junkie]

My DS1054Z arrived at the end of last week in the UK mid December batch.

The firmware is 4.02/SP3, board version 0.1.1

The 5uS (and multiples of 5), and AC coupled trigger jitter are both very bad.

The good news is this 'scope has a 3dB upper bandwidth of 98 MHz, measured with an RF signal generator. It is straight out of the box, not hacked.

[RoTTe]

Hi,
If anyone wants to play with it, here's my version of pyRigolWfm for DS1054Z and maybe other scopes from the series, unfortunately it doesn't calculate the voltages correctly (I failed to find all the necessary data in the WFM files), but can be used to dump raw data to CSV files, get some info about the WFM file, plot the data vs time and FFT.
https://github.com/michal-szkutnik/pyRigolWfm1000Z

To plot data vs time plus FFT:

Code: [Select]

wfmutil.py plot -rf filename.wfm
To export raw data to CSV with time in first column:

Code: [Select]

wfmutil.py csv -r input.wfm output.csv
To export raw data to CSV without time (sample values only):

Code: [Select]

wfmutil.py csv -rt input.wfm output.csv

Want to compare notes ?  I hate to work in parallel with the same thing. To calculate the voltage values you need the probe attenuation and scale. I saw your code, but the ProbeAttenTimesRange that you uses is the scale Y (V/div) on my notes.

Above I post some captures from my code that export the voltage (without the offset but this is trivial). Also calculate the time scale and total length of wfm file.

Edit: Yep, there is some diferences from your examples and my wfm files.

CH1-1kHz-0OFFSET-1Vdiv.wfm

Code: [Select]

General
=======
Device model             : DS1054Z
Firmware version         : 00.04.02.SP3
Horizontal scale         : 500000000 ps
Trigger delay            : 0 ps
Sampling rate            : 2.000000e+06 samples/s
Enabled channels         : [0] (zero-based indexes)

Channel CH1
===========
Label                    :
Enabled                  : True
Probe attenuation        : 14153201435541.504x
Y grid scale             : 1.000e+00 V/div
Y shift                  : 0.000e+00 V
Y inverted               : False
Sampling rate            : 2.000e+06 samples/s
Time delay               : 0.000e+00 s
No. of recorded samples  : 0


[sergey]

Seems now there are at least few guys working on the waveform formats..

I've tried to look into DS2000 waveform, but achieved less than marmad i guess.. Only had couple of evenings yet tho.

Anyway, RoTTe, Michal_S, is there any chance format of new DS1000z and DS2000 are the same?

[RoTTe]

Seems now there are at least few guys working on the waveform formats..

I've tried to look into DS2000 waveform, but achieved less than marmad i guess.. Only had couple of evenings yet tho.

Anyway, RoTTe, Michal_S, is there any chance format of new DS1000z and DS2000 are the same?

Even among versions (firmware) of DS1054Z some things -apparently- changes. My approach is really simple: take a lot of examples with few changes every time. I'm curious about how an WFM file from one scope works in the same scope after a firmware update. The basic struct is there, but the offset changes. Two wfm files taken with the scope stopped with seconds apart have variations.

I sent a PM to you, if you want to give a try.

[sergey]

Even among versions (firmware) of DS1054Z some things -apparently- changes. My approach is really simple: take a lot of examples with few changes every time. I'm curious about how an WFM file from one scope works in the same scope after a firmware update. The basic struct is there, but the offset changes. Two wfm files taken with the scope stopped with seconds apart have variations.

I sent a PM to you, if you want to give a try.

Geez, didn't think of firmware version would matter. Makes me curious: after upgrading the firmware, would previously saved waveforms loadable?

Yeah, saw the PM. Will send you the files soon.

[poida_pie]

Got an email today:

Dear poida_pie,
This is JiangQin from customer service Dept of RIGOL. I'm glad to help you.
The attachment is the latest firmware for DS1000Z. Please follow the
instructions in the attached document to update the product.

PS:the latest firmware is 00.04.02.03.00


(See attached file: DS1000Z(ARM)update.rar)(See attached file: How to
update DS1000Z.docx)

Best Regards?

??
============================
Application Engineer
Marketing Department
RIGOL Technologies
Tel: +86-10-80706688-821
Mail: jiangqin@rigol.com
Website:www.rigol.com
============================

The instructions read:
How to update DS1000Z
For updating please do following operation.
1. copy .gel file to the root of USB flash
2. Power on DS1000Z
3. Then almost at the same time press "Help" button continue ,until "Single" button be lighted
4. Insert USB flash
5. CH1 button is blinking
6. After a while ,all buttons be lighted
7. Reboot your unit and check the current firmware you will see it changed.
8.Press Utility –>self-cal to perform self-calibration .

Immediately I start wondering... Press "Help" on powerup? That does not do anything at all on the DS1000Z
Many attempts to product the result at step 5 were unsuccessful as you would expect.
The DS1000Z does not have the bootloader like the DS2000 series.
Has anyone here managed to get the bootloader enabled on a DS1000Z in accordance with the instructions
given to me by Rigol?

It seems someone at Rigol is confused.

So anyway I wait for it boot up, trying again since the keyboard locked up, again.
Then insert thumb drive.
 
When I insert the thumb drive, it says "An older version detected. Update? Model DS1000Z, Version 00.04.02.03.00"
Well let's see if it works.
The result is
"Update was failured!
Please check your flash driver and the update file, then try again!"

Two different thumbdrives were tried, both of which have been used to save waveforms in the past.

Another most excellent result. Still stuck at 00.04.02.SP3 with the 70ns jitter and alternate bootup keyboard lockouts.
Has anyone here received the latest f/w and tried it on a SP3 afflicted DS1000Z? What resulted?

 

[Michal_S]

Want to compare notes ?  I hate to work in parallel with the same thing. To calculate the voltage values you need the probe attenuation and scale. I saw your code, but the ProbeAttenTimesRange that you uses is the scale Y (V/div) on my notes.

Hi,
This "ProbeAttenTimesRange" is a name for something I didn't care to find a better name for ;-) It seems that my parser is not valid for SP3 rev. firmware.  I will send a PM to you regarding my findings, some questions, and maybe some WFM samples from my scope.

[seronday]

The instructions read:
How to update DS1000Z
For updating please do following operation.
1. copy .gel file to the root of USB flash
2. Power on DS1000Z
3. Then almost at the same time press "Help" button continue ,until "Single" button be lighted
4. Insert USB flash
5. CH1 button is blinking
6. After a while ,all buttons be lighted
7. Reboot your unit and check the current firmware you will see it changed.
8.Press Utility –>self-cal to perform self-calibration .

There is a Bootloader.
I have used the following process on an early DS1074Z with the original Bootloader.
The only difference is that the USB stick is plugged in before power On.

1. When Powered Off,    plug in USB drive with firmware to be installed.
2. Power ON with HELP button pressed and then press HELP button several times immediately.
3. Most lights flash On and Off and RIGOL splash screen appears as normal.
4. Then CH1 light starts flashing and the file is loaded from USB.
5. The loading time is fairly quick. When finished, the normal start up continues.
6. When start up process is finished, the UpDate Firmware screen appears with a message
that the version on USB is the same as already Installed.
7. At this point you may wish to select OK to install again or else Restart the DSO and check the Firmware Ver.

There appears to be a time out function, such that if no update file is found, the normal start up process continues.

Good Luck.

[EEVblog]

So Rigol have released a final version of the firmware to some people and it's still not fixed?

[poida_pie]

So Rigol have released a final version of the firmware to some people and it's still not fixed?

That is correct.
Version reported is 00.04.02.03.00 for model DS1000Z when I insert the thumb drive.
It does not update successfully. Using 2 different thumb drives proven to have worked when I updated using the SP3 beta.

It's looking to me that I will have to
1- sniff the SPI data going to the ADF4360-7 PLL chip, to determine the needed parameters so as to
2 - design a PLL loop LP filter using the software simulator tool
3 - stick that in the DSO