Author Topic: DS1054Z CSV export ?bug? [SOLVED]  (Read 11033 times)

0 Members and 1 Guest are viewing this topic.

Offline highwaymanTopic starter

  • Contributor
  • Posts: 24
  • Country: us
DS1054Z CSV export ?bug? [SOLVED]
« on: July 07, 2015, 02:54:55 am »
For the sake of simulating data capture, I placed all 4 inputs to the test point (3V @ 1kHz) on the DS105z (SoftWare Ver:00.04.03).  I captured the output to a CSV file.

I exported the screen shot at 12M [1] (LONG wait) points.
CH1 Min = 2.16 Max = 5.12
CH2 Min = 2.56 Max = 5.60
CH3 7.38E+08 steady for ALL points
CH4 7.38E+08 steady for ALL points

What the!?  Channel 1 & 2 are offset by 2.2-2.6V.  Why?  Channel 3 & 4 have no usable data.  Try again.

I lowered the data rate to 3K and exported the CSV.  The average was:
CH1 7.38E+08  steady for ALL points
CH2  Min = -2.08 Max = 1.04
CH3  Min = -2.32 Max = 0.80
CH4  Min = 5.84 Max = 8.88

Question 1: Does this happen on other DS1054z scopes?
Question 2: Why do I get 7.38E+08 for all points of different inputs?
Question 3: My MAX - MIN is the expected 3 volts.  But why isn't it 0-3 volts?

[1] CORRECTION: Setting was AUTO.  Actual was likely 6M.  Felt like 12M.
« Last Edit: July 09, 2015, 11:21:17 pm by highwayman »
 

Offline alsetalokin4017

  • Super Contributor
  • ***
  • Posts: 2055
  • Country: us
Re: DS1054Z CSV export ?bug?
« Reply #1 on: July 07, 2015, 05:39:13 am »
Could it have something to do with your choice of Trigger? You apparently have your scope set to trigger on the AC Line, not one of the signal channels. This seems very strange to me. The scope isn't displaying a stable set of traces this way. But...

I've tried to duplicate your settings, and when I use 3k points and AC Line trigger source, or trigger on a signal channel, I get values that DO make sense in the CSV file for all channels.

For all 4 channels, I get lows of -8.00e-02 and 0.00e+00  (slight noisy baselines) and highs of mostly 2.88e+00 and 2.96e+00 with some peaks at 3.02e+00.

I didn't try the 12M points. How did you get 12M points with all 4 channels active anyway? I can only get 6M points maximum on mine with all 4 channels active.

So I'm saying I can't seem to duplicate your CSV file errors on my scope. Software version 00.04.03.SP1, "unlocked" to 100MHz bw.



« Last Edit: July 07, 2015, 05:47:24 am by alsetalokin4017 »
The easiest person to fool is yourself. -- Richard Feynman
 

Offline highwaymanTopic starter

  • Contributor
  • Posts: 24
  • Country: us
Re: DS1054Z CSV export ?bug?
« Reply #2 on: July 07, 2015, 09:56:39 am »
Mine is also at 100MHz BW.
Thanks for taking the time to give it a try.  I made a correction to my original post.  The first storage was probably 6M, as you indicated.  The original trigger was set to CH4.  See attached.

I sent a request for the latest firmware (00.04.03.SP1?)
 

Offline Warhawk

  • Frequent Contributor
  • **
  • Posts: 819
  • Country: 00
    • Personal resume
Re: DS1054Z CSV export ?bug?
« Reply #3 on: July 07, 2015, 10:16:24 am »
For the sake of simulating data capture, I placed all 4 inputs to the test point (3V @ 1kHz) on the DS105z (SoftWare Ver:00.04.03).  I captured the output to a CSV file.

I exported the screen shot at 12M [1] (LONG wait) points.
CH1 Min = 2.16 Max = 5.12
CH2 Min = 2.56 Max = 5.60
CH3 7.38E+08 steady for ALL points
CH4 7.38E+08 steady for ALL points

What the!?  Channel 1 & 2 are offset by 2.2-2.6V.  Why?  Channel 3 & 4 have no usable data.  Try again.

I lowered the data rate to 3K and exported the CSV.  The average was:
CH1 7.38E+08  steady for ALL points
CH2  Min = -2.08 Max = 1.04
CH3  Min = -2.32 Max = 0.80
CH4  Min = 5.84 Max = 8.88

Question 1: Does this happen on other DS1054z scopes?
Question 2: Why do I get 7.38E+08 for all points of different inputs?
Question 3: My MAX - MIN is the expected 3 volts.  But why isn't it 0-3 volts?

[1] CORRECTION: Setting was AUTO.  Actual was likely 6M.  Felt like 12M.

I will check it out for you.
Only CSV error I know is a wrong sampling rate in the exported file.
btw: When you are about to convert CSV file to a simulation, you might check out the tiny link in my signature  ;)

« Last Edit: July 07, 2015, 11:42:43 am by Warhawk »
 

Offline Warhawk

  • Frequent Contributor
  • **
  • Posts: 819
  • Country: 00
    • Personal resume
Re: DS1054Z CSV export ?bug?
« Reply #4 on: July 07, 2015, 10:58:15 am »
btw this is the link for the latest firmware:
http://beyondmeasure.rigoltech.com/acton/fs/blocks/showLandingPage/a/1579/p/p-0019/t/page/fm/0

Ignore the description  ;)

Offline Warhawk

  • Frequent Contributor
  • **
  • Posts: 819
  • Country: 00
    • Personal resume
Re: DS1054Z CSV export ?bug?
« Reply #5 on: July 07, 2015, 12:14:02 pm »
Here you go,
1) snapshot from the scope
2) results in the simulation (when sampling rate is modified by hand)
3) scope version

note: I have not hacked it yet, so only a 3M record is used.
« Last Edit: July 07, 2015, 03:23:55 pm by Warhawk »
 

Offline highwaymanTopic starter

  • Contributor
  • Posts: 24
  • Country: us
Re: DS1054Z CSV export ?bug?
« Reply #6 on: July 08, 2015, 01:33:26 am »
Hmmm...A LITTLE progress, but I still can't properly export any CSV longer than 1200 lines.

If I export the screen
STORAGE -> DataSrc ->  Screen
then my data is good (0, 3V), but lacks 'deep memory'.  It terminates at 1200 lines no matter what setting my Mem Depth is.  I am half happy that I can get some good data exported.

If I try to export memory
STORAGE -> DataSrc -> Memory 
I get a longer CSV file but I still get 1-2 channels that are stuck at 7.38E+08.

Initially, I ~assumed~ that memory was all the data points that make up the points on the screen at the Mem Depth resolution.  So I tried to test the opposite.  I recorded some wave forms.

Utility -> Record -> On -> Record
In this mode, the memory was not rolling.  I had hoped that record would record over any older points and continue until 'Stop' was pressed.  Instead, after a number of points, the record stops.
So I export Memory to CSV.  I still get 1-2 channels that are stuck at 7.38E+08.

So...What exactly does Export Memory to CSV do?  Maybe that might be the point that I am missing.
 

Offline Warhawk

  • Frequent Contributor
  • **
  • Posts: 819
  • Country: 00
    • Personal resume
Re: DS1054Z CSV export ?bug?
« Reply #7 on: July 08, 2015, 07:27:34 am »
I assume that export memory does the whole acquisition memory dump. If I were you, I would try two things:
a) restore default settings and try again
b) uninstall options (hack) and try again

Maybe you could tape the whole setup (from factory defaults to exporting the file). Then I (we) could try to reproduce it.


Offline alsetalokin4017

  • Super Contributor
  • ***
  • Posts: 2055
  • Country: us
Re: DS1054Z CSV export ?bug?
« Reply #8 on: July 08, 2015, 12:43:26 pm »
Yes, that's right.  STORAGE -> DataSrc -> Memory  looks like the whole acquisition buffer. When I tried it with 3k points set in Mem Depth, I got a file that was 3002 lines long, with 3000 data points for each channel, and two header lines at the top giving Channel Numbers, Units, and Start and Increment (time). When I selected STORAGE -> DataSrc ->  Screen I got 1202 lines.

However, all my channels contained valid data in both cases. My scope is "fully optioned", firmware 00.04.03.SP1, I started from STORAGE>Default, then set up as in highwayman's first screenshot above.
The easiest person to fool is yourself. -- Richard Feynman
 

Offline alsetalokin4017

  • Super Contributor
  • ***
  • Posts: 2055
  • Country: us
Re: DS1054Z CSV export ?bug?
« Reply #9 on: July 08, 2015, 01:02:38 pm »
@highwayman: The RECORD system works like a stopmotion video of screenshots. You have control over the number of frames and the interval between frames in the record. It's a different system altogether from what you get in the Storage system. Take a look at the Record Opt menu, you'll see where to set the number of frames and the interval between. Ditto for Playback options. The number of frames (screenshots) that can be recorded and played back depends on several things, you may only have enough memory for a few frames if you have all 4 channels and deep Acquire memory set. With 3k points set and 4 channels on, I get a max recording length of 4967 frames. With 6M points set, I only get a max length of 2 frames.
« Last Edit: July 08, 2015, 01:04:38 pm by alsetalokin4017 »
The easiest person to fool is yourself. -- Richard Feynman
 

Offline highwaymanTopic starter

  • Contributor
  • Posts: 24
  • Country: us
Re: DS1054Z CSV export ?bug?
« Reply #10 on: July 09, 2015, 01:13:30 am »
Thank you both for following my instructions above so far.  I reset the scope from Ultra Sigma.  Scope powered up as it did when it first arrived.  I made a video of the power up of the scope to looking at the CSV file with pauses in between.

Part 1-4.

1/4: Set up and capture to CSV (3:47)
https://youtu.be/6kYTvWeJCIs

2/4: CSV saved (0:15)
https://youtu.be/20dC5TTmVjk

3/4: Load CSV on computer (0:34)
https://youtu.be/5Om-CrlkWl8

4/4: Summarize CSV data (0:59)
https://youtu.be/cedSIzeUcsY

Bottom line:  I still get the same error as previous times.  Maybe I am doing something wrong that I missed.  Any input is appreciated, especially if I missed a setting in video 1.  Video 2-4 are summarized below.

CH1 is 7.38E+08 for all
CH2 is 7.38E+08 for all
CH3 ranges from -3 to 0
CH4 ranges from 7 to 10

It is not shown in the video, but CH3 and CH4 alternate between high and low at the same time, so it is follow the trace, just with a random offset.
 

Offline Warhawk

  • Frequent Contributor
  • **
  • Posts: 819
  • Country: 00
    • Personal resume
Re: DS1054Z CSV export ?bug?
« Reply #11 on: July 09, 2015, 07:32:28 am »
Excellent ! I will try to reproduce it this afternoon (your morning)  ;)

Btw: did you try to open the CSV file in notepad ? That excel import... just sayin'  :-X

Offline alsetalokin4017

  • Super Contributor
  • ***
  • Posts: 2055
  • Country: us
Re: DS1054Z CSV export ?bug?
« Reply #12 on: July 09, 2015, 07:51:51 am »
heh.... I would be suspicious of the MS Excel too.... I use Linux/LibreOffice Calc....

But it doesn't look to me like you are doing anything "wrong" on the scope that would cause the corruption of the CSV file.

Have you installed the most recent firmware from the link that Warhawk provided above? The filename of the .gel file and the zip folder name are misleading, but once the firmware is installed the scope should report

Software Version 00.04.03.SP1

in the System Info screen.


If I were you I'd try the following three things:

First, install the new firmware and re-test.  (ETA: Run the Self-Cal procedure after the new install, it may take a couple of tries to complete the self-cal successfully.)
If still fail...
Second, try a different USB stick and re-test.
If still fail...
Third, try a different spreadsheet program (or look at the CSV file in a text editor like NotePad)

If still fail... contact Rigol Tech Support....   :(
« Last Edit: July 09, 2015, 07:55:21 am by alsetalokin4017 »
The easiest person to fool is yourself. -- Richard Feynman
 

Offline Warhawk

  • Frequent Contributor
  • **
  • Posts: 819
  • Country: 00
    • Personal resume
Re: DS1054Z CSV export ?bug?
« Reply #13 on: July 09, 2015, 06:32:48 pm »
I've just checked my scope 1074. I followed your video step-by-step including even horizontal offset for each channel. However my file looks good.
See it in the attachment below. I use MS office2007 and file is not automatically separated to the columns. I must use data import to get the same result as you have but when I do so, it is still OK. No data corruption.


I am going to update my scope to the last firmware now,so I will check it one more time :-)

Update: Nope, even with the newest FW and hacks installed, it works as it should be.
I have a strange feeling regarding MS excel...  :scared:
« Last Edit: July 09, 2015, 08:04:03 pm by Warhawk »
 

Offline highwaymanTopic starter

  • Contributor
  • Posts: 24
  • Country: us
Re: DS1054Z CSV export ?bug?
« Reply #14 on: July 09, 2015, 11:20:39 pm »
SUCCESS!!
 I can't thank everyone enough for their help.  I was certain that the error existed between the o'scope buttons and the closest chair.  Initially, I contacted Rigol directly for the latest firmware.  They informed me that I had the latest.  Well...

Now that I have 04.03.SP1 (it was 04.03), my CSV opens in Excel or Notepad with no problems.   :-+
All four of my signals range between ~0 and ~3.

Thanks again!
 

Offline Warhawk

  • Frequent Contributor
  • **
  • Posts: 819
  • Country: 00
    • Personal resume
Re: DS1054Z CSV export ?bug? [SOLVED]
« Reply #15 on: July 10, 2015, 07:51:39 am »
I am glad you could sort it out.
Do I get it right that the problem was because of the buggy firmware ? When did you get the scope ? Is it possible that you got some development FW ? I have not noticed that 04.03 was out there, only 04.03.SP1.

All's well that ends well.

(and I learnt a new idiom)  :D

Offline highwaymanTopic starter

  • Contributor
  • Posts: 24
  • Country: us
Re: DS1054Z CSV export ?bug? [SOLVED]
« Reply #16 on: July 10, 2015, 10:15:31 am »
That is correct.  I ordered the scope on 6/28/2015 from TEquipment.  If the 04.03 firmware was a development firmware, then it happened at Rigol.  The upgrade to the SP1 firmware (yesterday) was the first time I loaded any oscilloscope firmware.

Hopefully, it was just a bad initial firmware upload.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf