Is it possible to export data from a digital storage oscilloscope ?

[pfm]

Hi,
I am researching before buying a used DSO but there is one thing that I am not quite clear about DSOs.
If I understand correctly DSOs are able to "store" the sampled data. Is it possible to copy this data(via cable or memory cards) out to lets say a PC where a software program can be used to analyze it ? If so could you please give me some model#s of such scopes ? As I said I am looking for a used scope perhaps in the $300-$500 range or less if possible.

[olsenn]

Most, if not all, digital scopes can export data. I believe from your question you are not referring to a screen capture, but rather the measured values in some sort of spreadsheet. Many means of transporting this data exist, but in its most basic form, I would say you should use the bundled software to store the raw values, either as a text file, excel spreadsheet, or a proprietary format... or you may use VISA (or alternative) commands to capture data in LabView and do all kinds of neat things with it.

[Mechatrommer]

As I said I am looking for a used scope perhaps in the $300-$500 range or less if possible.

dont buy used, buy new! Rigol DS1052E can be had for somewhere more or less than $400. export data capability is "highly confirmed".

[sacherjj]

Both exporting data and doing screen grabs on the Rigol are simple.  I capture to a USB stick.  I haven't connected the scope to a PC directly though.

[pfm]

Thanks for the quick responses!
I was referring to the data, not the screen image. I intend to do FFT spectrum analysis on that data. Never really looked at Rigol. Looks like a few fans here. I will check it out.
A 50Mhz analog bandwidth should be fine but I do want a high sampling rate(which the DS1052 does have) as I'd like to "see" the waveform of some digital signals (spdif) which a typical 24bit 96Khz stream could be well in the order of 9+Mhz.

[wkb]

Here is a sample from a 1052E

[sacherjj]

Thanks for the quick responses!
I was referring to the data, not the screen image. I intend to do FFT spectrum analysis on that data. Never really looked at Rigol. Looks like a few fans here. I will check it out.
A 50Mhz analog bandwidth should be fine but I do want a high sampling rate(which the DS1052 does have) as I'd like to "see" the waveform of some digital signals (spdif) which a typical 24bit 96Khz stream could be well in the order of 9+Mhz.

There is also a simple hack to unlock the 1052 up to 100 MHz.

[olsenn]

I was referring to the data, not the screen image. I intend to do FFT spectrum analysis on that data. Never really looked at Rigol. Looks like a few fans here. I will check it out.

No need to extract data for this; the Rigol scopes have FFT built directly into their DSP processors.  I would go for the DS1102E by the way; it's 100MHz instead of 50MHz and it only costs an extra 50 dollars.

A 50Mhz analog bandwidth should be fine but I do want a high sampling rate(which the DS1052 does have) as I'd like to "see" the waveform of some digital signals (spdif) which a typical 24bit 96Khz stream could be well in the order of 9+Mhz.

24-bits per sample, at 96000 samples persecond yeilds a clock bit rate of 2.3MHz. A general rule of thumb is to take ten times the sample frequency for a square wave to keep it's waveform visible; so even a 25MHz scope will be fine for this, but I would still go with the 100MHz Rigol DSO

[Ajahn Lambda]

A 50Mhz analog bandwidth should be fine but I do want a high sampling rate(which the DS1052 does have) as I'd like to "see" the waveform of some digital signals (spdif) which a typical 24bit 96Khz stream could be well in the order of 9+Mhz.

The DS1052E has on-board FFT functionality, and this model is infamous for its ability to be 'upgraded' to 100MHz via firmware changes (it's relatively simple, don't worry).  Even though it's not exactly as great as a true spectrum analyzer, which is probably the right tool for the job, it can absolutely be useful for seeing the frequency content of signals.  I've used it several times already for finding the noise on some SMPSs, and I just love it.  Almost brought a tear to my eye when I found that, since I'd not even known this device could do FFTs (it was a Christmas gift, and I simply needed a simple, cost-effective 'scope).      


To answer your initial question, regarding the export of data. . . I really do not know, unfortunately.  I still have to piddle around with the I/O setup and the software a bit more - I've controlled it with the PC and captured the screens over USB - but I have to believe there's a way to get that 'raw' signal data for which you're looking.  I'll have to RTFM to find out, though I do know the scope can recall and display a 'reference' waveform  Why I remembered THAT, I don't know. Damn ADD....   

[pfm]

Yes the Rigol does look like a pretty good deal for <$350 from Rigol direct.

Now what about Atten ? Looks like a similar brand ? Any feedback here on Atten scopes ? are they good ?
Any others I could consider ? Owon ?

Olsenn, spdif has its own data structure that adds on more bits, plus two channels for stereo so make that twice. Its definitely 6+Mhz for 96/24.

Sigepblue, detection of precsense of noise on audio signal lines is one of the things I will be doing with the scope. Now the measurement  of the level (voltage) of the noise doesn’t have to be ‘precise’ but I do want the scope to be able to atleast ‘detect’ very small levels of noise i.e. its dynamic range or vertical resolution should be good enough. As long as it shows up on the spectrum with fair accuracy I’d be happy. I realize using the gain/attn setting (mv/div) would come into play and I would desire the gain to be of good quality so as not to add its own noise and distortion to it. Speaking of which what would be the dynamic range of the Rigol ?

[alm]

There have been some complaints about Atten scopes, try using Google to search for other threads on this forum. The responses to the Owon SDS series have been quite positive, but I seem to remember some firmware issues with FFT on the SDS scopes. No idea if this was fixed yet. The Owon PDS scopes are a lot worse. At least one review on the Owon SDS scope was posted on this forum, and there are several threads about them.

Dynamic range is going to be an issue. Theoretical dynamic range would be ~48 dB (8 bits), you would probably be lucky to get 40 dB, since there will usually be at least 1 bit of residual noise. Any DSO will have a limited dynamic range (although some slightly better, especially in averaging mode), this is where a real spectrum analyzer shines.

[Ajahn Lambda]

Sigepblue, detection of precsense of noise on audio signal lines is one of the things I will be doing with the scope. Now the measurement  of the level (voltage) of the noise doesn’t have to be ‘precise’ but I do want the scope to be able to atleast ‘detect’ very small levels of noise i.e. its dynamic range or vertical resolution should be good enough. As long as it shows up on the spectrum with fair accuracy I’d be happy. I realize using the gain/attn setting (mv) would come into play and I would desire the gain to be of good quality so as not to add its own noise and distortion to it. Speaking of which what would be the dynamic range of the Rigol ?

It's fairly decent IME.  In fact, I've used it to service audio (guitar) amplifiers and it seems to work great for that.  Guitar pickups output a signal in the range of about 200mV to 800mV or so; higher than that upon initial plucking (heh), and less when left to decay naturally.  The FFT can have a vertical scale in dBVrms or mVrms, with a minimum scaling of 2mV per division, IIRC.  This should be enough for most audio applications I can think of, and any audible noise in that signal should also be detectable with proper scaling and/or filtering if required.

Be aware that the FFT window chosen can also affect the representation.  From the user manual:

The oscilloscopes provide four FFT windows. Each window is a trade-off between frequency resolution and amplitude accuracy. What you want to measure and your source signals characteristics help determine which window to use.[/size]

You have your choice of four windows:  Rectangle, Hanning, Hamming, and Blackman.  If you download the user manual, that information is on page 2-12.

Check out this application note from Rigol NA regarding FFTs (includes FFT capturing!):
http://na11.salesforce.com/servlet/servlet.FileDownloadPkb?file=00PA0000001Ax4c&orgId=00DA0000000IWhw&pPid=501A0000000DTWI

[Mechatrommer]

check out this thread if you want to get the idea at what level the ds1052e can be used for "remote monitoring" via pc. Other Than FFT Feature Useful for an Oscilloscope

not just fft, you can imagine anything of data processing, but you only get 8bit values out of the dso (minus the noise and noneffective bit, i never measure it accurately). i'm currently building bode plotter which read the data from it. the other things in my mind, application for probe mismatch compensation (phase shift between ch1 and ch2, and magnitude mismatch if feed with same signal). other idea such as post-processed data display (filtered, BW "de-attenuation" recompensation/correction etc). you also can bring fft and lissajou display to a new level, other than the "a bit crappy" built in.

edit: in case you didnt notice, turning fft on in dso will limit maximum sampling rate (ie cannot do 1GSps), doing fft in pc software can avoid this.

[pfm]

The onboard fft with windowing options is great. What is the fft size (1024, ..8192, 16384..) btw ? I don’t see a setting for that and just going by some screenshots I have seen of the fft it seems like the default size might be a bit low ?

I see that data can be saved/exported in a variety of formats – REF, Waveform file, Bitmap file and CSV. Can someone please explain to me how these are different especially in terms of the data they contain ? Particularly if there are differences in terms of the “originality” of the sampled data  - meaning if there is a difference between what the scope actually captured sample-by-sample v/s only a downsampled or downconverted subset of that data.

One more question -
What is the minimum sensitivity of the scope ? in other words what is the smallest signal it can measure with meaningful accuracy ?  2mv/Div times 8 Divs = 8mv full scale and at 8bits thats 31.25uV (per bit) or lets say more practically at 7 bits it would be about 62.5uV. Am I thinking this through correctly ?

Thanks everyone for your inputs on this matter, its helping me build confidence in the choice I am making for the scope, and I truly appreciate it. I just want to make sure I know what I am getting.