New Rigol DS1054Z oscilloscope

[janoc]

Does anyone know how to dump the entire recorded data into CSV file(s)?

One of the options in the Storage menu once you choose CVS format is Source: [Display|Memory], you need to select "Memory."

Saving a 6Mpts record to USB memory stick this way takes forever since you are looking at a 100+MB file and a USB stick write speed of about 150KB/s.


I tried exporting waveform data to CSV so I could import it in OpenOffice:Calc and do a DFT on it for power factor measurement. My first attempt was with 6Mpts but after the horrendously slow export and seeing how slow Calc is at dealing with 1M cells per channel, I decided to drop to 60k - sampling at over 1000xFs is already plenty overkill.

I have done that just this week, trying to do FFT on the data - dumping the entire memory took only a few seconds. On the other hand, I was not recording 6Mpts neither.

You can record the data as "waveform", that gives you a binary .wfm file that I believe was reverse engineered before, so that could be used too in a pinch.

There is also a way to download the memory content by USB, but I haven't played with that yet.

[inline]

So I ordered yesterday afternoon from Tequipment and got it delivered this morning...same State...pretty sweet.  I played with the scope a bit today and all is great, but had a few questions.  I'm coming from a 2 channel scope and noticed that the Rigol DS1054Z has no "External Trigger" input like my previous 2 channel scope.  Am I missing/not understanding why it's not there?

[c4757p]

External trigger is commonly omitted on four-channel scopes. Even the four-channel Tek scopes I've had don't have it. I guess it's seen as something of a hack to get a bit more performance out of an otherwise limited two-channel instrument...

[kwass]

External trigger is commonly omitted on four-channel scopes. Even the four-channel Tek scopes I've had don't have it. I guess it's seen as something of a hack to get a bit more performance out of an otherwise limited two-channel instrument...

Surprisingly Rigol's 1000B low-end, older, 4-channel series scopes have an external trigger input.

http://www.rigolna.com/products/digital-oscilloscopes/ds1000b/

[Alex Eisenhut]

External trigger is commonly omitted on four-channel scopes. Even the four-channel Tek scopes I've had don't have it. I guess it's seen as something of a hack to get a bit more performance out of an otherwise limited two-channel instrument...

Oh yeah? Well my 547 with 1A4 has TWO external trigger inputs! Bahaha!

[janoc]

External trigger is commonly omitted on four-channel scopes. Even the four-channel Tek scopes I've had don't have it. I guess it's seen as something of a hack to get a bit more performance out of an otherwise limited two-channel instrument...

That's actually interesting, because the page 18 of the manual (from: http://www.batronix.com/pdf/Rigol/UserGuide/DS1000Z_UserGuide_EN.pdf) says:

"Chapter 5 To Trigger the Oscilloscope
Introduce the trigger mode, trigger coupling, trigger holdoff, external trigger and various trigger types of the oscilloscope."

Of course, the chapter 5 doesn't describe external trigger! (and the instrument doesn't have one).

There are more such gems in the manual, e.g. the option to switch the SPI/I2C decoding source to "Memory" instead of just screen - the option does actually exists on the instrument, but is permanently greyed out.

On the other hand, if you have 4 channels, it is probably not a too big pain to simply use one of the channels for triggering. On a two channel scope it would have been more limiting.

[kkessler]

Does anyone know how to dump the entire recorded data into CSV file(s)?

One of the options in the Storage menu once you choose CVS format is Source: [Display|Memory], you need to select "Memory."

When I do that, I just get one frame.  For example, I set my memory depth to 12K, and when I record the waveform, I get some 5000 frames, of 12K points.  When I save to a CVS file and select memory, I get 1 CSV file with 12,000 points in it, not a csv file with 12K * 5000 points, or 5000 CSV files of 12K points.

[tautech]

External trigger is commonly omitted on four-channel scopes. Even the four-channel Tek scopes I've had don't have it. I guess it's seen as something of a hack to get a bit more performance out of an otherwise limited two-channel instrument...

The Siglent SDS2000 series 2 & 4 Ch versions have an Ext trigger rear BNC as standard and it is available in menus for various uses. 

[DanielS]

When I do that, I just get one frame.  For example, I set my memory depth to 12K, and when I record the waveform, I get some 5000 frames, of 12K points.  When I save to a CVS file and select memory, I get 1 CSV file with 12,000 points in it, not a csv file with 12K * 5000 points, or 5000 CSV files of 12K points.

I haven't tried messing with segmented memory yet. With the number of little things Rigol seems to have gotten wrong in the DS1000z, I would not be surprised if omitting to enable dumping segmented capture memory (at least without writing an SCPI/VISA app to walk through them and download them individually to a PC) was one of them.

[Mark_O]

Surprisingly Rigol's 1000B low-end, older, 4-channel series scopes have an external trigger input.

Yes, they do.  They also have a lo-rez 320x240 display, and a whopping 8k samples/channel.    And cost 3x as much!

I think losing the ExtTrig input was a worthwhile tradeoff.

[EEVblog]

External trigger is commonly omitted on four-channel scopes. Even the four-channel Tek scopes I've had don't have it. I guess it's seen as something of a hack to get a bit more performance out of an otherwise limited two-channel instrument...

Yes, that's always been the design intention of external trigger. It gets you extra functionality without the cost of a full brown front end.
Many old school 4 channel analog scopes like Tek and Kikusui would have have the extra 2 channel with limited analog front end functionality. e.g. just a two voltage level attenuator. Because front ends are expensive.

[janoc]

When I do that, I just get one frame.  For example, I set my memory depth to 12K, and when I record the waveform, I get some 5000 frames, of 12K points.  When I save to a CVS file and select memory, I get 1 CSV file with 12,000 points in it, not a csv file with 12K * 5000 points, or 5000 CSV files of 12K points.

Which is normal - if you set your memory depth to 12kpts, that means you are recording 12k samples and you get all those 12k samples in the CSV file, i.e. 12k number rows + 2 header lines.

I am not sure where did you get the idea about the 5000 * 12k points. It simply dumps either whatever is on the screen, if you select "Screen" as the source or the entire memory of the scope, if you select "Memory". There is no concept of "frames" there - what are you referring to?

Apropos, the "Sequence On/Off" makes the scope automatically append a number to the file name when you want to save next waveform, so you get "NewFile1.csv", "NewFile2.csv", etc. instead of always generating "NewFile.csv" as the default file name.

[DanielS]

I am not sure where did you get the idea about the 5000 * 12k points. It simply dumps either whatever is on the screen, if you select "Screen" as the source or the entire memory of the scope, if you select "Memory". There is no concept of "frames" there - what are you referring to?

I'm pretty much certain he was thinking about segmented memory recording. At 12k points per frame / segment / trigger event, that would be 500 waveforms to fill the 6Mpts memory per channel though.

I have not tried segmented memory yet but I'm guessing the memory dump function only dumps the currently selected segment.

[accumulator]

I'm finally saying goodbye to my Tek boat anchor - does anyone know if the tequipment discount is still live and how to get one?

[nanofrog]

I'm finally saying goodbye to my Tek boat anchor - does anyone know if the tequipment discount is still live and how to get one?

PM sent.

[janoc]

I'm pretty much certain he was thinking about segmented memory recording. At 12k points per frame / segment / trigger event, that would be 500 waveforms to fill the 6Mpts memory per channel though.

I have not tried segmented memory yet but I'm guessing the memory dump function only dumps the currently selected segment.

Hmm, I am not sure whether it is possible to dump the data recorded using the "Record" function of the scope as a CSV. I haven't played with that yet, though.

[Jidis]

So....after a bit of back-and-forth it seems clear that the DS1054Z indeed has something like 100MHz bandwidth after inputting the codes.

What we really need is somebody who can do a test before/after, preferably with two scopes side by side.

I don't mean to change the subject, but how does the cutoff point for those specs typically work anyhow? I wondered a few months back as I've got an (unhacked) 50MHz 1052E here. I was trying to see how close to their labeled speed some random crystals were here and was somehow able to see the output of an 80MHz one. Is it normal for scopes to go that far past their frequency spec, or is 50 just the point where they should begin to lose something?

Thanks!

[donmr]

I don't mean to change the subject, but how does the cutoff point for those specs typically work anyhow? ...
Is it normal for scopes to go that far past their frequency spec, or is 50 just the point where they should begin to lose something?

Below the Nyquist limit of the sampling, the scope's signal level drops off at some rate determined by the input circuitry.  The standard is to specify the bandwidth as the frequency at which the signal is 3db down (1/2 of the power) from DC.  You can often see signals at higher frequencies, but with reduced magnitude.  And even at the BW freq. the signal is down by 3db so you can't measure the voltage accurately

Don

[Jidis]

Thanks Don!

[Solder_Junkie]

So....after a bit of back-and-forth it seems clear that the DS1054Z indeed has something like 100MHz bandwidth after inputting the codes.

What we really need is somebody who can do a test before/after, preferably with two scopes side by side.

I have a DS1054Z and a commercial grade Marconi RF signal generator that is flat within a dB or so to 512 MHz. Using a through 50 Ohm terminator on an "unhacked" 1054 and 50 Ohm coax, shows a -3dB bandwidth of between 95 and 98 MHz. Opening up the scope to mimic an 1104Z, with "100 MHz" bandwidth, gives a measured -3dB bandwidth in excess of 130 MHz. Individual scopes may vary, but my scope has well over it's specified bandwidth out of the box without hacking.

I forgot to add, the bandwidth varies a little depending on the level fed to the scope, however the differences in bandwidth between say 100 mV RMS and 1 V RMS are small.

The opened scope doesn't stop until you get up to the 400 MHz region, however the 1 GHz sample rate throws a low ball if you try to run more than one channel above 200 MHz.

[tango17]

I'm planning to buy the DS1054Z. Is the tequipment/eev discount still available? Anyone have the code?

Thanks!

[robrenz]

PM SENT

[rgreenfield]

Looking to get one of these Rigol DS1054Z scopes.  Can someone pass along the discount code please?

Thanks!

[Guni]

Search function is very useful or just reading few posts before 
A few posts per page on this thread are about discount code....

[nanofrog]

Search function is very useful or just reading few posts before 
A few posts per page on this thread are about discount code....

PM sent.