Siglent oscilloscopes. Is there any possibility of long time recording?

[Benbow]

Hello everyone!

I am interested in whether Siglent oscilloscopes have such useful option as long time recording, for example  SDS1000... serie. Do they have any possibility of long time recording their own waveforms? For subsequent analysis of it.
Just in case I installed their EasyScopeX software to look, but I have not found any signs of recording possibility at all. There was only snapshot possibility found.

Probably someone has more information or experience.

thanks in advance.

[Benbow]

I have already quite a few expensive the same electronic equipments having sporadic issues. Perhaps the same elements become out of order sporadically. And it is not soldering issue.
Of course, it is wide known that it is very difficult job to figure out sporadically non working element. There should be confidence that really sporadically non working element was replaced and issue will never occur again.
Even successful repair three of them defrays oscilloscope purchasing. Such as Rigol or Siglent series in low price range <$500.
I need just 24 hour real time recording. I need to see signal alteration within 24 hour at least, in some places.
I know that Textronix's software has that useful option. But they have not reasonable prices.
It is not affordable luxury.
 

[colorado.rob]

24 hours of recording at what sample rate?  Are you wanting to capture specific events over a 24 hour period, or do you really need to capture 24 hours of continuous data?  How do you intend to use/analyze the data?

[nctnico]

The GW Instek GDS-1000B series has a data log function which can record up to 100 hours using consequtive acquisitions which can be saved to a USB stick or a Windows network share as an image or raw waveform data (the raw waveform data can be read back into a trace). The GDS1054B fits your budget.

[boggis the cat]

I am interested in whether Siglent oscilloscopes have such useful option as long time recording, for example  SDS1000... serie. Do they have any possibility of long time recording their own waveforms? For subsequent analysis of it.

The SDS 1202X-E does not appear to have such a capability.  Possibly higher priced models may do, but I am not aware of this capability being present.  The Rigols don't have it 'natively', either, but possibly you could use the various software options out there to create a logging script.  You could do this for any 'scope, in principle, so it depends on whether you can find something already available or are willing to write your own solution.  (There may be caveats that could cause problems, such as the Rigol 1054Z having some remote interface bugs.)

A lot of Japanese 'scopes seem to have a datalogging type capability built in, but they can be difficult to find (and are not cheap).

As nctnico has pointed out, the GW Instek GDS 1054B is worth a look as it has this feature already available and is relatively inexpensive (less than the Rigol 1054Z in the US).  If you aren't keen on spending a fair amount of effort building your own solution then I would look at this 'scope first.

[DaJMasta]

What kind of capture rate are you looking at?  If it's extremely low, you may be able to keep capturing chunks through the USB using custom software.  Otherwise, you may be able to capture a full-sample block then program something to just do that as often as possible - you will miss significant parts for the transfer time, but if you're looking at long term changes in a steady signal, it could still be valuable.

[rf-loop]

The GW Instek GDS-1000B series has a data log function which can record up to 100 hours using consequtive acquisitions which can be saved to a USB stick or a Windows network share as an image or raw waveform data (the raw waveform data can be read back into a trace). The GDS1054B fits your budget.

Time gap between these "consequtive" acquisitions in this case, and samplerate?

Example with Siglent, using waveform history buffer:
2s/div, memset 14k, 2 channel in use (example for battery charging temp and voltage). 250ks/s.
Gap between acquisitions around 1s and every acquisition 28s. In wfm history buffer with this settings max 1887 acquistions. every single acq. time stamped.
With lower samplerate up to around 1 week "consequtive" acquisitions.
With low timebases,  gap between acquisitions is roughly one horizontal div. (14div acquisition 1div time gap and then next acquisition until wfm buffer is full and after then like FIFO)

If need long continuous acquisition stream, it can not do over 1400 second.

Later I will try make table for slow timebases and different acquisition length settings. (for normal wfm history buffer and also for segmented memory acquisition.

But then. If people know what there must NOT exists then it is perhaps more clever to set scope for trig or detect himself this unwanted situation and go to home and let scope run until it happened and then next working day look what it have found. One can use simple hammer for many many different things and one can use it only for one thing.... it depends, skills, experience and education.

[nctnico]

The difference between data logging mode and letting a scope run is that you still have the data in case when power gets interrupted. The gaps are relatively short versus the total time (the GW Instek is very fast when it comes to saving data). In an ideal world you want to trigger on events but if you don't know the problem then that is nearly impossible. When enabling peak-detect you can see there where high frequency peaks even at a low samplerate.

[rf-loop]

The gaps are relatively short versus the total time...

Relatively short? I'm so old that one year is relatively short.

Single acquisition length? Sample rate? Time gap between stored acquisitions? Of course these acquisitions have time stamps so you can easy tell real data instead of marketing phrases? But, with Siglent there is real numbers - data, as you know -  and not only marketing phrases.

[Benbow]

24 hours of recording at what sample rate?  Are you wanting to capture specific events over a 24 hour period, or do you really need to capture 24 hours of continuous data?  How do you intend to use/analyze the data?

Yesterday I was looking for more info on data logging using oscilloscopes.
this is piece of article (below) from Textronix web site.

Ways to do each type of data logging

    Seamless data logging cannot generally be done on oscilloscopes, because oscilloscopes are inherently triggered devices. If you require seamless data logging, you will need to look at either an analog data logger, or a digitizer instead of an oscilloscope.
    Triggered and measurement data logging can be done with an oscilloscope, in conjunction with software that will save the data to a computer. Wavestar, the TekXL toolbar, and SignalExpress will do both types of logging, or you can write a program that will save data over time. Wavestar and SignalExpress can be downloaded from our Software Download section.

I would say that not everything goes smoothly even in middle/high price range.

Judging by this article I would say, I would prefer so-called seamless data logging. Because, in my opinion, it is easier to understand what occured with the signal or how much alterated the signal in the measuring point. Especially if you don't know which event you are expecting.

here is block-diagram (below, made myself) of problem device. As shown there are some points for signal measurement. I would like to add that these devices may work whole day or more with no signal loss, than suddenly become non working and quickly become working. But in order to start again a production process it needs hours. The worth of new device ~1200 Euro.
I still don't know whether issue appearance depends on ambient temperature, load, current consuption, etc. First I need to see somehow that particular device really makes the loss of signal at its RS-485 output. Or when master tells to keep silent the slave device keeps on translation, or there is just distortion of signal.
The next question is where there was a loss and at (before/after) which measur. point. It is so simple in theory.

So I am asking by which way it can be caught that issue.






Also thank you guys for youre replies.

[seronday]

The Siglent SDS1000CML or CNL series DSO may do what you want.

I have used an Atten ADS1062CAL which is a rebadged Siglent SDS1000CNL for this once.

In Roll Mode at time base speeds of 100ms/div and slower they have a long term Recording function.  Like a chart recorder.
The recording is done to internal memory , so this sets the length of the recording.
At the slowest time base of 50S/div the recording length is approximately 37hrs with only one channel on.
The sample rate at this time base speed is 50 samples / sec.
Peak detect mode can be enabled, so transient changes can be captured.

The recording can be replayed either in Roll mode or one frame at a time and can be copied to a USB drive as a .csv file.

The information in the User Guide does not fully explain how this record function works.