Siglent SDS2000X Plus

[Martin72]

But is there a way to average FFT? And while we are at it: FFT min hold / max hold?

Yes...

[Martin72]

Maybe this is a result of using a less capable FPGA, but honestly I would assume this (and other things like the missing qualified trigger) could be more of a way of differentiation between the SDS2000X+ and the SDS5000X

Makes sense, why spend money for the SDS5000 when you could get it almost all with the SDS2000X+....

[Someone]

You make it sound as if by some magic MegaZoom uses all the memory (for sampling data of the current acquisition) all the time, but this is, plain and simple, wrong. Also, MegaZoom is not magic, it's actually a pretty simple, although clever, idea to maintain very high update rates - something which was valued by people coming from analog scopes.

I never stated that it uses all memory for each displayed trace. However the Megazoom ASIC can never know what is the last acquisition in advance

It doesn't. Hence why it only uses the full memory for the last acquisition *after* you press the STOP button (or the one acquisition in Single mode).

May I ask, if think this kind of Megazoom scope, how it is if trigger mode is normal so that it do acquisition only when trigger event happen. Then suddenly there do not exist anymore trig events. Scope stop acquisition and waiting next trig event displaying last acquisition result. Is this last one now there short or long memory (if this miraculous "clever parallel acquisition system" is doing magic things like some may dream)
Next later no anymore trigger events and user press stop.

After then he think he zoom and pan to captured details outside of display as here have discussed around this case lot of... 

I do not want assume anything, so I ask.

The simple answer, with infrequent triggers the last capture in run mode will be "short".

In run mode the acquisition memory depth is more or less the visible screen multiplied by the sample rate (automatically adjusted for maximum sample rate at all times).
In single mode the acquisition memory depth is the maximum available, and may extend outside the visible screen.

The trick is there is a short window of a fraction of a second after pressing stop (from run mode) when the scope will automatically/invisibly trigger a new single capture and show that instead.

[nctnico]

The trick is there is a short window of a fraction of a second after pressing stop (from run mode) when the scope will automatically/invisibly trigger a new single capture and show that instead.

Buzzzz. Wrong answer!!!

Get a Keysight Megazoom scope and try it yourself. Send in a signal which is a burst and you'll see you get the full memory length for the last trigger event without pressing stop. FFS I owned one so I'm pretty damn sure how it works.

[Someone]

The trick is there is a short window of a fraction of a second after pressing stop (from run mode) when the scope will automatically/invisibly trigger a new single capture and show that instead.

Buzzzz. Wrong answer!!!

Get a Keysight Megazoom scope and try it yourself. Send in a signal which is a burst and you'll see you get the full memory length for the last trigger event. FFS I owned one so I'm pretty damn sure how it works.

Have you tried it with a variety of their scopes? What I described is exactly what happens on several different models of their megazoom IV scopes, I know as I have tested it and confirmed this behaviour across different models. Just to humour you tried it again with a pulse burst, no different.

When switching from run to stopped, it doesn't have any captured acquisition data around the visible window, unless another trigger arrives in the hundred or so ms after stopped was pressed.

Different scopes have different memory management and ways of operating, stop the presses!

[nctnico]

Did you test with a short time/div? Otherwise there will be no data outside the screen because there isn't any data. Or simply zoom in?

[Someone]

Did you test with a short time/div? Otherwise there will be no data outside the screen because there isn't any data. Or simply zoom in?

Unlike you I'm not trying to be obtuse. The description and test were in reply to the polite and considered question by rf-loop. There are all sorts of minor details and specifics that could make a long winded report, perhaps you would like to undertake that or pay someone to compile it for you?

Of course this was under the situation of a short visible window where the available memory depth exceeded the visible span, the question and answer are clear enough but you come in with FUD and your simply incorrect accusations.

[nctnico]

Well, the Agilent 7104A I owned had the Megazoom III ASIC and not the IV so there may be (very unexpected) differences. Unfortunately I no longer have the 7104A to test but I'm 99.9% sure it works the way I described.

Edit: I dug around in some manuals. It seems the Keysight 3000X series limits the memory to the size of the screen by default. However there seems to be a 'Digitizer mode' which allows to set the memory to a defined length at the cost of losing various functions. I have not found such a limitation in the manual for the Agilent 7104A.

[rf-loop]

Now I only want to have this math trace on the screen, so I press the channel one knob, knob lightning is off, channel one trace on the screen is off too, math trace of channel one is still visible.

May I ask why you use channel one knob for shut off channel one trace display. Channel one knob is for on/off whole channel (and its displayed trace).

In other Siglent model (example SDS1x04X-E) there is menu setting where can find trace display on/off what leave channel normally working but can turn off ONLY trace display, not affect at all how this channel is still fully working in background only its trace display light  off.
If this function is missing in SDS2kXPlus FW it must be mistakenly forgotten. But  if so... now need wait bit more because Siglent office holiday time is extended now least one week more so all delays now more. How all continue, next week give more information about 2019-nCoV case here in China.

[Martin72]

May I ask why you use channel one knob for shut off channel one trace display

As I explained before, I'm used to doing it this way on lecroy scopes.
Interesting:
The SDS2000X+ is the only series, where averaging and eres are in the math menu.
All other siglents, 1000/2000/5000, got them in the acquisition setup.
Why doing this different by the 2000X+
Edit:

In other Siglent model (example SDS1x04X-E) there is menu setting where can find trace display on/off

And it seems, it´s the only model.

[rf-loop]

May I ask why you use channel one knob for shut off channel one trace display

As I explained before, I'm used to doing it this way on lecroy scopes.
Interesting:
The SDS2000X+ is the only series, where averaging and eres are in the math menu.
All other siglents, 1000/2000/5000, got them in the acquisition setup.
Why doing this different by the 2000X+
Edit:

In other Siglent model (example SDS1x04X-E) there is menu setting where can find trace display on/off

And it seems, it´s the only model.

Trace can shut off also in SDS1x02X-E models. Feature added in last FW (1.3.26)
Also I think it was not originally in 4 channel models but I can not find what time it was added.

Lack of this feature need complain Siglent. Although it is not most urgent.

Eres
Because real ERES is math function. Is it also so that it do not destroy original ADC data. (lets hope it is real ERES or least some main things, non destructive)

Please read what @Wuerstchenhund have said about ERES.

It is very different what is example some oscilloscopes HiRes acquisition is what use destructive hardware boxcar average. I do not know how much just SDS2kXPlus ERES works like LeCroy explained ERES but just because this change to math, perhaps more close LeCroy than other models.


http://cdn.teledynelecroy.com/files/appnotes/differences_between_eres_and_hires.pdf

http://cdn.teledynelecroy.com/files/appnotes/an_006a.pdf

[tautech]

Yes well a screenshot of a math trace from a hidden trace was posted early in this thread.
There is little reason to imagine this feature hasn’t made it to product release.
Post 34 IIRC

[maginnovision]

Other than FFT which was already confirmed I didn't see math without source trace displayed.

[Martin72]

Me too.

Here´s the first "list" of wanted updates for the sds2000x+:

- Adding the possibility of hiding channel traces for displaying math/decode functions only

- Adding trigger function "qualifying trigger A B"

- Labeling the F1 and F2 "channels" with the actual math function

- Fixing a performance problem when using math and memorydepth is above 20kpts

- Adding trigger mode "Alternate" if possible

Oh...this was it at all - Until now..

Another thing:

At home I got a lack of testsignals, e.g. serial decoding signals or signals with runts/glitchings.
So I think about to buy the siglent demo board, for having some basic test signals at home, not only for my actual scope.
What are your thoughts on this ?
It´s price doesn´t matter to me, I´m going to sell some of my not longer used test things, so it would cost me the half or less.

[nctnico]

Money is likely better spend on a USB logic analyser / pattern generator in order to  make patterns. Ika logic SQ100 seems to be quite cheap but I have no idea on the quality and usefullness of their software.

[TK]

I don't think it is any worth to buy a demo board when you can accomplish the same with an STM32 nucleo board or any Arduino compatible

[rf-loop]

- Adding trigger mode "Alternate" if possible

This is, how I can say it enough nicely... yes of course we can write to SantaClaus many kind of wishes... and so on.

But honestly, if we want in 4 channel scope true full Quad independent timebase Quad independent trigger "Alternate"  I believe simply, this gift do not come never to any already purchased Siglent 4 channel scope. And I believe also that not any future models... If drop independent timebase and think only single timebase individual triggers Alternate, not so hopeless wish technically... but, forecasting is difficult - especially to predict the future.

If 2 channel is enough and so that both have own memory and own ADC chip as is case when SDS2k works in interleaved mode, perhaps independent Alt is bit more easy..
 
Least I know some use for this feature but also can live without. For dedicated special needs mostly there is other solutions.

It is fun but Siglent have this kind of scope what have dual timebase alternate mode.
Old SDS1000CML/CNL. 
These have. Both channels can drive with independent t/div and both channels have also individual triggers. So it can capture and display very very different signals simultaneously. One weak point is that both channels have same samplerate what is limiting factor for extremely different speed signals. Naturally if other channel have 1Hz signal and other have 100kHz signal also this 100kHz channel wait until 1Hz waveform is captured and then it capture this 100kHz signal... and continue alternatively, just like old single beam analog scope  ALT works. (True dual beam analog scopes is Very different story, example Tek's 7044 where other trace trig do not need wait other...) Also both channels have fully independent trigger settings as also independent timebases (iIrr, if I remember right) and even both may have dual timebase, but these was old time analog oscilloscopes RollsRoyces (including price).

After old SDS1000 series various scopes it is not implemented in any other Siglent scopes what is sold as Siglent brand as far as I know.
Why exactly, I have no idea.

I hope Siglent  take seriously this Alt feature,  demand for this pops up repeatedly. It is not just sales brochure checkbox feature. It need be, performance and features (how does it work) of this mode is then other question.

[Martin72]

This is, how I can say it enough nicely…

That´s very kind from you.
It´s indeed a wish for christmas, not important you can "simulate" it with other trigger functions.
Unlike more important are the channel hiding thing and the lost of performance when using math and memorydepth is greater than 20k.
This siglent is brandnew (it seems to me, I´m the only one who got it - Or I bought the only part in stock, after my buy it wasn´t avaible  ), let´s wait for response from siglent.
By the way, I wrote to siglent eu support a couple of days ago.
No answer.

[rf-loop]

By the way, I wrote to siglent eu support a couple of days ago.
No answer.

Perhaps these kind of things are not very urgent and are forwarded to engineers in China who all are holiday and after then extended holiday due to this exceptional situation.

This is partially second hand info but... Also even when Siglent European office is in Germany also there China new year festival and holiday time affect to some things. And now more due to severe virus epidemic in China. Originally Siglent HQ (and afaik also factory) in Shenzhen start after holidays 3.2.  but due to 2019-nCoV epidemic here in China this time table is obsolete. At this time HQ is designed to return and start 10.2. (and as we know situation here in China it may change or perhaps it do not start with full power, depending how it affect peoples traveling back to work and other things related to mainland China virus epidemic. Main thing is now do all for attenuate this virus spreading and take care that average infection spreading system "gain" goes as fast as possible well under 1 so that epidemic suffocates. This goes highly over normal business short time needs.)

This exceptional situation may affect many things and sure it affect less urgent things highly although sales and repair service in European office works as normally as possible.

[Martin72]

Sure, actually they´re much more important problems and it´s understood, that this here is maximum unimportant thing against.
But..
A short "we´ve received your message" ( most common automatic reply)or "we´ve forwarding this" would be nice, especially when they are in germany.

[Martin72]

Money is likely better spend on a USB logic analyser / pattern generator in order to  make patterns. Ika logic SQ100 seems to be quite cheap but I have no idea on the quality and usefullness of their software.

I had a look on it and downloaded/run the software - Remembers me of the bitscope thing.
I think, I would buy the demo board instead, when I can refinance it by selling no longer used things.
Otherwise not.

[nctnico]

I don't see much practical use in a demo board which generates pre-defined signals. First of all you can't use it as a tool and secondly it is very unlikely to uncover odd behaviour of your oscilloscope.

[tautech]

So I think about to buy the siglent demo board, for having some basic test signals at home, not only for my actual scope.
What are your thoughts on this ?
It´s price doesn´t matter to me, I´m going to sell some of my not longer used test things, so it would cost me the half or less.

Have you studied the manual ?
https://www.siglenteu.com/accessory/stb-3/

It produces more than just a few 'ordinary' signals some of which will test your abilities to get stable triggering.
As you fiddle with a few brands it might be a useful tool to compare capabilities.
I've had one for a good few years and use it often when demonstrating DSO's to customer when they call to collect or buy.

[Martin72]

I´ve ordered this thing now.
Welectron gives you 45 days for testing..

[TK]

So I think about to buy the siglent demo board, for having some basic test signals at home, not only for my actual scope.
What are your thoughts on this ?
It´s price doesn´t matter to me, I´m going to sell some of my not longer used test things, so it would cost me the half or less.

Have you studied the manual ?
https://www.siglenteu.com/accessory/stb-3/

It produces more than just a few 'ordinary' signals some of which will test your abilities to get stable triggering.
As you fiddle with a few brands it might be a useful tool to compare capabilities.
I've had one for a good few years and use it often when demonstrating DSO's to customer when they call to collect or buy.

For a distributor, it makes sense to have one.  For the end user, no thank you