Siglent 1104x-e or 2202x-e?

[Fungus]

I'd like to know if, in your opinion, 2Gsa/s is overkill for a 200Mhz scope.

It depends on the rolloff of analog front end. On a 200MHz 'scope there will probably be visible signal past 500Mhz.

[tautech]

I'd like to know if, in your opinion, 2Gsa/s is overkill for a 200Mhz scope.

Late to this.....
I would normally recommend the cheaper SDS1202X-E as a first scope but in your case you have enough understanding to want to select a better product.
The SDS2202X-E offer 2x mem depth and 2x sampling, Bode plots and MSO, WiFi plus several other features over it's little 1202X-E brother.
If there's need for 4ch's the 1104X-E will be a better choice and then the feature set is closer but there is still 2 vs 1 GSa/s and 28 Mpts vs 2x 14 Mpts.

The decider for me and I guess you see it too, the SDS1202X-E is WYSIWYG whereas the SDS1104X-E and SDS2**2X-E are base platforms that WiFi, MSO or AWG can be added to if/when needed.

If we examine a usage scenario, 2 channels for SDS1104X-E and 2 ch's for SDS2202X-E, the sample rate will be the same: 1 GSa/s and memory depth equal too.
Either can be hacked and that's when SDS2202X-E will show a distinct advantage with BW @ 350 MHz. (SDS2352X-E)

[Fungus]

@MrW0lf: Your are showing a different effect here. You are zooming in on a signal without having enough information to display it properly.

Yes, I'm not sure what he's trying to prove there.

For audio work, even 100Mhz bandwidth is overkill.

Me? I'd turn on the 20Mhz filter if I was working on audio circuits with any of these 'scopes - to get rid of any distracting external noise.

[MrW0lf]

For 200MHz bandwidth 500Ms/s should be enough if the oscilloscope has a good sin x/x implementation. A higher samplerate would allow for a less steep roll-off of the input filter resulting in a slightly nicer frequency response but it is unlikely you'll have any benefit from it.

Theoretically perhaps in practice difference quite stark. Here I simulate 200MHz scope by enabling 200MHz HW LowPass and feeding square from 6GHz gen. At 500MSa/s pretty messy despite decent Sin(x)/x. 1G vs 2G difference mainly in timing resolution. So as rule of thumb I would say that for 100MHz scope you need 500MSa/s and for 200MHz scope 1GSa/s for "analog scope like" experience. Above that you will gain mainly timing resolution (which can be important sometimes).

[MrW0lf]

Yes, I'm not sure what he's trying to prove there.

I'm showing real world scenarios instead of dry theory, including importance of proper dot mode which will show analog like representation even if situation is totally hopeless for Sin(x)/x. But I understand this is new concept for you because your scope cant do it.

Edit: Attached same 200MHz edge square sampled at "hopeless" 200MSa/s due to maxed out memory at 10ms/. Dot mode saves situation and gives visual that scope having ~10x of memory would deliver (repetitive signals only).

[sn4k3]

Yes, I know I can hack 1104x-e to 200mhz. But:

i) The sample rate will still be 1Gsa/s vs 2 Gsa/s of the 2202x-e. I find 2Gsa/s attractive, but don't know if in practical terms there's a difference.

ii) If you want to be able to use the 200mhz bandwidth with 1104x-e you must also buy four 200mhz probes, so you should factor this in the equation.

iii) How can I be sure that the hack will effectively enable 200mhz bw without possible hidden quirks?

ii) Stock probes are 200mhz capable, i can confirm that, i have sds1104x-e with 200mhz.

Also: https://www.eevblog.com/forum/buysellwanted/fs-(uk)-siglent-sds1104x-e/msg2079202/#msg2079202



I took that great deal with all unlocked and options.

4 channel scopes are becoming for usefull than other features on most cases

[Andreax1985]

Another feature of SDS2000X-E series is 50ohm input termination... do you think it is a valuable addition to the scope?

[tautech]

Another feature of SDS2000X-E series is 50ohm input termination... do you think it is a valuable addition to the scope?

It is and it's nice to have it inbuilt but for the inexperienced or inattentive operator it has the vulnerability of being damaged with too much input. Just something to be wary of.   

[rf-loop]

Yes, I'm not sure what he's trying to prove there.

I'm showing real world scenarios instead of dry theory, including importance of proper dot mode which will show analog like representation even if situation is totally hopeless for Sin(x)/x. But I understand this is new concept for you because your scope cant do it.

Edit: Attached same 200MHz edge square sampled at "hopeless" 200MSa/s due to maxed out memory at 10ms/. Dot mode saves situation and gives visual that scope having ~10x of memory would deliver (repetitive signals only).

This is normal with Siglent's SARI.

Thanks god Siglent display modes - interpolations are fully post processed. Also in stop mode and in history buffer or segment buffer user can afterwards full freedom to select interpolation method or just real dots. It never flush out real ADC based samples.

Here is a rather extreme example of how  Siglent's well known  Sequential Acquisitions Random Interleaving - SARI works.

This is somehow like LeCroy's "Random Interleaved Sampling Mode - RIS" cousin. (do not still mix, they are still very different)

Here some old images (I do not translate here full text with these images) Sorry bit low gamma in images.


Original signal with 500MSa/s


Sample interval 20ns
fNyg 25MHz and signal 45MHz what also have lot of higher freq components (harmonics). Just totally hopeless - of course.

But then we have opportunity - SARI. Just turn off all interpolations, just true dots alone. Sequential acquisitions are still horizontally adjusted right related to trigger time position and -  here in next image is result, is our original signal.  Trigger engine works well and use "oversampling" and interpolation between decimated samples for adjust position. (of course if full sample rate then it do normal oversampling-fine interpolation between true samples as all modern scopes with full digital trigger engine do)


Sample interval 20ns.

Of course in this very extreme case there is not anymore enough dots (due to sampling interval and wfm/s*)  for draw continuous line but here persistence helps.
*One TFT frame do not include so many sequential acquisitions overlaid due to window zoom and main window time base. With more fast wfm/s rates of course dots density is much more high.

Of course SARI have also its limits and then more importantly, signal need be enough repetitive.
Do not mix this with some oscilloscopes ETS mode. SARI is not conventional DSO's ETS.

[Fungus]

Sigh. If only those cheap Rigols could have a dots mode:

Oh, wait a minute, they do...

[MrW0lf]

Oh, wait a minute, they do...

Nope, compare that to 1GSa/s trace and notice they are different eg it still somehow interpolated crap just drawn in different manner, or perhaps trigger system acting up. On Siglent 2GSa/s trace and even 100MSa/s trace are exactly of same shape eg there is no interpolation and trigger is rock solid. In fact this wast found out long ago:
https://www.eevblog.com/forum/testgear/new-rigol-ds1054z-oscilloscope/msg1821113/#msg1821113

[rf-loop]

Sigh. If only those cheap Rigols could have a dots mode:

Oh, wait a minute, they do...

And now run it so that input is nearly 2x fNyquist frequency square wave first harmonic and risetime around 1/5 of sampling interval (rt4ns @50MSa/s). I think you missed whole point totally.  In your image signal rising edge have around 2 samples. I do not know any single scope what can not do and reconstruct it (just turn Sinc on and your Riglol display very nearly same what you show with your dots+persistence. Rigol can not do at all what I show previously.
But yes, now we know it can also display dots. But it can not do at all what I show previously what is very different thing.
(btw, can  you show stopped scope same signal and dots image and then exactly same acquisition data then turn Sinc on display mode lines ... and also change Sinc off and lines... .. fun - is it)

[Gandalf_Sr]

Andreaux, as you can see there's LOT of opinions on which company does the best job taking ADC samples and displaying them and you  can quickly get lost in Nyquist theorum arguments etc.  The simple fact is that, for what your stated needs are, the DS1054Z does 99.5% of what you want to do and the UI is pretty good IMHO; if you want to pay more go ahead because, unless you get a bad one, they will all do what you want.

Also be aware that Tautech is a Siglent agent (it says so under his name).

[Fungus]

And now run it so that input is nearly 2x fNyquist frequency square wave first harmonic and risetime around 1/5 of sampling interval (rt4ns @50MSa/s).

Like this?

[tautech]

Also be aware that Tautech is a Siglent agent (it says so under his name).

And you sound like that's a bad thing ? 

[rf-loop]

And now run it so that input is nearly 2x fNyquist frequency square wave first harmonic and risetime around 1/5 of sampling interval (rt4ns @50MSa/s).

Like this?

Is this signal same as your previous message. If it is same signal, explain this previous image "sample fake dot" just after rising. (overshoot)
Then perhaps you did not understand what is 1/5...  Also I know why you want jump over other things what I ask...

I think this is waste of time. Years ago I try explain milk color to born blind people and soon I find it is least difficult of not even impossible until I fgind it is not even nessessary. After many times here in forum I feel bit same...

[Fungus]

Is this signal same as your previous message.

No, it was produced by user "TurboTom" using a real pulse generator (the previous was just me looking at an Arduino output pin)

I thought you'd already know this. In the original thread it's on the exact same same page as the images you (re)posted yesterday.

Here: https://www.eevblog.com/forum/testgear/new-rigol-ds1054z-oscilloscope/msg1818110/#msg1818110

[Andreax1985]

Andreaux, as you can see there's LOT of opinions on which company does the best job taking ADC samples and displaying them and you  can quickly get lost in Nyquist theorum arguments etc.  The simple fact is that, for what your stated needs are, the DS1054Z does 99.5% of what you want to do and the UI is pretty good IMHO; if you want to pay more go ahead because, unless you get a bad one, they will all do what you want.

Also be aware that Tautech is a Siglent agent (it says so under his name).

Is there any particular reason why I should go for the rigol DS1054Z (hackable to 100Mhz) in place of siglent 1104x-e (hackable to 200Mhz), apart from 100$ difference?

[Andreax1985]

I'd like to know if, in your opinion, 2Gsa/s is overkill for a 200Mhz scope.

Late to this.....
I would normally recommend the cheaper SDS1202X-E as a first scope but in your case you have enough understanding to want to select a better product.
The SDS2202X-E offer 2x mem depth and 2x sampling, Bode plots and MSO, WiFi plus several other features over it's little 1202X-E brother.
If there's need for 4ch's the 1104X-E will be a better choice and then the feature set is closer but there is still 2 vs 1 GSa/s and 28 Mpts vs 2x 14 Mpts.

The decider for me and I guess you see it too, the SDS1202X-E is WYSIWYG whereas the SDS1104X-E and SDS2**2X-E are base platforms that WiFi, MSO or AWG can be added to if/when needed.

If we examine a usage scenario, 2 channels for SDS1104X-E and 2 ch's for SDS2202X-E, the sample rate will be the same: 1 GSa/s and memory depth equal too.
Either can be hacked and that's when SDS2202X-E will show a distinct advantage with BW @ 350 MHz. (SDS2352X-E)

I've never read anywhere that SDS2202X-E is hackable to 350Mhz. Are you sure?

[tautech]

I'd like to know if, in your opinion, 2Gsa/s is overkill for a 200Mhz scope.

Late to this.....
I would normally recommend the cheaper SDS1202X-E as a first scope but in your case you have enough understanding to want to select a better product.
The SDS2202X-E offer 2x mem depth and 2x sampling, Bode plots and MSO, WiFi plus several other features over it's little 1202X-E brother.
If there's need for 4ch's the 1104X-E will be a better choice and then the feature set is closer but there is still 2 vs 1 GSa/s and 28 Mpts vs 2x 14 Mpts.

The decider for me and I guess you see it too, the SDS1202X-E is WYSIWYG whereas the SDS1104X-E and SDS2**2X-E are base platforms that WiFi, MSO or AWG can be added to if/when needed.

If we examine a usage scenario, 2 channels for SDS1104X-E and 2 ch's for SDS2202X-E, the sample rate will be the same: 1 GSa/s and memory depth equal too.
Either can be hacked and that's when SDS2202X-E will show a distinct advantage with BW @ 350 MHz. (SDS2352X-E)

I've never read anywhere that SDS2202X-E is hackable to 350Mhz. Are you sure?

Unproven as yet......but this guy is quite sure it can be done:
https://www.eevblog.com/forum/testgear/siglent-sds2000x-e/msg2018767/#msg2018767

[Fungus]

Is there any particular reason why I should go for the rigol DS1054Z (hackable to 100mhz) in place of siglent 1104x-e (hackable to 200mhz), apart from 100$ difference?

I think it's more than $100  , but.... you haven't mentioned a strict budget and for your stated work there's some reasons you might want to:
a) Better FFT
b) Lower front-end noise in the millivolt ranges
c) Possibility to link it to a signal generator (Siglent brand only, of course) and do frequency sweeps/Bode plots

All the talk about sample rate, etc., is a red herring but as usual the Siglent fan club has derailed the thread with a Rigol-Bashing frenzy and without reading your requirements.

Relevant:
* For audio work you should turn on the 20MHz filter and get rid of all the distracting radio noise from your surroundings that will distract you from real problems in the circuitry.

* For radio work you're interested in sine waves so discussion of rise times is pointless. With the bandwidth of either device you'll have more problems just getting the signal to the 'scope intact.

[Andreax1985]

I think it's more than $100  , but.... you haven't mentioned a strict budget and for your stated work there's some reasons you might want to:
a) Better FFT
b) Lower front-end noise in the millivolt ranges
c) Possibility to link it to a signal generator (Siglent brand only, of course) and do frequency sweeps/Bode plots

All the talk about sample rate, etc., is a red herring but as usual the Siglent fan club has derailed the thread with a Rigol-Bashing frenzy and without reading your requirements.

Relevant:
* For audio work you should turn on the 20MHz filter and get rid of all the distracting radio noise from your surroundings that will distract you from real problems in the circuitry.

* For radio work you're interested in sine waves so discussion of rise times is pointless. With the bandwidth of either device you'll have more problems just getting the signal to the 'scope intact.

how about the 50ohm input in the 2202x-e? do you see it as a big plus?

[tautech]

c) Possibility to link it to a signal generator (Siglent brand only, of course) and do frequency sweeps/Bode plots

You really need to get up with the play:
https://www.eevblog.com/forum/testgear/siglent-sds1104x-e-and-sds1204x-e-bode-plot-with-non-siglent-awg/

[Gandalf_Sr]

Andreaux, as you can see there's LOT of opinions on which company does the best job taking ADC samples and displaying them and you  can quickly get lost in Nyquist theorum arguments etc.  The simple fact is that, for what your stated needs are, the DS1054Z does 99.5% of what you want to do and the UI is pretty good IMHO; if you want to pay more go ahead because, unless you get a bad one, they will all do what you want.

Also be aware that Tautech is a Siglent agent (it says so under his name).

Is there any particular reason why I should go for the rigol DS1054Z (hackable to 100Mhz) in place of siglent 1104x-e (hackable to 200Mhz), apart from 100$ difference?

I don't know if there is a hardware difference between an official DS1054Z and a DS1104Z.

If you Google "Riglol" (not a typo) you will find a site (first hit for me) where you can enter the serial number and generate the code to unlock a DS1054Z up to a DS1104Z using option DSFR  (it will then report itself as a DS1104Z). Mine is flat to 130 MHz after this 'upgrade'. Some people claim it goes to 200 MHz (the probes I got were 200) but I found that, while I could see 200 MHz waveforms, the Y amplitude was falling - again, for what you say you want to use it for, it's all that you need.

[Gandalf_Sr]

Also be aware that Tautech is a Siglent agent (it says so under his name).

And you sound like that's a bad thing ? 

You're being a bit touchy - I just wanted to make Andreaux aware of your allegiance