Siglent SDS7000A 3 and 4 GHz 12 bit DSO's

[Martin72]

I am very excited to see the SDS7000, assuming it will have a faster UI, since it is using a I3.

And with 4 cores and 3.6Ghz not the badest one.
Before we bought the waverunner and the HDO6034A (both with i5 cpu), we got a hdo4000 (intel i3) on loan for two weeks.
The scopes based on a pc platform are simply in a different league in terms of overall performance.
I work with the HDO6034A almost every day and everything is so smooth, so fast, it has 8(!) math channels, until now I had 4 of them running at the same time and nothing has changed in speed, just great.
Last weekend at home on my HD I had 2(.... ) math channels active at the same time (both with FFT) and a few readings in the display.
Well...
You get what you paid for, even if it wasn't exactly little.
(Not to mention the unspeakable slowdown of the entire system when the Bode Plot is active).
But such comparisons bring you back down to earth when you think spending more money isn't necessary.

[JPortici]

The first thing my colleagues do is to reduce the memory depth....
It slows the system down - Nothing for unpatient users..

It also allows for more history waveforms i do it for both reasons. scope set at 12.5MPts, unless i need the samplerate at lower timebases

[tautech]

The first thing my colleagues do is to reduce the memory depth....
It slows the system down - Nothing for unpatient users..

It also allows for more history waveforms i do it for both reasons. scope set at 12.5MPts, unless i need the samplerate at lower timebases

SDS5000X right Jack ?
Have you not installed recent firmware that allows setting of a fixed sample rate ?

[JPortici]

The first thing my colleagues do is to reduce the memory depth....
It slows the system down - Nothing for unpatient users..

It also allows for more history waveforms i do it for both reasons. scope set at 12.5MPts, unless i need the samplerate at lower timebases

SDS5000X right Jack ?
Have you not installed recent firmware that allows setting of a fixed sample rate ?

Yes. I did, but i don't care for the function. I've always been content in how the scope operated out of the box and i'm used to it. I tried the other memory management modes but failed to find a scenario in which either would be an advantage over the Max Memory we all know and love (or a big enough advantage that i bother going through the menu to find the optimal setting from time to time)

[Performa01]

The first thing my colleagues do is to reduce the memory depth....
It slows the system down - Nothing for unpatient users..

It also allows for more history waveforms i do it for both reasons. scope set at 12.5MPts, unless i need the samplerate at lower timebases

SDS5000X right Jack ?
Have you not installed recent firmware that allows setting of a fixed sample rate ?

Yes. I did, but i don't care for the function. I've always been content in how the scope operated out of the box and i'm used to it. I tried the other memory management modes but failed to find a scenario in which either would be an advantage over the Max Memory we all know and love (or a big enough advantage that i bother going through the menu to find the optimal setting from time to time)

Your initial statement was very clear. You stick to 12.5 Mpts memory except when you really need longer captures = slow timebase settings where the samplerate drops below your requirement for the acquisition bandwidth. So what you want is sufficient samplerate rather than constant samplerate - the latter is only useful in some very special cases, e.g. where the scope bandwidth depends on the samplerate in a less familiar way, such as in ERES mode.

That's what generally is considered best practice, as 10-14 Mpts is some magic region for all scopes based on that current Zync platform...

[JPortici]

The first thing my colleagues do is to reduce the memory depth....
It slows the system down - Nothing for unpatient users..

It also allows for more history waveforms i do it for both reasons. scope set at 12.5MPts, unless i need the samplerate at lower timebases

SDS5000X right Jack ?
Have you not installed recent firmware that allows setting of a fixed sample rate ?

Yes. I did, but i don't care for the function. I've always been content in how the scope operated out of the box and i'm used to it. I tried the other memory management modes but failed to find a scenario in which either would be an advantage over the Max Memory we all know and love (or a big enough advantage that i bother going through the menu to find the optimal setting from time to time)

Your initial statement was very clear. You stick to 12.5 Mpts memory except when you really need longer captures = slow timebase settings where the samplerate drops below your requirement for the acquisition bandwidth. So what you want is sufficient samplerate rather than constant samplerate - the latter is only useful in some very special cases, e.g. where the scope bandwidth depends on the samplerate in a less familiar way, such as in ERES mode.

That's what generally is considered best practice, as 10-14 Mpts is some magic region for all scopes based on that current Zync platform...

the ERES hint gave me an idea, i'm still waiting for digital filters on the math (or on the input) channels, probably i can combine fixed sampling with ERES so i can achieve a low pass filter i can control
Did not know of this magic region but yes, 12.5Mpts proved to be a good compromize between enough samples (helps with SENT/CAN decoders at the standard frequencies) and UI Speed

[Martin72]

i'm still waiting for digital filters on the math (or on the input) channels

You got a 5000X, right ?

[tv84]

Yes.

[Martin72]

Argh, good morning...

OK, so I guess filters will come with one of the next firmware updates(memory management as a kind of prerequisite has it already).

[Martin72]

Yes. I did, but i don't care for the function. I've always been content in how the scope operated out of the box and i'm used to it. I tried the other memory management modes but failed to find a scenario in which either would be an advantage over the Max Memory we all know and love (or a big enough advantage that i bother going through the menu to find the optimal setting from time to time)

Me too.
"Growing up" with Lecroy scopes (93XX, waverunner LT, wavesurfer, waverunner 9000, HDO) I had never missed anything in that direction.
So the discussion about the memory was new to me when I entered the forum here.
Should I use the new filter function(on my 2000HD), this could change, but even then I don't need to worry, because the scope takes care of that for me - Even if it first asks whether it should change...
(With my two private Rigol Scopes Auto-Memory was the default setting, that will have its reasons).

[rf-loop]

Inside SDS7000A

Think about it.... 
On December 28, 2022, Siglent released two new products: SDS7000A digital oscilloscope and SFA8001 front-end amplifier chip.

SDS7000A is the first 4GHz bandwidth, 12-bit high-resolution digital oscilloscope in China, which raises the domestic high-resolution oscilloscope to a new height and fills the domestic gap in related fields! SFA8001 is the first self-developed oscilloscope front-end amplifier chip with a bandwidth of 8GHz released in China! The chip laid a solid foundation for the research and development of Siglent's 8GHz and subsequent higher bandwidth oscilloscopes.
It is expected that in the near future, a new generation of Siglent digital oscilloscope equipped with self-developed chips will meet with you. At the same time, the development of SFA8001 has also accumulated experience in the development of 16GHz bandwidth amplifier chips.



Siglent's first self-developed chip SFA8001
Amplifier chip: the core of the analog front-end circuit The three core chips inside the digital oscilloscope are: the front-end amplifier chip,
the ADC chip and the digital signal processing chip. Each of these three chips has its own role and has a direct impact on the performance of the oscilloscope.



As the first self-developed oscilloscope front-end amplifier dedicated chip of Siglent, SFA8001 has a high bandwidth of 8GHz and a low noise density of 1.9nV/√Hz, which can help oscilloscopes observe faster and smaller signals.



Self-developed chips: the only way to develop localization
Siglent's chip self-research road is rooted in product design needs, rather than the "muscle show" of corporate strength.
High-bandwidth front-end amplifier chips are difficult to directly implement through board-level circuits, and chipization is an inevitable trend. As one of the three core chips of oscilloscopes, the front-end amplifier chip is different from the ADC chip and the digital signal processing chip, and its application area is narrow and the design requirements are high, so it is difficult to purchase directly in the market.
Not only that, as a dedicated chip, the front-end amplifier chip often needs to be optimized according to the actual oscilloscope design when designing the circuit to better meet the product performance. Therefore, in the field of high-end oscilloscopes, analog front-end chips often exist in the form of self-developed by each oscilloscope manufacturer.



After that the next step is 16GHz...and so on...

The Emperor is far and the sky is high...run your own paths in freedom and fly higher...

Siglent, stock code: 688112.SH, Shanghai Stock Exchange (SSE).

[luudee]

Any pricing info yet ?


luudee

[Martin72]

Hi Rob,

https://www.eevblog.com/forum/testgear/rigol-hdo1000-and-hdo4000-12bit-oscilloscopes-launched-in-china/msg4708985/#msg4708985

To stop offtopic there...

In my opinion, the gap between the 2000X HD and the 7000A is too wide in terms of offering 12bit scopes for the western market.
I strongly expect the 7000A to cost at least 10000, it is a different league where it will play together with the 12 bit scopes from lecroy and Tek.
There is nothing in between the two extremes.
The 2000X HD is very good, I need to know that for various reasons.
But if you want a bigger screen, for example, or more math channels, as well as increased general performance, but don't want to drop back to 8 bit, you won't find it at Siglent.
I don't know if it's smart to leave the field to other brands in the western market.

[Martin72]

Listed on the chinese website:

https://www.siglent.com/products-overview/sds7000a/

100,000,000 Wfm/s ??   

[egonotto]

Listed on the chinese website:

https://www.siglent.com/products-overview/sds7000a/

100,000,000 Wfm/s ??   

Hello,

and "Low noise floor: only 4 μVrms at 220 GHz bandwidth"

I think they mean 220 μVrms at 4 GHz


Best regards
egonotto

[Martin72]

 

You´re right, it must be a failure in the description.
But apart from that, nice scope..
And only siglent knows why they´re offering two vertical resolution models.

[egonotto]

You´re right, it must be a failure in the description.
But apart from that, nice scope..
And only siglent knows why they´re offering two vertical resolution models.

Hello,
yes very nice scope and I think they mean 220 μVrms at 4 GHz.
 
Cleverscope CS328A even had three different resolutions on offer:

"CS328A   4M Samples, 10 bit sampler, 8 digital inputs, USB I/O
  The CS328 can be upgraded to 8M memory and 12 or 14 bit dual channel sampler, plus signal generator."

Red Pitaya was also once available with 10 or 14 bit

Best regards
egonotto

[tautech]

Listed on the chinese website:

https://www.siglent.com/products-overview/sds7000a/

100,000,000 Wfm/s ??   

Me thinks you need use a different translator:
main features
Bandwidth: 4GHz, 3GHz, 2GHz
Memory depth: up to 1Gpts/channel
Vertical resolution: 12-bit /10-bit
Real-time sampling rate: up to 20 GSa/s
Waveform capture rate: up to 1 million frames per second
Low noise floor: as low as 220 μVrms at 4 GHz bandwidth

[rf-loop]

And only siglent knows why they´re offering two vertical resolution models.

I know there is peoples who do not think money because from wall you can get more but there is still peoples and companies who may think also money...

But still we can see one single bit more do not double price but it doubles resolution.
And even more, here we get 2 bits more and we get quadruple resolution with same frequency bandwidth and samplerate.

It is totally different case if we reduce  samplerate (and BW) when we rise resolution. This is why.
Not because they just want to be uncomfortable.

Of course, everyone also understands how the market works, as well as why pay more if you don't need more.

12 bit, SDS6204 H12 Pro   2 GHz,  ¥229,880  Normal model
10 bit, SDS6204 H10 Pro   2 GHz,  ¥139,880  Economy model

The same realities are affecting the model SDS7kA as well.
and this 100 xxxx/s  need translates as 100*10000 and all who understand chinese numbering system and these "words" know it. But some translators perhaps not. When chinese say something what heards somehow as "millon" it is 10k (10000) and many times example prices are told in "millons" example in house markets. It is 10k not 1M.

[Martin72]

Me thinks you need use a different translator:

Microsoft Edge translating...

[tautech]

Me thinks you need use a different translator:

Microsoft Edge translating...

I find this very useful to decypher info from newly discovered products:
https://translate.google.com/

You can Copy/Paste can't you ? 

[Sighound36]

I'd be up for trying the 4ghz 12 bit models, have some budget left over after the recent home lab 4Ghz Wavepro HD that would be the model to go head to head with.

[Martin72]

12 bit, SDS6204 H12 Pro   2 GHz,  ¥229,880  Normal model(appx 32183€
10 bit, SDS6204 H10 Pro   2 GHz,  ¥139,880  Economy model(appx 19443€

The prices...Then I don´t want to know the ones for the 7000..
And then there is the 8 bit model here fore really less money than the 10 bit "Economy model" (appx 9980€ for the 2Ghz) 
It's hard to believe that all three ADCs are pin-compatible with each other, so they would have to have different layouts.
Apart from this, for 19000€ or 32000€ I could get a state of the art scope from lecroy

[TomKatt]

Stupid question from a 200 Mhz hobbyist...  What are you folks doing that requires 7-8 Ghz bandwidths?  Can you give an example of where you need that level of gear?

I imagine that just setting up a probing solution at those frequencies is an art unto itself.

[Martin72]

Just for example:

DDR3 electrical testing and timing analysis requires a 6 Series MSO oscilloscope with a recommended bandwidth of 8 GHz to cover entire range of DDR3 speed grades. However, for signal integrity testing and debug, a minimum bandwidth of 4 GHz would suffice most of the user's needs.

https://download.tek.com/datasheet/DDR3-and-LPDDR3-Measurement-and-Analysis-Software-on-6-Series-MSO-Datasheet-55W614702.pdf