Siglent's new product- MSO/SDS2000 Series

[Herman]

I notice that Siglent has released the 110,000 / sec high capture rate MSO/SDS2000 Series Super phosphor oscilloscopes. The original website is: http://www.siglent.com/cn/news/detail.aspx?id=100000061731257&nodecode=105002004

 SIGLENT MSO/SDS2000 Series Super Phosphor Oscilloscopes?SPO?

A brief parameter is as below:

Bandwidth 70MHz, 100MHz, 200MHz, 300MHz
Real time sample rate, Analog channel: 2GSa/s, Digital channel: 500Msa/s (MSO Function) 2/4 analog channels, 8 digital channels (MSO Function)
Built-in 25MHz function/arbitrary waveform generator
IIC, SPI, UART/RS232, CAN, LIN) (Decode Optional)
trigger functions (Pattern, Window, Interval, Dropout, Runt) and HDTV video trigger function


Now MSO/SDS2000 is launched only in China, I heard that it was sold very well in China now. I think it will soon appear on the market all over the world.
The detail datasheet of MSO/SDS2000 series is on the website
http://www.siglent.com/cn/product/detail.aspx?id=100000007854687&nodecode=105008001

[rf-loop]

English (chinglish) datasheet is here.
Here is now more than one or two threads in this forum about this new series.

[Herman]

thank you very much?it is appreciate?easier to understand than the Chinese edition

[olsenn]

Anybody have any pricing info?

How do Sigilent scopes typically compare to Rigol's?

[rf-loop]

This table I get from Siglent and I'm not responsible if there is errors.

[tinhead]

based on specs excellent DSO/MSO, let's hope the price will be on Rigol DS2xxx level (for comparable 2ch and same bw model).

[Carrington]

Sure, but certainly everyone is thinking the same, i.e. the Rigol DS2000 series is already hacked.
But for sure Siglent thought about that... What will be their strategy?

A teardown would be the beginning, because this can wake the restless mind of a possible hacker for the SDS2000 series.


I edit, because for now Rigol DS2000A can't be hacked.
Source: https://www.eevblog.com/forum/testgear/sniffing-the-rigol%27s-internal-i2c-bus/msg319032/#msg319032

[olsenn]

This table I get from Siglent and I'm not responsible if there is errors.

There are definately errors in that table. Some that I notice right away are:
      *The DS2000A DOES feature 50-ohm in addition to 1M
      *The DS2000A has (optional) 56M memory depth
      * The DS200A does NOT have a built in sig-gen; only the DS2000A-S does

And what about intensity grading, vertical sensitivity + noise floor, and most importantly... HACKS!!!!!

Many questions have yet to be answered, but good to see another company has stepped up to the challenge

[Hydrawerk]

https://app.box.com/s/dfas6a95v9ges74gdgli
SDS2000 has 110 000 waveforms per second, but under what circumstances? There is no specification in the datasheet. Note that Rigol DS2000A has the specification in the datasheet.
SDS2000 is more feature rich than my DSOX2000 but I doubt if the user interface is responsive.

[Hydrawerk]

Here is now more than one or two threads in this forum about this new series.

Yes. See more pictures here: https://www.eevblog.com/forum/testgear/spotted-new-birds-%28siglent%29/msg307635/#msg307635

[Carrington]

https://app.box.com/s/dfas6a95v9ges74gdgli
SDS2000 has 110 000 waveforms per second, but under what circumstances? There is no specification in the datasheet. Note that Rigol DS2000A has the specification in the datasheet.
SDS2000 is more feature rich than my DSOX2000 but I doubt if the user interface is responsive.

I would love to see a table for the SDS2000 like this :

or this


In China is on sale, I think, but I can't find any video with it into operation.

[olsenn]

Is it just me that thinks the waveform update rate of the DS2000 is pitiful? I mean it's rated for 50k wfm/s but one time base gets 46.8k, and the rest all get less than half of the rated spec! The average is more like 5k wfm/s!

Agilent's (x2000 I believe it was) is rated for 50k but actually maintains 50k across the majority of the time bases

[Carrington]

Is it just me that thinks the waveform update rate of the DS2000 is pitiful? I mean it's rated for 50k wfm/s but one time base gets 46.8k, and the rest all get less than half of the rated spec! The average is more like 5k wfm/s!

Agilent's (x2000 I believe it was) is rated for 50k but actually maintains 50k across the majority of the time bases

Yes, I knew, long ago I check that. Agilent in contrast to other manufacturers, has published it in detail. But not long time ago, DS2000 series in its category had no rival.

[marmad]

Is it just me that thinks the waveform update rate of the DS2000 is pitiful?

Yes, it's just you.

[Carrington]

Hi Marmad!
I suppose  that the tables above are correct. True?
Or is it just a small error with the coma?


Preliminary estimated list prices in Finland 30.10.2013. *)
Available December 2013

SDS2072 EUR 698.00
SDS2074 EUR 998.00

SDS2102 EUR 898.00
SDS2104 EUR 1288.00

SDS2202 EUR 1250.00
SDS2204 EUR 1700.00

SDS2302 EUR 1750.00
SDS2304 EUR 2200.00

Options **)
SDS-2000-FG /25MHZ Function/Arb. Generator option EUR 160.00
SDS-2000-DC /IIC, SPI, UART/RS232, CAN, LIN decoding option EUR 160.00
SDS-2000-PA /Power analysis option EUR 270.00

*) Prices may change depending if shipping costs or currency rates change.
**) Prices valid if order together with oscilloscope

Source: http://siglent.freeforums.org/siglent-sds-mso-2000-series-oscilloscopes-t88.html

110K wfm/s at 50ns
Source: http://www.861718.com:8080/webnews.newdetailview.do?iid=20131024120148590

[marmad]

I suppose  that the tables above are correct. True?

Yep - last time I checked. Nope - not any more. Waveform update rates have greatly improved at the smaller time bases on the DS2000 series.

Looking through the SDS2000 datasheet:

On the plus side, I like the option for intensity OR color temperature display.

On the minus side, it has less trigger options than the DS2000 series (USB is a very handy one that's missing); it makes abysmal use of the EXTERNAL trigger (just EDGE - same as the DS2000 - what's wrong with these designers?); but worst of all, I see absolutely no mention of segmented memory - which is (combined with the segment analysis tool), IMO, one of the best features of the DS2000 series.

I use the 56M of deep memory most often for recording segments, and if the SDS2000 is missing that feature, I'd never consider it as a viable purchase after having had the feature.

[Carrington]

Yes, you're right.

To manufacturers, I think it's time to specify in detail the wfm/s. Because with the series SDS2000 may happen as with the GDS2000 series:

[marmad]

People shouldn't get too obsessed with the wfrm/s speeds (within a certain range). The truth of the matter is that, in day-to-day use, you won't really notice much display difference (assuming you have intensity grading) - or increase your chances significantly of catching a glitch - until your wfrm/s rate jumps by a factor of ~10x or more.

Here are the average test times for catching a repeating signal fault with a probability of 99.9% (T=100 ns, GlitchRate=10/s).

Acquisition Rate	Test Time
100 wfrm/s	19h:11m:08s
1,000 wfrm/s	01h:55m:00s
10,000 wfrm/s	00h:11m:31s
50,000 wfrm/s	00h:05m:45s
100,000 wfrm/s	00h:01m:09s
1,000,000 wfrm/s	00h:00m:07s

[Carrington]

Another way to think:

Scope’s effective blind-time % = 100 x (1 – UW)
U = Scope’s measured update rate
W = Display acquisition window = Timebase setting x number of divisions

Example, at a 100ns/div timebase setting and 14 divisions:
  - 100K wfrm/s -> effective blind-time of 86.0%, i.e. it's showing 14% of the real-time waveform.
  - 50K wfrm/s -> effective blind-time is 92.5%, i.e. it's showing 7.5% of the real-time waveform.


http://cdn.rohde-schwarz.com/dl_downloads/dl_application/application_notes/1er02/1ER02_1e.pdf
http://cp.literature.agilent.com/litweb/pdf/5989-7885EN.pdf

[marmad]

Because with the series SDS2000 may happen as with the GDS2000 series:

The numbers you circled in red (for the GDS2000) may not be accurate: I got them from grego's video he shot while cycling quickly through the time base settings and reading the Trigger Out with his Picoscope. As far as I know, no one has yet done an accurate, controlled measurement of the GDS2000 wfrm/s rates at every time base.

But one number that IS certainly accurate - and it's a little disingenuous that Instek doesn't mention this in their specs - is the fact that the GDS2000 only achieves it's top-rated ~80k wfrm/s when sampling at 200MSa/s (instead of the full 2GSa/s).

[Carrington]

That's my point. Why hide it? 
If finally will we know it.

[marmad]

Example, at a 100ns/div timebase setting and 14 divisions:
  - 100K wfrm/s -> blind-time of 86.0%, i.e. it's showing 14% of the real-time waveform.
  - 50K wfrm/s -> blind-time is 92.5%, i.e. it's showing 7.5% of the real-time waveform.

You mean "effective" blind times - as opposed to "real" blind times (if the acquisition time is larger than the display window - i.e. 100ns x 14 in your example). The Agilent formula only matches the Rohde-Schwarz formula if you use the actual acquisition time (not the display window).

[Carrington]

Example, at a 100ns/div timebase setting and 14 divisions:
  - 100K wfrm/s -> blind-time of 86.0%, i.e. it's showing 14% of the real-time waveform.
  - 50K wfrm/s -> blind-time is 92.5%, i.e. it's showing 7.5% of the real-time waveform.

You mean "effective" blind times - as opposed to "real" blind times (if the acquisition time is larger than the display window - i.e. 100ns x 14 in your example). The Agilent formula only matches the Rohde-Schwarz formula if you use the actual acquisition time (not the display window).

Yeap, "effective".

[Hydrawerk]

Looking through the SDS2000 datasheet:
(...) I see absolutely no mention of segmented memory - which is (combined with the segment analysis tool), IMO, one of the best features of the DS2000 series.

I use the 56M of deep memory most often for recording segments, and if the SDS2000 is missing that feature, I'd never consider it as a viable purchase after having had the feature.

Yes, nothing in the datasheet. But in this table from rf-loop
they say that SDS2000 has a History function. It might be similar to segmented memory. In fact, Rigol does not use the name "Segmented memory", they call it "Record" instead.
EDIT: Table was made by Siglent (and shown by rf-loop.)

[marmad]

Yes, nothing in the datasheet. But in this table from rf-loop they say that SDS2000 has a History function. It might be similar to segmented memory. In fact, Rigol does not use the name "Segmented memory", they call it "Record" instead.

The table is wrong about some things, including the triggers on the DS2000 - so I wouldn't really trust it. The datasheet is supposed to list the full capabilities of the device - no mention of segments or recording or even a History function - so I would be surprised if it does it in any serious fashion. Compare it to the datasheet of the DS2000.