Author Topic: Siglent's new product- SDS2000X Series  (Read 26375 times)

0 Members and 1 Guest are viewing this topic.

Offline pascal_sweden

  • Super Contributor
  • ***
  • Posts: 1397
  • Country: no
Re: Siglent's new product- SDS2000X Series
« Reply #75 on: July 18, 2016, 02:34:49 am »
Thanks for clarifying!

I must admit that the GW-Instek implementation looks really nice! If I only had such an oscilloscope! :)

Hopefully Rigol and Siglent will implement their protocol decoding in the same way in the future.

Both Rigol R&D and Siglent R&D, could just buy a GW-Instek GDS2104E oscilloscope, to verify and test the current correct implementation from GW-Instek, and adapt their own incorrect implementation accordingly :)
« Last Edit: July 18, 2016, 03:16:10 am by pascal_sweden »
 

Online nctnico

  • Super Contributor
  • ***
  • Posts: 15469
  • Country: nl
    • NCT Developments
Re: Siglent's new product- SDS2000X Series
« Reply #76 on: July 18, 2016, 03:23:20 am »
Don't count on it. Because the current Rigol and Siglent scopes lack the processing power they will need to do decoding in the FPGA which might not have the resources and even if the resources are available implementing the decoding will take a significant amount of FPGA development work.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline pascal_sweden

  • Super Contributor
  • ***
  • Posts: 1397
  • Country: no
Re: Siglent's new product- SDS2000X Series
« Reply #77 on: July 18, 2016, 03:39:48 am »
Then I have to wait for another DSO manufacturer that launches a Zynq-7000 based bench oscilloscope :)

No matter how good the GW-Instek GDS2104E oscilloscope performs, I don't like the physical design,
and I do consider the physical design important, given that the oscilloscope is on the bench all the time :)

Even if the new Siglent SDS2000X series is not based on a Zynq-7000 SoC, it does have a pretty powerful CPU, doesn't it? What processor is it, and what are the specs of it? Maybe that beast can do the same job as the dual ARM Cortex-A9 architecture in the Zynq-7000 SoC?

Or one can only hope that Rigol is secretly working on a new oscilloscope series based on the Zynq-7000 :)
« Last Edit: July 18, 2016, 03:45:57 am by pascal_sweden »
 

Offline tautech

  • Super Contributor
  • ***
  • Posts: 13286
  • Country: nz
  • NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: Siglent's new product- SDS2000X Series
« Reply #78 on: July 18, 2016, 06:40:23 am »


No matter how good the GW-Instek GDS2104E oscilloscope performs, I don't like the physical design,
and I do consider the physical design important, given that the oscilloscope is on the bench all the time :)

Even if the new Siglent SDS2000X series is not based on a Zynq-7000 SoC, it does have a pretty powerful CPU, doesn't it? What processor is it, and what are the specs of it? Maybe that beast can do the same job as the dual ARM Cortex-A9 architecture in the Zynq-7000 SoC?

Daves teardown:
http://www.eevblog.com/forum/blog/eevblog-864-siglent-sds2000x-series-oscilloscope-teardown/
Avid Rabid Hobbyist
 

Online nctnico

  • Super Contributor
  • ***
  • Posts: 15469
  • Country: nl
    • NCT Developments
Re: Siglent's new product- SDS2000X Series
« Reply #79 on: July 18, 2016, 06:54:01 am »
The CPU in the SDS2000X is an AD BF531 which is the 400MHz entry level Blackfin DSP. It is extremely out of place for use as a generic purpose processor. Sure the specs boast impressive performance but you won't get those without hand  optimised assembly code OR a very good C compiler which can make use of the parallellism provided by the CPU core. It also doesn't have an MMU so there is no protection (or detection) against different tasks corrupting eachother.

@Pascal: You can buy the GDS2000E and put something like this (a tea cap) over it when not powered:

 :-DD
« Last Edit: July 18, 2016, 07:18:39 am by nctnico »
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 
The following users thanked this post: Krisztián

Offline rickv14623

  • Regular Contributor
  • *
  • Posts: 199
  • Country: us
Re: Siglent's new product- SDS2000X Series
« Reply #80 on: July 27, 2016, 01:04:18 am »
Found this review from Jack Ganssle today http://www.embedded.com/electronics-blogs/break-points/4442429/Siglent-s-SDS-2000X-oscilloscopes

Looks at SDS2000X next to a Keysight unit.
 

Offline Hydrawerk

  • Super Contributor
  • ***
  • Posts: 2338
  • Country: 00
Re: Siglent's new product- SDS2000X Series
« Reply #81 on: August 01, 2016, 01:58:39 am »
Thank you.
Amazing machines. http://www.youtube.com/user/denha (It is not me...)
 

Offline Hydrawerk

  • Super Contributor
  • ***
  • Posts: 2338
  • Country: 00
Re: Siglent's new product- SDS2000X Series
« Reply #82 on: August 01, 2016, 02:14:13 am »
Quote
One of the great things about digital scopes is the deep memory. With 140 Mpts there’s a lot of data captured.
but
Quote
There is no search capability for finding things in the buffer, which some scopes have.
That is not good.   :-- :--
http://www.embedded.com/electronics-blogs/break-points/4442429/Siglent-s-SDS-2000X-oscilloscopes
Amazing machines. http://www.youtube.com/user/denha (It is not me...)
 
The following users thanked this post: Krisztián

Offline pascal_sweden

  • Super Contributor
  • ***
  • Posts: 1397
  • Country: no
Re: Siglent's new product- SDS2000X Series
« Reply #83 on: August 01, 2016, 02:43:10 am »
Does the GW-Instek GDS2000E have this? Can the dutch guy promoting GW-Instek on this forum, make a video about this functionality? :)
 

Offline Hydrawerk

  • Super Contributor
  • ***
  • Posts: 2338
  • Country: 00
Re: Siglent's new product- SDS2000X Series
« Reply #84 on: August 01, 2016, 02:52:23 am »
There is a Search button on GW-Instek GDS2000E. You can look into the user manual.
Amazing machines. http://www.youtube.com/user/denha (It is not me...)
 

Online nctnico

  • Super Contributor
  • ***
  • Posts: 15469
  • Country: nl
    • NCT Developments
Re: Siglent's new product- SDS2000X Series
« Reply #85 on: August 01, 2016, 03:26:37 am »
There is a Search button on GW-Instek GDS2000E. You can look into the user manual.
And there it says it can search for triggers (edge, runt, bus decoding), signal peaks or FFT peaks.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline radar_macgyver

  • Frequent Contributor
  • **
  • Posts: 334
  • Country: us
Re: Siglent's new product- SDS2000X Series
« Reply #86 on: August 01, 2016, 04:10:01 am »
Another serial decoder bug to report, this time on the I2C decoder.

I'm using an STM32F469 to read data from a TH02 temperature sensor (HopeRF). I have the analog channels 1 and 2 hooked up to SDA and SCL, respectively. First, I noticed that, unlike for the serial decoder, I can't make an acquisition and then zoom in on the region of interest. When attempting to do so, the decoded results don't follow the zoom change.

To illustrate, the first attachment is a capture of a single I2C transaction. The micro's software does the transaction only once, so the scope is not re-triggered. Triggering is set to 'normal'. The decode looks fine. I then use the horizontal position to move the traces. Note that the decode results don't follow. I can also change the horizontal timebase, and the decoded results still don't match up.

I also noticed that if I change the horizontal timebase and make the micro do another I2C transfer, the scope captures it and displays it on the traces, but the bus decode does not change. If I do another transaction without touching the controls of the scope, it does the decode. Subsequent transfers are captured and decoded without error. It's as if the first transaction after changing the horizontal timebase isn't properly decoded. I noticed that this is only true of the horizontal timebase, not the horizontal position. It also happens when changing the memory depth (these images were all recorded at 7k, but the behavior is the same regardless of memory depth). It also behaves identically if I use the digital channels to read in the I2C bus.

From my earlier tests with Serial (UART) decode, I know that the scope handles zoom/pan properly when doing serial decode; it just doesn't seem to do so for I2C. Would be nice if they can roll this in to the next firmware update.
 
The following users thanked this post: tautech

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: Siglent's new product- SDS2000X Series
« Reply #87 on: August 01, 2016, 04:14:30 am »
Don't count on it. Because the current Rigol and Siglent scopes lack the processing power they will need to do decoding in the FPGA which might not have the resources and even if the resources are available implementing the decoding will take a significant amount of FPGA development work.

You don't know what you're talking about. The Rigol DS2000 has been doing decoding on it's entire memory since before your GW Instek GDS2000E was even brought to market. I believe I showed it when I made my comparison video with the Siglent SDS2000 2 years ago.
 

Online nctnico

  • Super Contributor
  • ***
  • Posts: 15469
  • Country: nl
    • NCT Developments
Re: Siglent's new product- SDS2000X Series
« Reply #88 on: August 01, 2016, 04:25:41 am »
@marmad: Can you shift the I2C start condition out of the screen to the left to prove it (ofcourse the start condition must still be in memory but just outside the display range)?
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: Siglent's new product- SDS2000X Series
« Reply #89 on: August 01, 2016, 04:37:24 am »
@marmad: Can you shift the I2C start condition out of the screen to the left to prove it (ofcourse the start condition must still be in memory but just outside the display range)?

The entire waveform is a single capture with the start condition that triggered well outside screen-left. Look at the little graphic at top-center of the screen: the gray/black sine wave shows the part of memory currently being displayed (the black section), with the orange T (trigger point) far to the left of the current part displayed on screen. Here's another image using zoom mode; the trigger is all the way outside of both the zoomed window (black) and unzoomed portions (blue).

« Last Edit: August 01, 2016, 04:39:18 am by marmad »
 

Online nctnico

  • Super Contributor
  • ***
  • Posts: 15469
  • Country: nl
    • NCT Developments
Re: Siglent's new product- SDS2000X Series
« Reply #90 on: August 01, 2016, 04:43:26 am »
Well at least that shows the DS2000 appears to decode the entire memory. That is different from what I've seen from the Rigol DS1000Z and Siglent oscilloscopes so far.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline pascal_sweden

  • Super Contributor
  • ***
  • Posts: 1397
  • Country: no
Re: Siglent's new product- SDS2000X Series
« Reply #91 on: August 01, 2016, 05:44:46 am »
So the Rigol DS2000A, the Owon XDS3000 and the GW-Instek GDS2000E all do decoding in full memory and not only for what is displayed on the screen.

But why can't they make this work on the Rigol DS1000Z in a similar way? Is it just a design decision as such in that firmware, or is it really not feasible from a technical hardware perspective?
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: Siglent's new product- SDS2000X Series
« Reply #92 on: August 01, 2016, 05:53:56 am »
So the Rigol DS2000A, the Owon XDS3000 and the GW-Instek GDS2000E all do decoding in full memory and not only for what is displayed on the screen.

But why can't they make this work on the Rigol DS1000Z in a similar way? Is it just a design decision as such in that firmware, or is it really not feasible from a technical hardware perspective?

Actually, the Owon XDS3000 doesn't do it (I made a mistake and deleted my previous message) - I hadn't tested well enough. It also decodes just what is displayed.

But in terms of the Rigol DS2000(A) versus the Rigol DS1000Z, I'm pretty sure that it's related to what nctnico mentioned earlier: speed. People seem to think that the DS2000A is quite similar to the DS1000Z series - but in fact, in terms of speed, they're like night and day. The DS2000A has a FPGA dedicated for display processing (and extra display memory) that the DS1000Z is lacking, so using the DS1000Z after using the DS2000(A) feels as slow as molasses.
« Last Edit: August 01, 2016, 05:57:15 am by marmad »
 

Offline pascal_sweden

  • Super Contributor
  • ***
  • Posts: 1397
  • Country: no
Re: Siglent's new product- SDS2000X Series
« Reply #93 on: August 01, 2016, 06:11:03 am »
If you buy a new Rigol DS2000A or MSO2000A these days, will it include the Yaigol PLL fix from in the factory? Or does it still include the PLL hardware issue, which is masked with a semi workaround firmware?
 

Offline Hydrawerk

  • Super Contributor
  • ***
  • Posts: 2338
  • Country: 00
Re: Siglent's new product- SDS2000X Series
« Reply #94 on: August 01, 2016, 06:18:58 am »
So the Rigol DS2000A, the Owon XDS3000 and the GW-Instek GDS2000E all do decoding in full memory and not only for what is displayed on the screen.
What about Rigol DS4000?
Amazing machines. http://www.youtube.com/user/denha (It is not me...)
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: Siglent's new product- SDS2000X Series
« Reply #95 on: August 01, 2016, 06:32:36 am »
Here is a very clear and easy-to-understand illustration of the kind of problems nctnico has been trying to explain which can result from only decoding display memory:

A single I2C READ of 19 consecutive bytes of data.

There are two images for each DSO - one with the I2C READ command at the far left of the respective display area (50h), and the second one with the I2C READ command just beyond the left edge of the screen.

You can see that the Owon (which is only decoding display memory) loses the ability to decode any part of the string as soon as it loses the READ command from display memory (since it doesn't know what the data is in reference to).










« Last Edit: August 01, 2016, 06:36:18 am by marmad »
 

Offline pascal_sweden

  • Super Contributor
  • ***
  • Posts: 1397
  • Country: no
Re: Siglent's new product- SDS2000X Series
« Reply #96 on: August 01, 2016, 06:43:36 am »
Regarding the Serial Decode option. Does the Rigol DS2000A support 9-bit serial data?

Note that LeCroy does support this:
http://teledynelecroy.com/options/productseries.aspx?mseries=261&groupid=88

So if Wuerstchenhund reads this, he is going to be a happy man! :)
« Last Edit: August 01, 2016, 06:56:44 am by pascal_sweden »
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: Siglent's new product- SDS2000X Series
« Reply #97 on: August 01, 2016, 06:59:18 am »
Regarding the Serial Decode option. Does the Rigol DS2000A support 9-bit serial data?

Isn't that something you can find out from the online manual? Obviously, this discussion of decoding on other DSOs is slightly off-topic to the thread - but at least it was related to a specific question about the capabilities of the SDS2000X series. But asking other questions about other DSOs is not, so let's leave those for other threads, ok?
 

Offline pascal_sweden

  • Super Contributor
  • ***
  • Posts: 1397
  • Country: no
Re: Siglent's new product- SDS2000X Series
« Reply #98 on: August 01, 2016, 07:06:22 am »
Okey, fair enough. I have checked in the manual of Rigol DS2000A and it does support 9 bit data!
Strangely enough, it is not supported on the Rigol DS1000Z, unless it is a typo in the User's Guide.

Rigol DS2000A:
Press Data Bits to set the data width of each frame. It can be set to 5, 6, 7, 8 or 9 and the default is 8.

Rigol DS1000Z:
Press Data to set the data width of each frame. It can be set to 5, 6, 7 or 8 and the default is 8.

Maybe now one can confirm here if the same is true for the Siglent SDS2000X series! :)
« Last Edit: August 01, 2016, 07:22:05 am by pascal_sweden »
 

Offline tautech

  • Super Contributor
  • ***
  • Posts: 13286
  • Country: nz
  • NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: Siglent's new product- SDS2000X Series
« Reply #99 on: August 01, 2016, 07:17:17 am »
Regarding the Serial Decode option. Does the Rigol DS2000A support 9-bit serial data?

The latest manual suggest the SDS2kX supports 7 and 10 bit but the manual is older than the latest FW, so I'd need to check a unit to be sure.............. :popcorn:

Edit
Only 7 or 10 bit with current FW 38.7
« Last Edit: August 01, 2016, 07:28:31 am by tautech »
Avid Rabid Hobbyist
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf