Author Topic: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators  (Read 28820 times)

0 Members and 1 Guest are viewing this topic.

Offline EEVblog

  • Administrator
  • *****
  • Posts: 30642
  • Country: au
    • EEVblog
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #25 on: September 19, 2015, 12:53:26 am »
BTW, video uploading now. Almost an hour long.
I found a few firmware issues with it, but the price is very compelling, and otherwise it was pretty good.
I think the main drawback is the 1GS/s sample rate. Not good enough when two channels are enabled on the 200MHz model.
Got to be the cheapest way to get a 500uV/div scope as well.
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 18581
  • Country: nl
    • NCT Developments
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #26 on: September 19, 2015, 01:04:53 am »
BTW, video uploading now. Almost an hour long.
I found a few firmware issues with it, but the price is very compelling, and otherwise it was pretty good.
I think the main drawback is the 1GS/s sample rate. Not good enough when two channels are enabled on the 200MHz model.
Why not? 200MHz sampled at 1GS/s is still 5 points per period which is enough to reproduce the signal with light interpolation.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 17083
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #27 on: September 19, 2015, 02:16:32 am »
BTW, video uploading now. Almost an hour long.
I found a few firmware issues with it, but the price is very compelling, and otherwise it was pretty good.
I think the main drawback is the 1GS/s sample rate. Not good enough when two channels are enabled on the 200MHz model.
Got to be the cheapest way to get a 500uV/div scope as well.
Link please Dave.
Avid Rabid Hobbyist
 

Offline Berni

  • Super Contributor
  • ***
  • Posts: 2672
  • Country: si
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #28 on: September 19, 2015, 07:41:08 am »
Why not? 200MHz sampled at 1GS/s is still 5 points per period which is enough to reproduce the signal with light interpolation.

Having both channels enabled drops down the sample rate by half. Those 2.5 points per period is indeed not enough.
 

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 3143
  • Country: cn
  • Born with DLL21 in hand
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #29 on: September 19, 2015, 07:43:41 am »

1. I found a few firmware issues with it, but the price is very compelling, and otherwise it was pretty good.

2. I think the main drawback is the 1GS/s sample rate. Not good enough when two channels are enabled on the 200MHz model.

3. Got to be the cheapest way to get a 500uV/div scope as well.

1. There is many FW issues (also I have found many)  but most of these are not severe class. But of course also less severe need fix. Some more severe need repair asap. But not at all this nearly catasrof what was case when SDS2000 was first time launched and there was fina├Âlly so much that it was more wise to not fix and fix over fix, they write new whole revision.  I do not believe after looking SDS1000X that this is case. I believe this is level where normal bug fix sequence is enough.
One bug is example in "SPO engine".  (example: CH1 on) If set example to 50ns/div (all fast timebases) and measure wfm/s speed. It may show average (average over one TFT update cycle) 61kwfm/s and burst speed over 70kwfm/s.  Now, if turn example vertical shift or adjust V/div, wfm/s drops markable. If you return vertical settings and signal etc all are agen same, it do not return to original speed at all. If V/div and vertical shift keep changed or not but cycle CH2 on and off it returns normal speed or cycle CH1 off and back on same result, until you agen touch vertical settings. (speed drop is 10-20% or something like this...depends perhaps combination of other settings)
This is just one example of bug what need immediately repair because reason may be hidden in some important place in SPO engine and perhaps somehow connected to some other bug (I have least one supect)

2. True.  500MSa/s Nyq. is 250MHz and 250/200 is 1.25.
It is barely enough in theory IF: There is perfect true Sin(x)/x math and there is perfect "brick wall" corner at 200MHz. There is not so much problems with Nyquist Shannon rule 2 due to fact that there is just one ADC for 500MSa/s (not interleaved what may lead to more or less phase errors between true samples).  But in practice it is not well enough in practice.  Minimum in practicde is somewhere 1.33 - 2.5 For 200MHz sinewave BW 600MSa/s (and- "more better")  and for other waveforms if analog front end before ADC  BW shape is tight brick wall (what it is not in any mid or low end scopes what I know)

I can tell that Siglent Sinx/x is good and it do not destroy true sampled data but not mathematically ideal perfect.  But also it do not make it good for 2 channel full 200MHz.

If look Rigol DS1000Z 100MHz (hacked or original) all 4 channel in use it have 250MSa/s.  Nyquist 125MHz.
125/100 is 1.25. Same as in Siglent with all 2 channels in use with 200MHz.  But here is very poor fake Sin(x)/x.
It can not say it is good 200MHz if both channels are in use simultaneusly. 1.25 is not enough. 1.35 ... 1.5 and more is of course better. But as all, it also depends many things.
It need always tightly keep in mind that if talk "200 MHz or 20MHz" scope it is stated for pure sinewave.
200MHz: IF take 20MHz ideal square wave (it do not exist)  to scope input it have harmonics 60, 100, 140, 180, 220...etc.  9th harmonic is inside and even this is some amount attenuated and 11th go over.
THis is if there is not any limitinmg factor coming from sample rates. When there still exist harmonixs what go near or over nyquist limit and ADC itself see these it start produce aliases.  Corners start wobbling etc.

3. It (AFAIK) is true analog side 500uV/div.   Example in Rigol DS1000Z most low is 5mV/div. After then down to 1mV/div is digital side multiply...   Also Rig DS2000 do not have analog side 500uV/div.

(I have not measured yet Siglent true BW using 500uV/div. If it have true analog 200MHz BW then I can accept its natural noise in this scope class. Also there is amount of "1/f" like noise.  If it is  BW rejected, then it is bit noisy. But still not true bad, as we understand all realities about noise when analog front end is still quite cheap circuit solution. (good state of art or even high-end  analog input stage easy cost over this whole scope price)


« Last Edit: September 19, 2015, 08:14:24 am by rf-loop »
If practice and theory is not equal it tells that used application of theory is wrong or the theory itself is wrong.
-
Harmony OS
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 30642
  • Country: au
    • EEVblog
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #30 on: September 19, 2015, 08:15:01 am »
Why not? 200MHz sampled at 1GS/s is still 5 points per period which is enough to reproduce the signal with light interpolation.

The 1GS/s halves to 500MS/s when both channels are on.
In the video I show why you need 4 times the bandwidth.
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 30642
  • Country: au
    • EEVblog
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #31 on: September 19, 2015, 08:19:27 am »
3. It (AFAIK) is true analog side 500uV/div.   Example in Rigol DS1000Z most low is 5mV/div. After then down to 1mV/div is digital side multiply...   

First time I've heard of that. Looks like true 1mV to me, but I have not deliberately checked.
Got any evidence of this?

Quote
Also Rig DS2000 do not have analog side 500uV/div.

Yes it does, it's one of the big banner features.
 

Offline Bryan

  • Frequent Contributor
  • **
  • Posts: 572
  • Country: ca
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #32 on: September 19, 2015, 08:54:42 am »
BTW, video uploading now. Almost an hour long.
I found a few firmware issues with it, but the price is very compelling, and otherwise it was pretty good.
I think the main drawback is the 1GS/s sample rate. Not good enough when two channels are enabled on the 200MHz model.
Got to be the cheapest way to get a 500uV/div scope as well.

Hmm, checked and no video yet.
-=Bryan=-
 

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 3143
  • Country: cn
  • Born with DLL21 in hand
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #33 on: September 19, 2015, 12:15:54 pm »
SDS2102X 
input set for 50 ohm  Signal from HP E4421B what have, including my cable,  enough level accuracy for this kind of rough preliminary test. (I have not corrected cable attenuation because I do not also know this gen flatness table)

relative level reference 1MHz Level set for 6div pp on the scope screen.
Measured 50mV - 200mV/div  -3dB point is  275MHz or higher.

500uv/div (p-p 6div) -3dB point over 250MHz

(it can easy look it is not made By digital side magnification. Just stop and (horz and or vertical) zoom and look all separate ADC raw points vertical dispersion, distribution)


If practice and theory is not equal it tells that used application of theory is wrong or the theory itself is wrong.
-
Harmony OS
 

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 3143
  • Country: cn
  • Born with DLL21 in hand
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #34 on: September 19, 2015, 12:50:56 pm »
3. It (AFAIK) is true analog side 500uV/div.   Example in Rigol DS1000Z most low is 5mV/div. After then down to 1mV/div is digital side multiply...   

First time I've heard of that. Looks like true 1mV to me, but I have not deliberately checked.
Got any evidence of this?

Quote
Also Rig DS2000 do not have analog side 500uV/div.

Yes it does, it's one of the big banner features.

Banners and facts - sometimes not from same world. (And I do not direct this to Rigol but all, from shoes and trousers  to spectrum analyzers)

AFAIK DS2000 500uV/div is derived from 1mV/div using digital side multiplication but this I can not proof, only heard and also made some opinion based to information and test images on the forum

DS1000Z is different case and this I have on my lab table for occasional less important purposes(DS1074Z) and here more:
https://www.eevblog.com/forum/testgear/rigol-ds1074z-weird-signal-level-problem/msg563208/#msg563208

-First about Rigol's  Sin(x)/x  accidental FW bug or error in design. (perhaps image cosmetics have been more important than more serious things in T&M equipment)
-And then about 1 and 2mV/div things.
If need proof more about vertical level things then I will later disassembly it and do measurements in analog front end before ADC.,
« Last Edit: September 19, 2015, 01:03:35 pm by rf-loop »
If practice and theory is not equal it tells that used application of theory is wrong or the theory itself is wrong.
-
Harmony OS
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 18581
  • Country: nl
    • NCT Developments
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #35 on: September 19, 2015, 01:28:34 pm »
Why not? 200MHz sampled at 1GS/s is still 5 points per period which is enough to reproduce the signal with light interpolation.
The 1GS/s halves to 500MS/s when both channels are on.
In the video I show why you need 4 times the bandwidth.
Aha. On a side note: according to signal theory you'll have all the information of the part of a signal which falls within the Nyquist limit. Its just our eyes which cannot connect the dots! I did a lot of hands on research in this area and found out that a good signal reconstruction algoritm (note I avoid the word interpolation because that suggests adding information which isn't there) can reconstruct signals up to 0.45fs (500MS/s*0.45=225MHz). A condition for this to work is to have good anti-aliasing filtering.
« Last Edit: September 19, 2015, 02:17:43 pm by nctnico »
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline hans

  • Super Contributor
  • ***
  • Posts: 1046
  • Country: nl
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #36 on: September 19, 2015, 03:57:08 pm »
After watching the video I'm also missing Nth edge trigger near the options.

The Rigol has it (optional), but it's very useful to trigger of e.g. SPI bus clocks from a micro that is also doing interrupts etc. If you're debugging data in a long transfer, there is no guarantee the time base will be stable at all.
However I must admit that probably for that kind of work a cheap LA is better than any built-in digital decoding etc.
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 18581
  • Country: nl
    • NCT Developments
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #37 on: September 19, 2015, 04:16:01 pm »
After watching the video I'm also missing Nth edge trigger near the options.

The Rigol has it (optional), but it's very useful to trigger of e.g. SPI bus clocks from a micro that is also doing interrupts etc. If you're debugging data in a long transfer, there is no guarantee the time base will be stable at all.
However I must admit that probably for that kind of work a cheap LA is better than any built-in digital decoding etc.
The trigger hold-off function can also be used to tirgger on the beginning of an SPI transaction.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 17083
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #38 on: September 19, 2015, 07:47:50 pm »
BTW, video uploading now. Almost an hour long.
I found a few firmware issues with it, but the price is very compelling, and otherwise it was pretty good.
I think the main drawback is the 1GS/s sample rate. Not good enough when two channels are enabled on the 200MHz model.
Got to be the cheapest way to get a 500uV/div scope as well.

Hmm, checked and no video yet.
Here it is:

https://www.eevblog.com/forum/blog/eevblog-797-siglent-sds1000x-review/
Avid Rabid Hobbyist
 

Offline hans

  • Super Contributor
  • ***
  • Posts: 1046
  • Country: nl
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #39 on: September 19, 2015, 08:49:41 pm »
After watching the video I'm also missing Nth edge trigger near the options.

The Rigol has it (optional), but it's very useful to trigger of e.g. SPI bus clocks from a micro that is also doing interrupts etc. If you're debugging data in a long transfer, there is no guarantee the time base will be stable at all.
However I must admit that probably for that kind of work a cheap LA is better than any built-in digital decoding etc.
The trigger hold-off function can also be used to tirgger on the beginning of an SPI transaction.

I did not meant to say start of SPI trigger (in that case yes; bursts will be triggered quite well from holdoff), but more exactly Nth clock pulse in a waveform.

E.g. a SPI or I2C bus where the uC master controls it, but also has (uncorrelated) other tasks/interrupts going on that may cause the waveform part you want to see jitter around a lot.
You could manually move the horizontal on each trigger, but even better is to trigger at the exact bit you want to see or monitor.

Arguably triggers are quite important to look at, because there is little use to a scope if you can't trigger on your signal reliably. I don't think Nth edge will be missing for a lot of people, but it was a difference I noticed and personally value a lot.
« Last Edit: September 19, 2015, 08:51:34 pm by hans »
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 17083
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #40 on: September 20, 2015, 07:26:12 am »
Dave, did you get the -S, inbuilt AWG model?

No, just the normal one. It has the connector on the back though.
This might indicate the AWG HW is installed and the -S version as available from the factory is AWG enabled before shipping.  :-//
I wonder why Siglent just didn't offer AWG as a purchasable option as they do with the SDS2000 series AWG option  :-//

Might have to rip the covers off mine next week when it arrives.  >:D
Avid Rabid Hobbyist
 

Offline Wuerstchenhund

  • Super Contributor
  • ***
  • Posts: 2811
  • Country: gb
  • Only sparsely active but available via PM
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #41 on: September 20, 2015, 07:55:56 am »
@tautech

I may be wrong but didn't you mention you also have an SDG200X AWG? If so could you post some scope screenshots showing a 1MHz, 15MHz and 25MHz square wave please?
 

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 3143
  • Country: cn
  • Born with DLL21 in hand
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #42 on: September 21, 2015, 06:50:00 pm »
Here SDS1000X wfm/s table.
More is under work. (sinx/x and vectors on,  fast segment acquire speed and other factors, history buffer things)
All speeds are averages. (not top speeds between single wfm, minimum average time is 2s. With slower time bases much more.

NOTE!  In table, model is accidentally wrong. There read SDS2202X

Correct model number is of course SDS1202X


Tnx for Mark-O about note this error.



« Last Edit: November 16, 2015, 12:18:12 pm by rf-loop »
If practice and theory is not equal it tells that used application of theory is wrong or the theory itself is wrong.
-
Harmony OS
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 17083
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #43 on: September 22, 2015, 06:13:01 am »
Dave, did you get the -S, inbuilt AWG model?

No, just the normal one. It has the connector on the back though.
This might indicate the AWG HW is installed and the -S version as available from the factory is AWG enabled before shipping.  :-//
I wonder why Siglent just didn't offer AWG as a purchasable option as they do with the SDS2000 series AWG option  :-//

Might have to rip the covers off mine next week when it arrives.  >:D
Well I don't have to open her up, as shipped the basic SDS1000X models do have the the AWG option HW and 30 trial AWG usages, just like the SDS2000.  :-+
For simple use the SDS1202X performs just as you'd expect and IMHO is a very nice unit. Just as Dave said these units are quite solid, much heavier than the older SDS1000 series. No doubt there maybe some bugs in these just released units but as yet I've not stumbled across one.  :)

Might post more findings after dinner.

Requests?

@tautech

I may be wrong but didn't you mention you also have an SDG200X AWG? If so could you post some scope screenshots showing a 1MHz, 15MHz and 25MHz square wave please?
You are and I've asked Steve and his guys if they would oblige.
Avid Rabid Hobbyist
 

Offline Timur Born

  • Regular Contributor
  • *
  • Posts: 65
  • Country: de
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #44 on: September 22, 2015, 01:55:09 pm »
With the current price-drops on SDS2000 series units (in EU/Germany), how does the X compete vs. its supposedly bigger brothers?
 

Offline Timur Born

  • Regular Contributor
  • *
  • Posts: 65
  • Country: de
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #45 on: September 22, 2015, 02:00:42 pm »
Dave lists the 1GSa/s sampling rate as the main drawback. Does it limit bandwidth enough to forgo the 200 MHz model and just go for the 100 MHz one (150 EUR price difference before VAT)?
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 30642
  • Country: au
    • EEVblog
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #46 on: September 22, 2015, 02:16:26 pm »
Dave lists the 1GSa/s sampling rate as the main drawback. Does it limit bandwidth enough to forgo the 200 MHz model and just go for the 100 MHz one (150 EUR price difference before VAT)?

If you really need 200MHz real bandwidth on both channels at once, I wouldn't recommend buying this scope. Single channel, no problem.
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 30642
  • Country: au
    • EEVblog
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #47 on: September 22, 2015, 02:18:25 pm »
Well I don't have to open her up, as shipped the basic SDS1000X models do have the the AWG option HW and 30 trial AWG usages, just like the SDS2000.  :-+

DOH!
I just shot a teardown and completely forgot to look for the AWG circuit  :palm:
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 18581
  • Country: nl
    • NCT Developments
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #48 on: September 22, 2015, 04:51:15 pm »
You still got the scope? Any chance to test whether the decode only decodes what is on screen or decodes the entire memory?
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline Siglent America

  • Regular Contributor
  • *
  • Posts: 221
  • Country: us
Re: NEW - Siglent X-Series Oscilloscopes and Arb Waveform Generators
« Reply #49 on: September 22, 2015, 05:46:52 pm »
Dave, did you get the -S, inbuilt AWG model?

No, just the normal one. It has the connector on the back though.
This might indicate the AWG HW is installed and the -S version as available from the factory is AWG enabled before shipping.  :-//
I wonder why Siglent just didn't offer AWG as a purchasable option as they do with the SDS2000 series AWG option  :-//

Might have to rip the covers off mine next week when it arrives.  >:D
Well I don't have to open her up, as shipped the basic SDS1000X models do have the the AWG option HW and 30 trial AWG usages, just like the SDS2000.  :-+
For simple use the SDS1202X performs just as you'd expect and IMHO is a very nice unit. Just as Dave said these units are quite solid, much heavier than the older SDS1000 series. No doubt there maybe some bugs in these just released units but as yet I've not stumbled across one.  :)

Might post more findings after dinner.

Requests?

@tautech

I may be wrong but didn't you mention you also have an SDG200X AWG? If so could you post some scope screenshots showing a 1MHz, 15MHz and 25MHz square wave please?
You are and I've asked Steve and his guys if they would oblige.

Here are screen shots of a SDG2122X outputting a square wave at 1 MHz, 15 MHz and 25 MHz into a SDS2304 (300 MHz) oscilloscope.
Steve
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf