Author Topic: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown  (Read 14228 times)

0 Members and 1 Guest are viewing this topic.

Offline R_Gtx

  • Regular Contributor
  • *
  • Posts: 52
  • Country: gb
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #25 on: November 26, 2017, 05:39:34 pm »
I think Sbus (near USB connector) is there with reasons?

p.6 of the QuickStart guide confirms that the S-Bus, a modified HDMI connector, is for digital inputs:
 

Offline rhb

  • Super Contributor
  • ***
  • Posts: 2677
  • Country: us
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #26 on: November 27, 2017, 03:54:39 am »
Interestingly, the Instek MSO-2204EA exhibits the same double trace behavior as the Siglent when tapping on the BNC in normal mode.
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 16347
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #27 on: November 27, 2017, 04:10:21 am »
Interestingly, the Instek MSO-2204EA exhibits the same double trace behavior as the Siglent when tapping on the BNC in normal mode.
Is it consistent ? Each and every event ?
Avid Rabid Hobbyist
 

Online Brumby

  • Supporter
  • ****
  • Posts: 9402
  • Country: au
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #28 on: November 27, 2017, 05:25:33 am »
Here's a question:
Is the multiple waveform just a temporary artefact that goes away after a time - the specific interest being how long that time can be?

In the Keysight scope the artefact was visible ... but only for a sub second time interval.

Could it be that the Siglent will clean its act up ... but it might be less disciplined and take an unspecified time period to do it?

In this video, we see this occur quickly at times - and then at around the 18:50 mark, it takes around 7 seconds.  Perhaps the times it "didn't" were simply from Dave not waiting long enough.

I'm not excusing the behaviour - just looking at a possibility.
 

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 3071
  • Country: cn
  • Born with DLL21 in hand
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #29 on: November 27, 2017, 07:25:39 am »
Here's a question:
Is the multiple waveform just a temporary artefact that goes away after a time - the specific interest being how long that time can be?

In the Keysight scope the artefact was visible ... but only for a sub second time interval.

Could it be that the Siglent will clean its act up ... but it might be less disciplined and take an unspecified time period to do it?

In this video, we see this occur quickly at times - and then at around the 18:50 mark, it takes around 7 seconds.  Perhaps the times it "didn't" were simply from Dave not waiting long enough.

I'm not excusing the behaviour - just looking at a possibility.

In video there is also one mysterious.

When it have multiple waveforms captured (of course, this is mostly not one trigger event) and display these on the screen, it display continuously many seconds when Dave talk.  Then it suddenly "clean" display and only one wfm is displayed. Just at this moment there can see that   :bullshit: Trigger indicator blinks shortly. This is unexpected. If look carefully this short blink can see. What it is? Why?

I have tested this now many ways with SDS1104X-E  brother SDS1202X-E.
In normal mode it leave all these wafeforms continuously on the screen what are captured and displayed in last display frame. What are in previous display frame, they disappear (of course but human eye fading bit slowly). After then if no anymore trigger events it keep these on the screen continuously and infinite. Until trigger event or user hit example run/stop what leave last one on the screen.

A.
If user want first trigger event single waveform displayed and then stop. Trigger mode Single.

B.
If user want last event(s) there is two way and trigger mode Normal in use.

 - B1.  Acg mode fast.
It leave after last trigger event, this last frame visible independent of if it have single or multiple events. (bug or not but this it do and I like it do this - because it have more tools for look details)

- B2.  Acg mode slow.
It capture only single trigger event to one display frame. After last trig, it display this last frame where is last captured waveform. If user want this, just use this.

If it is too complex, on the markets can find more simple scopes.
But I like that scope is slave and I am mastrer who select how scope do. Not vice versa.

If there is multiple frames captured after acquistion started (fast or slow) and trig Normal mode. Last frame is displayed if no anymore trigger events.
User can stop oscilloscope pressing Run/stop or History. At this point it clean and leave last wfm to display.

If it was fast mode and last frame and previous frames have single or multiple waveforms these every waveform can look separately using history view. (up to max buffer depth - depending settings)
This is because waveform history buffer is not "frame recorder". It record every single waveform, idependent of how many waveform is in run mode overlayed in one display frame. Also every waveform have individual time stamp.






« Last Edit: November 27, 2017, 07:44:22 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
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 16347
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
« Last Edit: November 27, 2017, 03:41:50 pm by tautech »
Avid Rabid Hobbyist
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 18010
  • Country: nl
    • NCT Developments
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #31 on: November 27, 2017, 11:39:04 am »
Interestingly, the Instek MSO-2204EA exhibits the same double trace behavior as the Siglent when tapping on the BNC in normal mode.
Disable persistence and the effect is gone!
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 10196
  • Country: 00
Re: EEVblog #1044 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #32 on: November 27, 2017, 11:46:26 am »
Dave: Are we going to see individual reviews of the new bottom tier 4 channel scopes
I think that's likely.
Shootouts are big jobs, but separate videos and shootout is even more work.

Are there enough "new bottom tier 4 channel scopes" for a shootout? I thought there was only one.

Plus: At 40% more money+three years more recent then the current champion (Rigol DS1054Z) it would only be news if it lost the shootout.

A few comparisons is OK but a video review + bug hunt of this new Siglent seems more appropriate.
« Last Edit: November 27, 2017, 11:53:52 am by Fungus »
 

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 3071
  • Country: cn
  • Born with DLL21 in hand
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #33 on: November 27, 2017, 12:21:08 pm »
Interestingly, the Instek MSO-2204EA exhibits the same double trace behavior as the Siglent when tapping on the BNC in normal mode.
Disable persistence and the effect is gone!

If this is just because persistence on why it do not fade out during persistence time when it ageing exept if persistence time is set for infinite. Or is there bug in persistence system?
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 Fungus

  • Super Contributor
  • ***
  • Posts: 10196
  • Country: 00
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #34 on: November 27, 2017, 02:16:40 pm »
If it's "persistence" then all traces should disappear after a time, n'est pas?

 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 18010
  • Country: nl
    • NCT Developments
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #35 on: November 27, 2017, 02:56:58 pm »
Interestingly, the Instek MSO-2204EA exhibits the same double trace behavior as the Siglent when tapping on the BNC in normal mode.
Disable persistence and the effect is gone!
If this is just because persistence on why it do not fade out during persistence time when it ageing exept if persistence time is set for infinite. Or is there bug in persistence system?
Off-topic: I think this is a bug in the GW Instek (already reported so let's see how quickly they fix it). If more triggers arrive then the old traces will fade away but when no new triggers arrive the old traces don't fade away.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2272
  • Country: dk
  • Country: Broken Britain
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #36 on: November 27, 2017, 04:10:07 pm »
I've done lots of tapping of sensors to test their response, and a multiple bouncing first contact is the norm even with very hard surfaces. So I'm sure a bouncy screwdriver making contact with a plastic mounted BNC bounces and gives multiple triggers.

If using something like a small screwdriver, holding it gently near its middle while swiping it passed the BNC should allow it to bounce away after its first contact.

Slowing the scope to 10ms/div will probably show the multiple triggers are just screwdriver bounces!
CML+  That took much longer than I thought it would.
 

Offline nidlaX

  • Frequent Contributor
  • **
  • Posts: 649
  • Country: us
Re: EEVblog #1044 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #37 on: November 27, 2017, 10:17:47 pm »
Are there enough "new bottom tier 4 channel scopes" for a shootout? I thought there was only one.

Plus: At 40% more money+three years more recent then the current champion (Rigol DS1054Z) it would only be news if it lost the shootout.

A few comparisons is OK but a video review + bug hunt of this new Siglent seems more appropriate.
There's enough demand to warrant a shootout between the Rigol DS1054Z, GW Instek GDS-1104B, Uni-T UPO2104CS, Micsig TO1104, and Siglent SDS1104X-E. You could even throw in the shitty 4 channel Hantek bench scope along with the Owon and Hantek 4 channel PC based scopes.
 
The following users thanked this post: jgalak

Offline lukier

  • Supporter
  • ****
  • Posts: 615
  • Country: gb
    • Homepage
Re: EEVblog #1044 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #38 on: November 27, 2017, 10:50:05 pm »
There's enough demand to warrant a shootout between the Rigol DS1054Z, GW Instek GDS-1104B, Uni-T UPO2104CS, Micsig TO1104, and Siglent SDS1104X-E. You could even throw in the shitty 4 channel Hantek bench scope along with the Owon and Hantek 4 channel PC based scopes.

That would be great, but I'm not sure if Dave has Micsig TO1000 series one, AFAIR some older one he got ages ago.
 

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 3071
  • Country: cn
  • Born with DLL21 in hand
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #39 on: November 28, 2017, 06:20:41 pm »
@ EEVblog  (Dave)  one question

My previous message:
In video there is also one mysterious.

When it have multiple waveforms captured (of course, this is mostly not one trigger event) and display these on the screen, it display continuously many seconds when Dave talk.  Then it suddenly "clean" display and only one wfm is displayed.

When I testing SDS1104X-E today I find something "strange".

When you make this video and after then time is in video around 18:57 (time when it suddenly clean image and leave last one) after start: How long time scope was continuously on before this 18:57 time in video. Over around 30min or less?
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: 3071
  • Country: cn
  • Born with DLL21 in hand
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #40 on: November 29, 2017, 06:27:18 am »
@ EEVblog  (Dave)  one question

My previous message:
In video there is also one mysterious.

When it have multiple waveforms captured (of course, this is mostly not one trigger event) and display these on the screen, it display continuously many seconds when Dave talk.  Then it suddenly "clean" display and only one wfm is displayed.

When I testing SDS1104X-E today I find something "strange".

When you make this video and after then time is in video around 18:57 (time when it suddenly clean image and leave last one) after start: How long time scope was continuously on before this 18:57 time in video. Over around 30min or less?

Without Dave's answer.
My strong supposition:
In video, before time position 18:57 scope was not continuously on over 30 minutes.

It is explained here. This is my opinion based with my test result, not officially confirmed by Siglent.
https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/msg1361981/#msg1361981
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 David Hess

  • Super Contributor
  • ***
  • Posts: 10044
  • Country: us
  • DavidH
Re: EEVblog #1044 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #41 on: November 30, 2017, 05:32:43 pm »
What's with the 4 channels? I usually find 2 enough.

Since on these oscilloscopes triggering occurs after digitizing the signal at the full resolution and sample rate, (1) the difference between a full vertical input and a trigger input is only input signal conditioning (2) and memory.  If the signal conditioning only includes a single high input impedance attenuator which is common now, then the signal conditioning is likely identical also.  So the change to support a full vertical input instead of a trigger input is likely only the memory to support it.

Oscilloscopes which use analog triggering have a greater difference between their vertical and trigger inputs so the incremental cost to implement 4 channels is greater.

(1) Run a test on an cheap DSO by triggering from an MSO input while displaying the vertical input of the same signal to see why this is.

(2) With only one high impedance attenuator and no input ground coupling or precharge, the signal conditioning of a cheap DSO vertical input shares more in common with a trigger input than a vertical input of past oscilloscopes.
 

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 3071
  • Country: cn
  • Born with DLL21 in hand
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #42 on: November 30, 2017, 06:32:40 pm »

Appears it "ARM"s itself every 15 seconds, and that refreshes the screen.  Bizarre.


At this time (this may change in later FW) it works like this (simplified):

If all channels have 5mV/div or more it do not "refresh" the screen after 15 seconds.

If one or more channels use 500uV/div - 2mV/div 
And if oscilloscope have powered less than 30 minute it use automatic front end/ADC self adjust (tiny selfcal) every 15 seconds because it do not have reached thermal equilibrium. It do not detect this situation, it is just blind  30min timer.

After 30 minutes continuous power on this process ends.

I do not know why this is not at this time under user control si that user can select if it use or not this autoadjust.  In some oscilloscopes what have autoadjust there is also selction if user enable or disable it so that it do not disturb some measurement  work. But here is not selection now.


In normal mode  it show last display frame if not anymore trigger events.
If it is in fast acquisition mode, depending signal there may be more than one waveform overlayed in one display frame.
If want be sure there only one last wavefom, user can select acquistion mode slow. It capture only one waveform to one display screen.
If want be sure it stop and display first  trigged wafeform user select trigger mode Single, no matter if acquisition mode is slow or fast.

If trigger mode was Normal and acquisition mode fast and if there exist multiple waveform in last display frame and user want see then only last, he can push stop or history button.  After then he can also look in waveform history buffer and look and analyze (using measurements, FFT, cursors) these separate waveform in time order. In history buffer they have also time stamps.

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 Dwaine

  • Frequent Contributor
  • **
  • Posts: 265
  • Country: ca
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #43 on: December 01, 2017, 12:42:10 am »
Image Name:   Linux-3.19.0-xilinx-svn7597

Why do these manufactures always use old dev/build environments?  What a shame.   At least 4.x.  The 3.19 kernel is EOL.
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 16347
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #44 on: December 01, 2017, 02:04:47 am »
Image Name:   Linux-3.19.0-xilinx-svn7597

Why do these manufactures always use old dev/build environments?  What a shame.   At least 4.x.  The 3.19 kernel is EOL.
What's to say it can't be updated later in a firmware release ?
Possible or not ?
Avid Rabid Hobbyist
 

Offline nidlaX

  • Frequent Contributor
  • **
  • Posts: 649
  • Country: us
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #45 on: December 01, 2017, 02:34:23 am »
Image Name:   Linux-3.19.0-xilinx-svn7597

Why do these manufactures always use old dev/build environments?  What a shame.   At least 4.x.  The 3.19 kernel is EOL.
What's to say it can't be updated later in a firmware release ?
Possible or not ?
Test equipment manufacturers are near to last in terms of keeping their products updated and secure. LAN and WiFi connected test equipment are probably great attack vectors for industrial espionage.
 

Offline madires

  • Super Contributor
  • ***
  • Posts: 4832
  • Country: de
  • A qualified hobbyist ;)
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #46 on: December 01, 2017, 10:32:28 am »
I'd consider any T&M devices insecure by default. If you have to protect your R&D secrets design your network accordingly.
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 10196
  • Country: 00
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #47 on: December 01, 2017, 10:53:42 am »
Image Name:   Linux-3.19.0-xilinx-svn7597

Why do these manufactures always use old dev/build environments?  What a shame.   At least 4.x.  The 3.19 kernel is EOL.

What advantages would 4.x bring to an oscilloscope?
 

Offline SaKhan

  • Contributor
  • Posts: 14
  • Country: ch
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #48 on: December 01, 2017, 11:06:46 am »
I guess not many. Siglent as many other vendors probably backport the necessary bug fixes and features from the newer kernels in order to have a stable and established ecosystem. There are so many changes between 3.19 and 4.x that unless we have a config from their production kernel it's really hard to tell.
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 10196
  • Country: 00
Re: EEVblog #1042 - Siglent's $499 SDS1104X-E 4CH Oscilloscope Teardown
« Reply #49 on: December 01, 2017, 01:09:40 pm »
I guess not many. Siglent as many other vendors probably backport the necessary bug fixes and features from the newer kernels in order to have a stable and established ecosystem. There are so many changes between 3.19 and 4.x that unless we have a config from their production kernel it's really hard to tell.

How many of the changes would even be applicable in this context:

https://kernelnewbies.org/Linux_4.0

If it ain't broke, don't fix it.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf