Siglent Technical Support join in EEVblog

[markus_jlrb]

Dears,

by the way,

is the labeled part of the attached tracking plot

discontinued due to a calibration lookup error ?

If so how I could correct it?

Measurment to do this is available but where inside
the OS have I correct it?

Thanks Markus

 

[rf-loop]

After this all, I will kindly recommend that you now delete all things what you think about how to use this spectrum analyzer.  Then after self reset. Start from empty table. Start carefully lessons study and exercises how to use this spectrum analyzer. Lessons aafter lessons start from bottom to top, step by step.
It looks like there is now so many things that this course is not easy using only forum.
Sorry if I feel arrogant but with my poor english I can not use all polite sentences.

Start from lesson about what are separate traces and how they work and how to use settings and what are these all settings.
If there is some bug-error it is always possible but with your examples I can not find but only normal working equipment (and possible operator errors or you did and set still something what you did not tell)

Example here you wonder how trace A disappear in your third image. Of course it disappear, It is broken if it do not this.


Then you wonder this step what small step can see in trace. Of course there are steps. TG Output level is adjusted  by steps and these steps are in factory produced TG output level calibration table inside system. It need dosome steps for compensate level errors over freq band and steps are not there very small. These all are in TG out cal data table but please do not touch them (yes if you go inside you can even manually edit these tables)  if you do not really know how it works and if you do not have right test instruments etc. stay away from factory cal data.
You perhaps have somehow missed normalize function what you really need do every time first for your reference setup (external test setup what you use)  Normalization is function what you really need know and understand what it do and also how to use it right way.
This IS very important function. After you know how traces work and how to handle and set these trace things then before you start anything more with TG, you need be familiar with Normalize function. Also it need know how to use it and what you can not do after normalization without need do it again. 

[markus_jlrb]

Dear "rf-loop"

to be kind and to give you a proper answer
I would like to know more precise what you meant with "After this all"
to follow your thoughts.

My first contact with SA's is dated to 1981 and I believe to have some
general (mayby not an expert) knowledge in SA technique.

So please help me to follow you and your comment.

Many thanks in advance for your kind response.

By the way my opinion is that the SA FW has a geaphical
but not a measurment feature that cause the trace to dissapear.

Markus

[rf-loop]

Dear "rf-loop"

to be kind and to give you a proper answer
I would like to know more precise what you meant with "After this all"
to follow your thoughts.

My first contact with SA's is dated to 1981 and I believe to have some
general (mayby not an expert) knowledge in SA technique.

So please help me to follow you and your comment.

Many thanks in advance for your kind response.

By the way my opinion is that the SA FW has a geaphical
but not a measurment feature that cause the trace to dissapear.

Markus

Also I have first contact with SA but around it was 70 century.

It helps of course to understand some things. But, it do not give lot knowledge how to do operations with example tis Siglent.
There need learn details how to operate Siglent and what it do in different situátions with different settings etc. History experience do not teach these things.

Now may I ask why you wonder trace A is deleted in this image. (yes now I can see you have explained it bit in your original message "But what I found today is that Trace A is shifted/hidden behind Trace B and if I press
Blank softkey the trace B will be deleted and just behind trace B that overlaped trace A
the trace A occure again but at the positionof trace B.")

A: action / P: picture
A1: TG=0dBm / scale=1db /  ref=3dBm / Sweep single / Trace button pressed / trace A selected / Clear Write
P1: Trace A was sweeped and frozen on the screen.

Yes frozen because not new sweep. But it is still not fixed to screen and late new sweep overwrite it.
What to do. As you see trace A is still in normal write mode. At this point, before go to next step, you need stop trace A writing so that it do not follow data anymore. --» Select trace A mode: View. After then it is fixed and it do not "listen" new data and (new sweep) do not overwrite it. If you now change level level things, fixed trace A also follow, If you change horizontal, fixed trace A do not re scale. It is not fixed to freq axis. IN some cases it is useful but some cases it is more useful if it do scaling and is fixed to freq axis, as it is fixed to level axis.

So, I mean, study and exercise Siglent opertator things. "Know your equipment".  Of course it do not mean that need learn what is spectrum analyzing and analyzers. (exept that after analog sweeping SA with analog filter IF need understand how differently full digital IF works and affect things in some situations. Good to know what all is meaning of samping "bucket". THis SA do not exactly sweep! It is hopping. IN some cases this may be very important to know before think that there is bug. But this have nothing now to do with this your example.

Overall this is good example.

So, how ever you do but if I want look first some circuit with TG - SA and want keep this trace for example my own needings as some kind of reference... and after then I want change circuit between TG - SA and look how it differ from first.

Connect circuit and do sweep or sweeps using traces A and B or just A...  then freeze trace A to display selecting trace A mode View.

Make changes in circuit and continue sweeping with trace B writing. Now you can follow how it looks like related to A example when you do adjustments.

Also it need realize how important is Normalization function

A2: trace B selected / inserted DUT between TG and SA
P2: Trace A label changed to Trace B only

Just as it works and I can not see any problem in image. There is only one menu what need display what trace settings you are doing. (If you mean this menu label).
But also this make me wondering why and what you think is wrong? So, for undestanding better I will do this operation as you did and with thinking-quessing what you perhaps want or expect to see with these 3 steps. So I handle this later after I quess what you want see.




Then this

A3: Clear Write selected
P3: while trace B is printed trace A will be deleted (was my first impression)

Here happend just how it works. In this sweep both traces A and B write same data to screen. Both traces are in Clear write mode. Just as your previous images tell. Also trace A indicator in panel left side tell it is not freezed to View mode. 

Traces fixed dislay overlay priority is: Starting from bottom: A, B, C and D. If all traces are on and writing with equal settings there can see only D.


But there is also one detail in your explanation what I need retest when I have time to use SSA before I say anything about it.

[markus_jlrb]

Dear "rf-loop"

many thanks for your detailed explanation.

What I was aware was that the TG is sweeped by steps (hopping) and a fft is done for the pice of time domain alias spectrum.
As well as i was aware of the normalization needed for qualified and precise  measurements - but this was not needed for demonstration.
What I was not awere was the fact that it is necessary too push the view button in a single sweep mode.

During my first steps with this SSA I always pushed the view button to froze the previous trace but this was always a bit tedious to do it
every time. so I switched to the single sweep mode.

So now I understand that it is necessary always to activate the view mode to hold the trace inside the trace memory.
But from my understanding this should be done automatically if I switch to a different trace - hopefully you admit.

As you mentioned before the normalization my way of thinking was a bit different - sorry for that.
I plot the first trace without the DUT (TG/SA connected together) and then I change to the next trace and connected the DUT between.
After the plot of the next trace I changed to the TRACE/MATH function and done the calculation of the difference to see the result.
I have to mentioned that I miss a combination of X+-Y+-<Constant> only X+-Y+-REF is available now but this will be perhaps a
feature in the future - thanks for that in advance.

Markus


 
 
 

[rf-loop]

Dear "rf-loop"

many thanks for your detailed explanation.

What I was aware was that the TG is sweeped by steps (hopping) and a fft is done for the pice of time domain alias spectrum.
As well as i was aware of the normalization needed for qualified and precise  measurements - but this was not needed for demonstration.
What I was not awere was the fact that it is necessary too push the view button in a single sweep mode.

During my first steps with this SSA I always pushed the view button to froze the previous trace but this was always a bit tedious to do it
every time. so I switched to the single sweep mode.

So now I understand that it is necessary always to activate the view mode to hold the trace inside the trace memory.
But from my understanding this should be done automatically if I switch to a different trace - hopefully you admit.

As you mentioned before the normalization my way of thinking was a bit different - sorry for that.
I plot the first trace without the DUT (TG/SA connected together) and then I change to the next trace and connected the DUT between.
After the plot of the next trace I changed to the TRACE/MATH function and done the calculation of the difference to see the result.Now you see trace B show this filter shape and you can compare these and example do adjustments so they match. Or run some other procedures and measurements.

Also when use spectrum (also using TG) it can run so that traces are connected to different detectors etc. 
Example trace A detector pos peak and draw max hold. Trace B detector negative peaks and draw min hold.
Trace C normal and clear write and trace D detector video average and clear write (D not with TG when Normalize, it is reserved)

With TG in use or not. THis is fast freq hopping system. It do not sweep like analog SA at all. So, also detectors need understand carefully and understand why wrong detector for signal under test may give totally bad result. Using
I have to mentioned that I miss a combination of X+-Y+-<Constant> only X+-Y+-REF is available now but this will be perhaps a
feature in the future - thanks for that in advance.

Markus

Yes, it is important to freeze trace using view. Also after single sweep.  Of course it can do automatically if they program it for this but how then if user do not want freeze it when he use multi trace for things what he want. Example some may want use multi trace so that every trace have is "connected" to different detector. In what circumstances SSA need do "auto freeze" and when user want it do not. It is more easy that operator is master and SSA is sub. It is perhaps whole mess for many different users if it do this kind of things automatically. There is so many different applications for what we use SSA.

Usually in simple case example with VSWR bridge. First connect it between TG - SA. Insert standard load to bridge port.
After then do normalization.
Now there is straight line. If want, user can set display line to this level if it is more nice for view.

After then this standard load out from bridge port (or cable end) and connect DUT (example antenna)

After then just run continuous sweeping and watch screen when do adjustments in antenna itself. All time can see difference to normalized level.

In some other case, if have example two filter what want compare or adjust they match with each others by simple meaning.

Do setup for filters connection between TA - SA  so that you can normalize this whole setup but just without filter using connector in place of filter,

Normalize.

Connect reference filter and let SA sweep. Example traces A and B. Then freeze trace A: set mode View. Now you have filter shape on the screen as your reference (trace A).

SA can sweep, but now only trace B is writing.

Take filter out and connect other filter for adjust to match with first one.

Trace A show reference filter shape and B follow nor this filter under check/adjustment.

[markus_jlrb]

Dear "rf-loop"

your absolutely right - for this kind of comparison measurements of antennas or filters
the continuouse sweep is an excellent mode. In my case I just compared DUTs statically
to play around with the SA so my perspective was different what leads to my statement.

Many thanks for your patience and explanations.

Enjoy your weekend and probably sure until next time when other issues arise during my
steps with this nice SA.

[pantelei4]

What is the news on the SHS800 series?
It's not Siglent Technical Support.   

[evgen.05]

Hello! Found a bug in sds 1202x-e. Single shot trigger not working on 50ms and higher. From 20ms to 2ns everything work fine.

[rf-loop]

Hello! Found a bug in sds 1202x-e. Single shot trigger not working on 50ms and higher. From 20ms to 2ns everything work fine.

Btw, why you tell it works from 20ms to 2ns. Why yiou did not tell it works 20ms to 1ns if you have used this scope.

Here it works rock solid also using 50ms/div and slower down to 100s/div. (long time for wait when testing)

Can you tell exactly in what mode you use scope.
Also it is good practice to tell FW version when report some possible problems independent of if error is user error or equipment error. Current FW version is 5.1.3.13

It nice if error reporting user can give as much as  possible (and also these what you perhaps think are not needed) information, not opinions but real data, starting from signal used for test  and ending to whole list of scope settings when firnd problem. This waysome other people here or in Siglent  can try repeat this problem and then it is very useful.



Input 20Hz   square wave and turn 50ms/div and set for ok signal level on the screen, example 4 - 6div height.
Turn Roll mode OFF!  and in trigger selections select Auto and push trigger level knob for 50%
Now select trigger Normal. Display looks as previously and you see it triggers and capture signal just as in auto mode (in this case)
Now push Sigle. You see it do single acquisition and stops.
What is now wrong?
Push again Single, again it do sigle shot.
You can change frequency and you can see it triggers just perfect and do just perfect one single shot for this

What I do not personally like at all is that default is that it force to roll mode always you go from 20ms/div to 50ms/div.
It urgently need feature that user can select if it do autoRoll mode or not. This defaulted auto Roll is terrible as long as user can not select if it do it automatically or not.

[evgen.05]

Hello! Found a bug in sds 1202x-e. Single shot trigger not working on 50ms and higher. From 20ms to 2ns everything work fine.

Here it works rock solid.

Can you tell exactly in what mode you use scope.
Roll or not Roll

Input 20Hz square and turn 50ms/div and set for ok signal level on the screen.
Turn Roll mode OFF!  and in trigger selections select Auto and push trigger level knob for 50%
Now select trigger Normal. Display looks as previously and you see it triggers and capture signal just as in auto mode (in this case)
Now push Sigle. You see it do single acquisition and stops.
What is now wrong?
Push again Single, again it do sigle shot.
You can change frequency and you can see it triggers just perfect and do just perfect one single shot for this

Yes Roll mode.
Tried to make how you told - works well. This behavior is very strange. Before this oscilloscope used Hantek. There is no this "feature" Roll mode and trigger in single shot mode worked without any "bugs"

[rf-loop]

Also Single works in Roll mode. But Roll mode may have less sample rate and it may work bit different what user may think, but it is special Rollmode single trig.
In normal and auto you can use full 14M but in Rollpaper mode you can use max 1.4M so samplerate is 1:10.
After it have been in Roll mode and you go to other mode remember to check memory maximum length setting. It is not memory lenghth, it is memory length maximum limit in Siglent. And Rollpaper mode force it to max 1.4M.


When user do Single in Roll mode there is one "trick" what need understand.
When you turn Roll mode to Sigle trigger mode it Roll. It do NOT wait trigger! But then after it get trigger it capture whole trace and stop so that trigger event is in trigger time position. Paper roll is rolling in Single mode until it meet trigger event!

[glenenglish]

Hello Sigilent support,
I am considering a sds2000X four channel.

1) SDS2304X : What is the sample rate per channel for a FOUR mode ? is it 2/2 = 1Gsps or 2/4 = 0.5 Gsps ?

2) your sds1202X-E , how does it do SPI decoding (4 wire) with only 3 inputs (ch1, ch2, trigger ) ? (or is it only MOSI or MISO one at a time ) ?

with thanks
glen

[tautech]

Hello Sigilent support,
I am considering a sds2000X four channel.

1) SDS2304X : What is the sample rate per channel for a FOUR mode ? is it 2/2 = 1Gsps or 2/4 = 0.5 Gsps ?

2) your sds1202X-E , how does it do SPI decoding (4 wire) with only 3 inputs (ch1, ch2, trigger ) ? (or is it only MOSI or MISO one at a time ) ?

with thanks
glen

I can answer both these.

1) 2 x 2 GSa/s divided by 4 = 1 GSa/s for all active channels. If ch 1 and 3 OR 4 are used 2 GSa/s is available for each of the 2 active channels. (4 ch models only)
Image below shows sampling rate and available memory depth (140 Mpts) halved when all are channels active.


2) Yes only MOSI OR MISO along with the clock.
At this time there is no provision to use EXT trigger in the Decode UI.

I can answer further questions by PM if needed as I have both these models.

Edit. Image added.

[rf-loop]

I am considering a sds2000X four channel.

1) SDS2304X : What is the sample rate per channel for a FOUR mode ? is it 2/2 = 1Gsps or 2/4 = 0.5 Gsps ?

Tautech allready answer this but still I will explain it bit more detailed so that there can not be any misunderstandigs also for possible other readers.

SDS2xx2X  2 channel models have 2GSa/s for 1 channel in use and 1GSa/s for both channels simultaneously. There is 140M memory. It is also shared with these two channels. For one channel max 140Msamples and 70M for both channels when both channels are use at same time.  Because sample rate is divided by 2 and memory length is divided by 2 acquisition time length is same.
This is because there is one two channel ADC chip. When 1 channel is in use this ADC works internally interleaved mode and samplerate is 2GSa/s and when it is used as 2 channels mode it have 1GSa/s for both channels simultaneously.

SDS2xx4X  4 channel models have two similar 2 channel systems.
 
It can say that there is  2 x  2 channel systems. We can think Ch1 and Ch2 is pair 1 and Ch3 and Ch4 is pair 2.

If in pair 1 is only one channel in use and same time in pair 2 is also 1 channel in use both these channels can use 2GSa/s and both these channels have max 140M acquisition lenght.

If least one pair have both channels in use then also other pair max is 1GSa/s.
So, if 3 or 4 channels are in use simultaneously maximum is 1GSa/s and max memory for every channel is 70M.

Using 16 channel  LA do not "eat" analog channels resources and vice versa. 

This same ADC and memory principle is also used in later coming SDS1xx4X-E models. Exept that ADC's there are 1GSa/s
 (or 1Ga/s as Siglent have mistyped in front panel sticker). 

[kvd]

On the SDS-1202X-E I noticed that the numeric entry window for the FFT Center Frequency reacts just opposite to turning of the Intensity/Adjust knob as all the other numeric entry windows.

This is quite annoying. Numeric entry is already quite cumbersome, even without this opposite behaviour. Siglent please correct this in the next firmware issue.

I am running 5.1.3.13 now.

[Plasmateur]

Hello, I was wondering if you could confirm my findings about the SDG2042X

I see two peaks on a signal analyzer when both channels are on.

It seems that the CH.1 has more isolation from channel CH.2 when both channels are on.

Example of my setup today.

I set CH.1  to 40MHz between 5-30mV @ 50Ohm

I set CH.2 to 38MHZ between 5-30mV @ 50Ohm

I then tested the frequency spectrum of each channel on a signal analyzer.

Signal analyzer indicates maximum isolation on CH.1 to be around 50dB with respect to CH.2 when both channels are on, and CH.2 has around 40dB with respect to CH.1 when both channels are on.

I switch frequencies on each channel and adjusted frequencies as well. Same results.

[rf-loop]

Hello, I was wondering if you could confirm my findings about the SDG2042X

I see two peaks on a signal analyzer when both channels are on.

It seems that the CH.1 has more isolation from channel CH.2 when both channels are on.

Example of my setup today.

I set CH.1  to 40MHz between 5-30mV @ 50Ohm

I set CH.2 to 38MHZ between 5-30mV @ 50Ohm

I then tested the frequency spectrum of each channel on a signal analyzer.

Signal analyzer indicates maximum isolation on CH.1 to be around 50dB with respect to CH.2 when both channels are on, and CH.2 has around 40dB with respect to CH.1 when both channels are on.

I switch frequencies on each channel and adjusted frequencies as well. Same results.

I can confirm it.
In my tests worst case what I can find without more long tests is:
And all values @50ohm  Open working channel terminated (in some cases effect is high)

CH1  38MHz, sine, 6.5mVpp
CH2  40MHz, sine, 6.5mVpp

Measured using SSA3021X
From CH1.   CH2 signal can see -57.5dB below CH1 signal.
From CH2.   CH1 signal can see -30.4dB below CH2 signal.


CH1  38MHz, sine, 6.6mVpp
CH2  40MHz, sine, 6.6mVpp

Measured using SSA3021X
From CH1.   CH2 signal can see -60.9dB below CH1 signal.
From CH2.   CH1 signal can see -41.8dB below CH2 signal.


0dBm level test:

CH1  38MHz, sine, 632.6mVpp 
CH2  40MHz, sine, 632.6mVpp

Measured using SSA3021X
From CH1.   CH2 signal can see -74.2 dB below CH1 signal.  (CH2 termination important!)
From CH2.   CH1 signal can see -77.9 dB below CH2 signal.  (CH1 termination important!)


My recommandation have always been that using function generator as rf generator best way is that set FG quite high level but not highest (example +10 ... +17dBm (10 - 50mW )  and use external attenuator for set low level signals. Not only for this reason here now but also overall signal quality.
This is also how professional RF generators work typically. They produce quite high level RF and after then, before output is attenuator.   

But, I hope Siglent do investigation why this leaks much more from

[Plasmateur]

Thank you or confirming rf-loop.

The reason why I was using such a low voltage was because of the 2nd harmonic I was seeing on the signal analyzer.

I will attempt to use 0dBm on both CH.1 and CH.2 and an external attenuator. Then I will check the 2nd harmonic on a signal analyzer. I might have to buy a low pass filter for the 2nd harmonic.

[Siglent America]

We have updated the firmware on both the the SDM3055 and SDM3065X DMMs.

SDM3055: 1.01.01.19
You can download the firmware from the below link:
 
http://www.siglentamerica.com/gjjrj-xq.aspx?id=6413&tid=15
 
The new firmware information as below:
1.      Add U disk recovery function.
2.      Modify Hold function arithmetic.
3.      Add the telnet SCPI function.
         Example: telnet 192.168.0.10  port: 5024
4.     Repair the external trig problem.
 
 
SDM3065X: 3.01.01.03
You can download the firmware from the below link:
 
http://www.siglentamerica.com/gjjrj-xq.aspx?id=6414&tid=15
 
The new firmware information as below:
1.      Add U disk recovery function.
2.      Add the update of the calibration data function.
3.      Fix the bug of a few abnormal DCV measurements when the input terminal is shorted.
4.      Repair the external trigger problem.
5.      Add the telnet SCPI function.
         Example: telnet 192.168.0.10  port :5024
6.      Some FPGA optimizations.

[flash2b]

Like I posted in the other thread related to the SDM3000 DMMs, I have update my SDM3055 to 1.19. This worked fine and even my reported 'beeper bug' was gone.

However I wonder what Siglent means with "Add U disk recovery function.". I browsed trough all the menus and I didn't see any new option in the file/usb section. Is this function some special thing where the device can recover from the USB when I fails to boot ?

Hope someone can give the answer.

[tautech]

Like I posted in the other thread related to the SDM3000 DMMs, I have update my SDM3055 to 1.19. This worked fine and even my reported 'beeper bug' was gone.

However I wonder what Siglent means with "Add U disk recovery function.". I browsed trough all the menus and I didn't see any new option in the file/usb section. Is this function some special thing where the device can recover from the USB when I fails to boot ?

Hope someone can give the answer.

I very much suspect it is.
Looking at the changelogs for several new FW versions we see this 'recovery function' being added to a # of instruments.
I've not bothered to ask as this seems the obvious explanation. (to me)

My 2c until informed otherwise. 

[borjam]

I very much suspect it is.
Looking at the changelogs for several new FW versions we see this 'recovery function' being added to a # of instruments.
I've not bothered to ask as this seems the obvious explanation. (to me)

My 2c until informed otherwise. 

If I remember well there were some issues with Siglent equipment failing to boot after a non graceful shutdown. It would certainly make sense to add some recovery option.

[tautech]

I very much suspect it is.
Looking at the changelogs for several new FW versions we see this 'recovery function' being added to a # of instruments.
I've not bothered to ask as this seems the obvious explanation. (to me)

My 2c until informed otherwise. 

If I remember well there were some issues with Siglent equipment failing to boot after a non graceful shutdown.

Only in the X-E (that I've seen mentioned) but AFAIK it was with the initial firmware, not any since.
There's no mention of the 'USB recovery' add on in the latest X-E FW 5.1.3.13.

It would certainly make sense to add some recovery option.

[MrW0lf]

Only in the X-E (that I've seen mentioned) but AFAIK it was with the initial firmware, not any since.

SDG2000X, after fw update + using using AWG software immediately afterwards.
Solution was mainboard replacement.