Siglent SDS1000X and (MSO) SDS1000X+ series oscilloscopes

[snik]

Have try diffrent voltage Ranges, from mV to Volts.  All the same drop. The 4 Volts was for example.

[Wuerstchenhund]

Have a SDS1102X for some time and was happy with it.
Yesterday I do some sweeps on this Oscilloscope and I recognize that I have over the full Frequency Range of the scope an continues Voltagedrop. For example at 10 MHz to 100 MHz at a Signal from 4Vpp, correct terminated, i have a Drop from  1.2 Vpp.

Aside from the question of cabling and termination, the SDS1102X is an 100MHz scope, which means at 100MHz the signal attenuation is specified as 3dB (0.707x Vs). In reality, most scopes are better than their specified BW, but there will still be some attenuation once you reach the specified BW limit.

Assuming your cable was fine and you terminated correctly, a drop of 1.2Vpp with a Vs of 4Vpp equals an attenuation of approx 3dB.

Sounds reasonable to me 

[snik]

When I reach the end of the scope frequency, that there is a voltage Drop that's okay for me that's the 3dB Rule which I know , but in lower frequency the voltage should be constant. But here the voltage drops from beginning. Even my cheap ass Hantek is better and constant at voltage in lower  freqency. Drops only in Higher Frequency.
The siglent voltage is for example at 50 MHz 3.2Vpp at 4V Signal. That could not be right

[Wuerstchenhund]

When I reach the end of the scope frequency, that there is a voltage Drop that's okay for me that's the 3dB Rule which I know , but in lower frequency the voltage should be constant. But here the voltage drops from beginning. Even my cheap ass Hantek is better and constant at voltage in lower  freqency. Drops only in Higher Frequency.
The siglent voltage is for example at 50 MHz 3.2Vpp at 4V Signal. That could not be right

No, that's doesn't sound right. Can you be sure that the drop is not caused by your signal source, or cabling/connectors/adapters? If so then this looks like a problem with the scope, i.e. a problem with your specific unit or a general bug/flaw.

[Performa01]

Yesterday I do some sweeps on this Oscilloscope and I recognize that I have over the full Frequency Range of the scope an continues Voltagedrop. For example at 10 MHz to 100 MHz at a Signal from 4Vpp, correct terminated, i have a Drop from  1.2 Vpp.

...

Is that normal at this scope ? I think not. Firmware is latest.
Has anybody an Idea what's wrong with it ?

You have plugged the probe into the signal generator output.

What bandwidth is that probe?
Are you sure the probe compensation is adjusted properly?

Even if the rated probe bandwith exceeds 100MHz by a fair margin, its frequency response will not be perfectly flat. From my experience, even 300MHz and higher rated probes can drop up to 2dB at 100MHz (and read up to 2dB high at 250MHz in return).

Make a direct connection with quality coax cable and switch the scope input to 50 ohms, then check again.

[rf-loop]

Did you properly terminate the BNC cable at the input of the oscilloscope? You have to select input impedance of 50 ohm in the oscilloscope to avoid reflections, which will lead to a reduced amplitude level.

In video it looks more like there is probe connected to SDG output. So if this is correct termination and for look scope frequency response...  no, there is nothing right for this purpose.

@snic ask if this is normal with this scope.

I can answer, if we do not know anything but what is in the video. Yes this is very normal for this kind of playing.

Btw, where is SDS1102X front panel knobs.

This is example about how oscilloscope frequence response can not measure at all. With this method can only "wonder" what all "miracles" happend...

Btw, oscilloscope probe in SDG output...  source impedance is 50ohm.
If look only one thing (and there is many what affect) probe cap is around 20pF if it is "usual" around hundred MHx 10x probe. 
If it is 20pF with 10MHz its reactance is around 800 ohm.  With 100MHz  reeactance is around 80 ohm. Think about it.

Oscilloscope frequency response measurement need constant level over frequency band in scope input terminal and exatly just there. (also this is much more complex thing if look even bit more deep.)  Example. Someone have scope what have 1Mohm input and input capasitance is say example 15pF.  Now one connect 50ohm feed thru terminator to input. Then he happy believe his input is now 50ohm input. Bullshit. Only for DC.

[snik]

The Probe is an agilent Probe with high frequency, didn't know exactly 250 or 300 MHz.
Have also used my 4 Coax-Cables with 50 Ohm terminated. All the same.
Was Sitting all the night when I see error  to figure out what's wrong but it seems it is the scope. But what could cause the Problem ?
I have also ooened the scope and compare it with Fotos from Daves teardown of the sds1202x but it looks okay at first overview.

[crazyguy]

Have a SDS1102X for some time and was happy with it.
Yesterday I do some sweeps on this Oscilloscope and I recognize that I have over the full Frequency Range of the scope an continues Voltagedrop. For example at 10 MHz to 100 MHz at a Signal from 4Vpp, correct terminated, i have a Drop from  1.2 Vpp.
I have done a short Video :

https://youtu.be/ivatJ9_jq6U

Is that normal at this scope ? I think not. Firmware is latest.
Has anybody an Idea what's wrong with it ?

Michael

Siglent SDS1102X v Rigol MSO1074Z Part A-Vert/Horiz by tomtektest
https://www.youtube.com/watch?v=JLzq3IyGpOQ#t=17m00s

In his test result the -3dB cut-off frequency is at ~173MHz

[rf-loop]

The Probe is an agilent Probe with high frequency, didn't know exactly 250 or 300 MHz.
Have also used my 4 Coax-Cables with 50 Ohm terminated. All the same.
Was Sitting all the night when I see error  to figure out what's wrong but it seems it is the scope. But what could cause the Problem ?
I have also ooened the scope and compare it with Fotos from Daves teardown of the sds1202x but it looks okay at first overview.

Do you know really this SDG flatness up to 100MHz. This is SDG2042X.

What you know about actual  real signal level in oscilloscope input terminal.  Do you have any instrument for check it.

This because I do not trust this SDG2042X level accuracy with 100MHz (and specially when load reactance is what ever)

One other thing. Do you have good quality feed thry 50ohm terminator. If you have, do test so that you use scope input high impedance with this terminator.  Then other test without it and scope internal 50ohm as termination. As short 50ohm coaxial as possible.  Also if you have example 10dB atenuator, put it to scope input when use scope internal 50ohm.
Then use (just for reference) 300mV-pp sinewave for test.

In some message you say Hantek cheap scope show it better, you think. How you know it show more right. it is other and good question.

If you can not see difference when you measure with this agilent probe connected directly to SDGoutput BNC and then good coaxial  terminated with 50ohm, sorry but it is hard to belive. I expect least some visible difference. Not only between 10 and 100MHz points but also in whole response shape. (with peak mode, roll, and SDG sync you can produce whole response curve.


Until you know what is signal level really in scope input we can not say anything sure.

Is it possible you can use real trusted leveled RF generator.

I have measured several SDS1102X scopes and everyone -3dB point related to 1MHz reference level is always over 140MHz with all V/div settings in two bottom V band. (whole V range is divided to 3 bands)

And for sidemark, I have measured lot of Hantek's (as I previously have sold these, and I teest every single scope what I sell, everyone. I know what its freq response hassle is - really. So if Hantek tell signal flatness is better, it can be true or it can be wrong. Of course it is possible it is more right but, with tens of Hantek experience, sorry but I do not believe anything what this crap show until I really know individual Hnatek what I have measured then I can somehow trust how much it lie. (all scopes lie, indepeendent of brand and price, one less than other etc..

IF this SDS1102X freq response is really out of specification then just send it to repair, it have warranty.
I believe you have not adjusted this front end  "accidentally" when you have opened it.  Btw, why you open it if you think something is wrong and there is warranty. And even more strange because you are in Germany and Siglent service is also there.

Specifications:
Bandwidth Flatness
DC ~ 10%(BW): ± 1 dB
10% ~ 50%(BW): ± 2 dB
50% ~ 100%(BW): + 2 dB / -3 dB

But first need measure it really.

[snik]

Do you know really this SDG flatness up to 100MHz. This is SDG2042X.
What you know about actual  real signal level in oscilloscope input terminal.  Do you have any instrument for check it.
This because I do not trust this SDG2042X level accuracy with 100MHz (and specially when load reactance is what ever)

But also in the original Range up to 40 MHz the Voltage drops at the SDS1102X...
Have check the SDG with two other scopes (Rigol DS1102Z and Hantek DSO5102P), there the Signal looks like it should be.
Also tested Signal at an HM5012 Spectrum Analyzer.

One other thing. Do you have good quality feed thry 50ohm terminator. If you have, do test so that you use scope input high impedance with this terminator.  Then other test without it and scope internal 50ohm as termination. As short 50ohm coaxial as possible.  Also if you have example 10dB atenuator, put it to scope input when use scope internal 50ohm.
Then use (just for reference) 300mV-pp sinewave for test.

Have got more 50Ohm Attentuators, tested all, all the same. Tested with Internal (Measeured 51Ohm). My Coaxes are two with 1 meter and one with an 0,5 meter. Tested all cables and also with Probe and also tested with the other scopes - at the other scopes no problem only at Siglent.

In some message you say Hantek cheap scope show it better, you think. How you know it show more right. it is other and good question.

At the Hantek the Voltage doesn't drop and shows the right Values until at the end at the Frequency range, but thats okay. Also at Rigol Signal is also bang on.

If you can not see difference when you measure with this agilent probe connected directly to SDGoutput BNC and then good coaxial  terminated with 50ohm, sorry but it is hard to belive. I expect least some visible difference. Not only between 10 and 100MHz points but also in whole response shape. (with peak mode, roll, and SDG sync you can produce whole response curve.

My Problem isn't some mV more or less. (Have you look at the Video ?) There is a constant linear Voltage drop - tarting at low frequency up to the end. With all cables, Frequencys and probes i tested. 
This problem doesn't occur at the other scopes.  Only at the Siglent.

Is it possible you can use real trusted leveled RF generator.

No, not in short time.

I have measured several SDS1102X scopes and everyone -3dB point related to 1MHz reference level is always over 140MHz with all V/div settings in two bottom V band. (whole V range is divided to 3 bands)
And for sidemark, I have measured lot of Hantek's (as I previously have sold these, and I teest every single scope what I sell, everyone. I know what its freq response hassle is - really. So if Hantek tell signal flatness is better, it can be true or it can be wrong. Of course it is possible it is more right but, with tens of Hantek experience, sorry but I do not believe anything what this crap show until I really know individual Hnatek what I have measured then I can somehow trust how much it lie. (all scopes lie, indepeendent of brand and price, one less than other etc..

As i wrote above, also tested at a Rigol and at a Analyzer. All your questions, that was what i also thinking about and tested.  The Hantek is really cheap creepy Stuff (That why i bought the Siglent) , but here it is bang on at the Signal Voltage, the Rigol too.

I believe you have not adjusted this front end  "accidentally" when you have opened it. 

No - there was nothing "accidentally" happen ( except the broken seal  ) The issue was before i open it ... 

Btw, why you open it if you think something is wrong and there is warranty. IF this SDS1102X freq response is really out of specification then just send it to repair, it have warranty.

Bought the Siglent from an chinese eBay Seller for a (very) good price. It was a Seller who have a small bunch of (Black ?) items here in germany (without cusoms duty). That's maybe the reason for the issue 
I will not asked Siglent Germany for warranty, because i didn't buy it from an offical Reseller. 
That's the price you pay some time with these cheap seller on ebay, but i knew that risk before 

Now i must search the failure (Or put it in the dumpster  ).
Was hoping somebody have a hint for me.

Thanks for the Help.

[nctnico]

You could always try to send it in for warranty. Any idea about the age of the scope?

[rf-loop]

Btw, why you open it if you think something is wrong and there is warranty. IF this SDS1102X freq response is really out of specification then just send it to repair, it have warranty.

Bought the Siglent from an chinese eBay Seller for a (very) good price. It was a Seller who have a small bunch of (Black ?) items here in germany (without cusoms duty). That's maybe the reason for the issue 
I will not asked Siglent Germany for warranty, because i didn't buy it from an offical Reseller. 
That's the price you pay some time with these cheap seller on ebay, but i knew that risk before 

Now i must search the failure (Or put it in the dumpster  ).
Was hoping somebody have a hint for me.

Thanks for the Help.

When it arrive how it was. Originally siglent sealed carton?
Did it come with calibration certificate?

Are there any signs that it was perhaps localized for China domestic markets?

There is some adjustments in front end but I'm not sure how much it can help.

But first, ask what Siglent think if you can get some help for it. Where ever buy but it is still Siglent original genuine SDS1102X product (or is it?).

(inside China domestic markets there are models:
SDS1052X
SDS1072X
SDS1102X
SDS1202X)

[Performa01]

Sounds like the input divider frequency compensation is not ok.
Have you tried with a smaller signal that doesn't require input attenuation? This might be the case for vertical settings 100mV/div and less, but I cannot know exactly, maybe RF-Loop does. If the issue disappears at the lower voltage ranges, it most likely is an input divider compensation issue.

Check the serial number and try tp find out if it is suspicios. Maybe the Chinese guy sold you a null-series unit that he somehow managed to acquire and that was never intended to be sold, maybe also never got properly calibrated/adjusted.

[snik]

Here a short update. "Fixed" my problem.

After some looking and measuring in the Frontend i recognize that at the AD8370 Output (Pin 8/9) is an Capacitor. So i compare with Dave's teardown pictures of the SDS1202X and ... there is no Cap. 

Look here :



Then i desoldered the cap at channel 2 to take some measurements and this is the result :


Here i have a 300mVpp Voltage at both Channels, Ch1 (yellow) is with Cap and Ch2 (Pink) is without. Looks pretty good 

So i think Siglent use this Capacitors for Bandwidth limiting.

Now i removed both and now i'm happy with the result 

[rf-loop]

Electronics world is full of big and small miracles.
  

btw, can you measure this capasitance value enough accurately?

[nctnico]

Still wondering why that capacitor was there. 50 or 70 Mhz version maybe which got the wrong sticker on the casing??

[snik]

Still wondering why that capacitor was there. 50 or 70 Mhz version maybe which got the wrong sticker on the casing??

Firmware say its's a SDS1102X. Under the mainboard is a sticker (See attachment). Maybe someone have a clue with the numbers.

btw, can you measure this capasitance value enough accurately?

The cap was something around 10/12pF.

[rf-loop]

Now i removed both and now i'm happy with the result 

What ever is reason but  most important is you are happy with result.
But AFAIK all  labels etc tell it is factory original genuine SDS1102X.

Btw, with external high quality feed thru terminator + 6dB attenuator, very short cable  and scope input 1Mohm I have measured normal factory original SDS1102X -3dB over 180MHz using 300mVpp high quality sine. Scope V/div of course 50mV/div with this signal.  Using longer cable, without extra attenuator and scope input internally 50ohm, result worse. BW shape worse and -3dB nearly 50MHz lower.   We can measure frequency response with so different tools etc.. 
Also, it is not real 50ohm impedance over freq range when only add parallel 50ohm resistor with 1Mohm 18pF input.  Reflections in cable may produce many fun "things"... in worst case also inside signal source leveling system.

[snik]

Btw, with external high quality feed thru terminator + 6dB attenuator, very short cable  and scope input 1Mohm I have measured normal factory original SDS1102X -3dB over 180MHz using 300mVpp high quality sine. Scope V/div of course 50mV/div with this signal.  Using longer cable, without extra attenuator and scope input internally 50ohm, result worse. BW shape worse and -3dB nearly 50MHz lower.

I have made only for the measurement an short RG316 Cable. Measurement was with an external Terminator but without an attenuator. You see on my picture above, now without the cap the scope can do much more BW 
The only reason for me, why Siglent put in the Cap, is for Bandwidth limitations. For what else ?  Would like to know if this capacitor is in all SDS1102X or if only mine was an exception. Can't really believe that. 
Without the capacitor the scope maybe can go over the 200 MHz but for now i have no HF signal generator to test that.
For the "normal" 100MHz BW the scope is now really good  and that is in the moment all what i need

[tautech]

For the inbuilt AWG on this series of DSOs (Plus and S models), new EasyWave software:
http://www.siglentamerica.com/USA_website_2014/Firmware&Software/Software/EasyWave_V100R001B01D01P34.rar

Note
This is NOT FW, it is PC software for AWG remote management over LAN or USB.

UI screenshot

[tautech]

So following on from general discussion in other threads about DSO channel offset and baseline noise here are a few screenshots showing real levels and how they might affect real measurements. If anything these might confirm some of the tricks Dave used in his vids "Why are DSO's so Noisy ?"
I say "tricks" but these are some of the most widely used simple methods to clean up a waveform if and when needs be.

SDS1102X+ 500uV/div, 50 Ohm input, DOT mode
FW P06
Source:
50 Ohm
SDG810 (Siglent's cheapest AWG series)
2mV @ 1 MHz

I have done these shots in 2 sessions, 1 of which the instruments were not warm or Autocal'ed so there is some little variation in the UI indicated levels.
Along with image labels the info in the GUI should explain all.

500 uV GND coupled


Open input showing channel offset, indicated onscreen when the "Set Trigger to 50%" is pressed.


Full BW 500uV/div and holdoff increased to 600ns for stable display (AWG jitter?)


20 MHz BW limit


BW Limit and 4 Averages


BW Limit and Eres


And finally Full BW, Colour and Infinite Persist



There's a multitude of minor measurement differences in these shots, the frequency counter and OSD are somewhat sensitive to trigger levels and noise at this low level (2 mV) input.
Enjoy.

[nctnico]

I would have used a 20dB or 40dB  attenuator on the generator. That way the noise from the generator gets reduced so you -in theory- only see the noise contribution of the scope.

[tautech]

I would have used a 20dB or 40dB  attenuator on the generator. That way the noise from the generator gets reduced so you -in theory- only see the noise contribution of the scope.

Various connection methods including a Tek 50 Ohm 10x BNC attenuator were attempted to evaluate the impact on the displayed waveform and as differences were so minor only straight 50 Ohm BNC to BNC was used.

[vpv]

Hello!
I bought Siglent SDS1102X and so far I'm very pleased.
I have one question: when I check in "Information", there is "PAS" OptionName.
What is PAS?
Thank you!

[tautech]

Welcome to the forum.

PAS is Power Analysis.
There is a whole suite of tools within used to measure and calculate efficiencies in SMPS design.
Use requires voltage and current probes.