Siglent SDS6000A DSO's 500MHz-2GHz

[tautech]

This is the SDS5000X FW.

It is ! 

Correct link:
https://int.siglent.com/upload_file/zip/firmware/Oscilloscope/SDS6000_V1.4.3.3_EN.zip

~120MB

[oewean]

Sorry for the wrong URL, updated now!

[bson]

Would you mind posting the SDS5000X link again?  Or was this the 4/27/22 release?

[tautech]

Would you mind posting the SDS5000X link again?  Or was this the 4/27/22 release?

Yes it was and it's an important update. 

[tv84]

SDS6000A board picture

[tv84]

Misc pics 1/3

[tv84]

Misc pics 2/3

[tv84]

Misc pics 3/3

[Martin72]

SDS6000A board picture

Description of the main parts (pic attached).

[tv84]

Following its little HD brother...   

[tautech]

Following its little HD brother...   

I'm surprised you didn't crack it sooner. 

[tv84]

I'm surprised you didn't crack it sooner. 

Your fault, you are not selling them fast enough, so small playground! 

Maybe now the ADCs price change... 

[Performa01]

Just recently I’ve demonstrated how to get a better frequency resolution at higher frequencies for the SDS2000X HD. One fo the methods was downconversion using the waveform math as a mixer. For this, a second channel had to be fed with the oscillator signal. See reply #273 in the following thread:

https://www.eevblog.com/forum/testgear/siglent-sds2000x-hd-12bit-(published-for-chinese-domestic-market-only)/msg4320658/#msg4320658

I’ve also mentioned that maybe one day we might get full length reference traces in our DSOs, which would enable us to do the downconversion without additional signal source. The SDS6000 already takes one step in that direction, as it has the Memory channels, which can be used as input for math expressions. A Memory trace can have up to 12.5 Mpts, so this is some serious amount of data.

Yes, it’s actually possible to do a clean downconversion from 29.6 MHz to 600 kHz, see attached screenshot.

SDS6204_FFT_512kpts_DC_M29MHz_125MSa_6.25MSa_29.6MHz_0dBm_Excl

Of course, this is not quite as crucial, as the SDS6000 has four times the FFT-length, hence four times the frequency resolution of an SDS2000X Plus/HD from the outset.

WARNING: Wnile this demonstrates the basic potential of memory traces, this particular example is unlikely to work with an 8-bit SDS6000A!

[PartialDischarge]

Does this scope have a real separate 50Ohm input?

I'm surprised at the low noise it has comparing other models....

[tautech]

Does this scope have a real separate 50Ohm input?

Vertical scale
(probe 1X): 1 MΩ: 0.5 mV/div – 10 V/div
                  50 Ω: 0.5 mV/div – 1 V/div

https://int.siglent.com/u_file/document/SDS6000A_Datasheet_EN01C.pdf

We have the 2 GHz model, do you want to see any particular measurement ?

[PartialDischarge]

We have the 2 GHz model, do you want to see any particular measurement ?

Can u show the pulse response with a fast pulse generator (at least 5ns rise) and 50Ohm termination both in 10ns/div and 10us/div?

[tautech]

We have the 2 GHz model, do you want to see any particular measurement ?

Can u show the pulse response with a fast pulse generator (at least 5ns rise) and 50Ohm termination both in 10ns/div and 10us/div?

Done in reply #206 with a Leo Bodnar 30ps pulser. 

[rf-loop]

We have the 2 GHz model, do you want to see any particular measurement ?

Can u show the pulse response with a fast pulse generator (at least 5ns rise) and 50Ohm termination both in 10ns/div and 10us/div?

 

U think it is some kind of FNIRSI & alii.........

[PartialDischarge]

 

U think it is some kind of FNIRSI & alii.........

I don't know what you find so funny, but I'm glad that your ignorance makes you laugh so hard! keep it at it, laugh is good!

[Performa01]

Does this scope have a real separate 50Ohm input?

I'm surprised at the low noise it has comparing other models....

What exactly do you want to know?

The SDS6000 is still a general purpose MSO with universal BNC inputs that can be switched from high-Z to 50 ohms.

The relevant question would be: what is the input return loss in 50 ohms mode?

The answer is: VSWR < 1.2 up to 1.2 GHz, VSWR < 1.4 up to 2 GHz (typical)

Regarding the noise, there is little difference between high-Z and 50 ohms inputs. Yet the 1 meg inputs are a bit more quiet, which shouldn’t be too much of a surprise because of the nice lowpass filter created by the input capacitance of the high-Z architecture.

According to the datasheet, the SDS6000 is limited to 1 GHz at input gain settings below 2.3 mV/div. Yet the actual gain difference is less than 3 dB at 2.5 GHz, so the noise bandwidth is almost the same. In any case, this is the reason why I provide measurements not only for the highest sensitivity of 500 µV/div, but also for 2.3 mV/div.

Input 1 meg, gain = 2.30 mV/div, BW = 2 GHz: 116 µVrms noise
Input 50 ohms, gain = 2.3 mV/div, BW = 2 GHz: 142 µVrms noise
Input 1 meg, gain = 500 µV/div, BW = 1 GHz: 116 µVrms noise
Input 50 ohms, gain = 500 µV/div, BW = 1 GHz: 128 µVrms noise

See corresponding screenshots

SDS6204_Noise_FFT_1M_2.3mV_2GHz
SDS6204_Noise_FFT_1M_500uV_1GHz
SDS6204_Noise_FFT_50_2.3mV_2GHz
SDS6204_Noise_FFT_50_500uV_1GHz

[Performa01]

I don't know what you find so funny …

Maybe it’s because you seem to want to challenge an upper midrange 2 GHz MSO with a 5 ns pulse – a test scenario more commonly found when checking the base functionality of a 200 MHz entry level DSO?

Anyway, here’s a 10 ns wide pulse with 1.5 ns rise/fall time, measured at 10 ns/div, see first screenshot.

SDS6204_Pulse_10ns_10-1.5ns

Next is the same pulse at 10 µs/div; we can see the repetition frequency of 1 MHz. We can also see the FFT-spectrum of the pulse, which tells us that even 1.5 ns transition times don’t require massive bandwidth for a faithful reproduction and 350 MHz should be plenty for an 8-bit system.
 
This also demonstrates the beauty of deep measurements, which provide accurate pulse width and rise/fall times even at slower time bases and longer records like the 0.5 mega points here.

SDS6204_Pulse_FFT_10us_10-1.5ns

[PartialDischarge]

I don't know what you find so funny …

Maybe it’s because you seem to want to challenge an upper midrange 2 GHz MSO with a 5 ns pulse – a test scenario more commonly found when checking the base functionality of a 200 MHz entry level DSO?

And hence my comment about his ignorance, because I'm not interested in the edge response but what comes after that to see if reflections happen, so 1ns or 5ns rise time is ok.
Also for the 10us/div I'm interested in the flatness of the pulse as this low frequency area is common to have non flat frequency responses in many scopes due to non-linear capacitances in the boards where the attenuators are built.
So never assume what other knows or does not know, and laughing at others for asking just shows sheer stupidity, besides the aforementioned ignorance.

[tv84]

So never assume what other knows or does not know, and laughing at others for asking just shows sheer stupidity, besides the aforementioned ignorance.

And what do your assumptions show? Intelligence? Education?

[PartialDischarge]

And what do your assumptions show? Intelligence? Education?

Oh I can tell you that you'll never see me laughing at someone for his questions, if I see someone here with basic or beginner questions either I pass for lack of time or do answer trying to help in what I can.

[Martin72]

You shouldn´t take some things too seriously, especially in forums...