Siglent SDS800X HD Review & Demonstration Thread

[markone]

The RBW can only be achieved by manually setting the sample rate in the menu.
-snip

Not stupid at all, some automation in FFT settings would be more than welcome, direct RBW setting would be first in the list as well as  a "blocking" mechanism for impossible parameters combination, simply to mimic a true SSA, putting the timebase setting as a "slave" parameter.

For instance, I hate a lot that when i loose the CF setting increasing the Timebase, as a consequence of this you have to re-set the CF manually, that is a huge PITA.

[gf]

1) link to compressed bin file 100MSa @ 2GSa/s (size is about 95MB) : https://file.io/su83ZcbhPr4d

Thank you!

I removed my previous post where I messed up the vertical scaling
I've also received a .bin file from Martin72, so I decided to plot them side by side with yours.
Attached are the corrected/updated plots.
The similarity is amazing, although it is not the same instrument.

For figure4, I did downsample with anti-aliasing filering (which the scope does not do).
Compared to figure3, some of the spurs disappear (those that fold down from 62.5...1000 Mhz to 0...62.5 Mhz in figure3).

[KungFuJosh]

They both have SDG2000X AWGs as the source? That is crazy similar for the signals.

[Martin72]

Yepp,

We have the same generator:

https://www.eevblog.com/forum/testgear/sds800x-hd-review-demonstration-thread/msg5483791/#msg5483791

[KungFuJosh]

We have the same generator:

I guess we all do lol.

Did you use the 3kXHD or the 800 for your results?

[Martin72]

Mostly the 800, except this one:
https://www.eevblog.com/forum/testgear/sds800x-hd-review-demonstration-thread/msg5494624/#msg5494624

[eTobey]

Is there a way, to create a math formula, so one could easily see, what the duty cycle (yellow) roughly is?

[ebastler]

Is there a way, to create a math formula, so one could easily see, what the duty cycle (yellow) roughly is?

A "Measurement Track" (section 18.5 of the manual) should do what you want: Calculates time-dependent measurements within one capture and plots them like a trace. 

[Performa01]

...
The commonly agreed industry standard for the maximum acceptable input VSWR is 1.5:1, which is equivalent to 14 dB return loss.
...

I've never heard of such a standard.  Could you provide a reference?  If you are making measurements, 1.5 seems really high. 

Looking at the manual for my LeCroy 7200, they spec the VSWR at 1.2, BW is 4GHz with a 125pSec transition for 50mV and up.  Isolation is 60dB min at 1GHz.

I have no reference to such a standard – my statement is based solely on the impression left by various RF literature, experience and common sense. Sorry if it has been misleading.

At least one example, showing that I’m not the only one who thinks 1.5:1 is the limit of acceptable (some might call it ideal) VSWR, can be found here:

https://www.ahsystems.com/notes/VSWR-return-loss-sheet.php

A VSWR of less than 1.5:1 is ideal, a VSWR of 2:1 is considered to be marginally acceptable in low power applications where power loss is more critical …

Very few oscilloscopes have the input return loss / VSWR specified; one of the few exceptions was the now discontinued 2 GHz R&S RTE. From the data sheet Version 16.01:

Input VSWR : input frequency ≤ 500 MHz 1.25 (meas.); input frequency > 500 MHz 1.4 (meas.)

I take from this, that a company, that is renowned to be an RF-specialist since 1933, has no problems specifying a VSWR of 1.4:1 for frequencies >500 MHz on an instrument that isn’t exactly a cheap low-end product. There exists no VSWR-specification for its successor, the MXO5, anymore…

Finally, the common-sense part: A VSWR of 1.5:1 means 4 % reflected power, i.e. 4 % power loss at the receiver end (scope input). That’s 0.17 dB = ~2% voltage error. It is less than the DC-accuracy specification of most oscilloscopes.


As can be seen from my measurements, the 500 (actually 570) MHz SDS2000X HD provides a VSWR of <1.25:1 up to ~750 MHz. That should be in line with your expectations based on the LeCroy 7200.

[joeqsmith]

Thanks for the response.  It could very well be my lack of any common sense and experience causing my confusion.

If I were considering the acceptable VSWR of my antenna system, I could see some general guidelines.   If I were trying to measure the output power of my amplifier, I would have a different criteria for the load match of my power meter (mine is 1.05 for example).   There is a reason you will find many application notes about using precision attenuators to improve port matching when making measurements.    Consider some parts may have a very good isolation between their input and output and the measurement uncertainty would be much less. 

My vintage LeCroy 8500A specs 50+/-2%.  20log10(50*0.02/50*2.02) = 40dB return loss, but that's at DC.   An 8-bit system like this is going to mask a lot of the errors which may be why manufactures didn't provide the details.   

I imagine as manufactures improve their products and users find more applications for them, spec's like return loss will become more important.   

...
The commonly agreed industry standard for the maximum acceptable input VSWR is 1.5:1, which is equivalent to 14 dB return loss.
...

I've never heard of such a standard.  Could you provide a reference?  If you are making measurements, 1.5 seems really high. 

Looking at the manual for my LeCroy 7200, they spec the VSWR at 1.2, BW is 4GHz with a 125pSec transition for 50mV and up.  Isolation is 60dB min at 1GHz.

I have no reference to such a standard – my statement is based solely on the impression left by various RF literature, experience and common sense. Sorry if it has been misleading.

At least one example, showing that I’m not the only one who thinks 1.5:1 is the limit of acceptable (some might call it ideal) VSWR, can be found here:

https://www.ahsystems.com/notes/VSWR-return-loss-sheet.php

A VSWR of less than 1.5:1 is ideal, a VSWR of 2:1 is considered to be marginally acceptable in low power applications where power loss is more critical …

Very few oscilloscopes have the input return loss / VSWR specified; one of the few exceptions was the now discontinued 2 GHz R&S RTE. From the data sheet Version 16.01:

Input VSWR : input frequency ≤ 500 MHz 1.25 (meas.); input frequency > 500 MHz 1.4 (meas.)

I take from this, that a company, that is renowned to be an RF-specialist since 1933, has no problems specifying a VSWR of 1.4:1 for frequencies >500 MHz on an instrument that isn’t exactly a cheap low-end product. There exists no VSWR-specification for its successor, the MXO5, anymore…

Finally, the common-sense part: A VSWR of 1.5:1 means 4 % reflected power, i.e. 4 % power loss at the receiver end (scope input). That’s 0.17 dB = ~2% voltage error. It is less than the DC-accuracy specification of most oscilloscopes.


As can be seen from my measurements, the 500 (actually 570) MHz SDS2000X HD provides a VSWR of <1.25:1 up to ~750 MHz. That should be in line with your expectations based on the LeCroy 7200.

[Performa01]

So now we can officially say that there is some margin of improvement in FFT performance, could you "disclose" if we will see something in regard with the incoming FW release?

Sorry, but even when it might appear that way, that issue isn't simply a programming error. Repair of the FFT would be relatively easy, but could affect other math and measurement functions in a negative way - and the perfect solution might take a lot of additional resources. So the viable solutions are still discussed internally and I cannot even know how much effort the final (safe) solution will be.

[eTobey]

Is there a way, to create a math formula, so one could easily see, what the duty cycle (yellow) roughly is?

A "Measurement Track" (section 18.5 of the manual) should do what you want: Calculates time-dependent measurements within one capture and plots them like a trace.

Thanks.

Thats kinda nice, but why didnt they include this in the math? Alignment is everything with a scope. 

2 Sentences in the manual, but 6 settings / buttons in the scope.

[Performa01]

Thanks for the response.  It could very well be my lack of any common sense and experience causing my confusion.

No need to be sarcastic.

I thought it was clear that I was talking about oscilloscopes and not power meters, which commonly have an order of magnitude better accuracy. Consequently, I would also prefer a VSWR of 1.05 for the power meter.

I thought it was clear that I was talking about oscilloscopes and not complete measurement setups. Nowhere did I state that a better VSWR could not be beneficial at times, and that it could not be achieved by adding an inline attenuator – even though in practice I’ve more often wanted to improve the output impedance match of a generator than the input match of an oscilloscope – but that’s just me.

[joeqsmith]

There is no reason an oscilloscope could not measure power.  In this thread we are seeing a lot more applications and I don't see this curbing anytime soon.  I'm not suggesting  an oscilloscope is a replacement for a power meter, network analyzer, spectrum analyzer....   Rather, modern scopes are slowly smearing the lines.   As their performance improves and the features they support grow, I suspect such details will become more important.   While a VSWR of 1.5 may be a good guideline when talking about antennas, it's not something I have ever considered when making a measurement in the lab.  Certainly not a standard.   Rather, I treat each measurement as a unique problem.

Oh well, sounds like I have my answer and not trying to upset you.  Thanks for the input.   

Thanks for the response.  It could very well be my lack of any common sense and experience causing my confusion.

No need to be sarcastic.

I thought it was clear that I was talking about oscilloscopes and not power meters, which commonly have an order of magnitude better accuracy. Consequently, I would also prefer a VSWR of 1.05 for the power meter.

I thought it was clear that I was talking about oscilloscopes and not complete measurement setups. Nowhere did I state that a better VSWR could not be beneficial at times, and that it could not be achieved by adding an inline attenuator – even though in practice I’ve more often wanted to improve the output impedance match of a generator than the input match of an oscilloscope – but that’s just me.