Products > Test Equipment

Siglent SDS800X HD Review & Demonstration Thread

<< < (74/87) > >>

markone:

--- Quote from: gf on May 11, 2024, 08:47:21 am ---
--- Quote from: markone on May 11, 2024, 01:13:38 am ---1MHz SPAN :

...
I would say that SDS800x FFT is not usable at low RBW (in relation to the central frequency), if you guys can say with certainty that there are no bugs in the FW FFT implementation my interest in this functionality, apart from a few cases at quite low frequencies, ends here.

--- End quote ---

* Could you capture 100M samples of this signal at full 2GSa/s sample rate and post the .bin file?
   My feeling is that what you see in the FFT graph is already contained in the samples. I tend not to blame the FFT for that.

--- End quote ---

100MSa are available @ 2GSa/s | 5ms/div, with this setting FFT works up to 25MHz, 50MHz CF is possible up to 2GSa/s | 2ms/div that corresponds to 40MSa, not relevant to you request but it's important to know.

My feeling is that here we have a "beat" phenomenon between ADC sample frequency and input signal frequency, not an expert in FFT math but I guess there are techniques to mitigate it.

Anyway, do you have an application able to dump the whole acquisition buffer ?


--- Quote from: gf on May 11, 2024, 08:47:21 am ---
* On a larger 0...1GHz span, how do the levels of the harmonics compare between FFT and SA?


--- End quote ---

It's all up to RBW setting related to CF, any SPAN/CF combination has a RBW thereshold under which the spurious fest start, anyway the SSA provides this spectrum image :




--- Quote from: gf on May 11, 2024, 08:47:21 am ---
* Does a larger V/div (200mV/, 500mV/,...) reduce the amplitude of the close-in spurs (and harmonics)?


--- End quote ---


Tried infinite settings, this is definitely not the case.


--- Quote from: ebastler on May 11, 2024, 06:27:20 am ---So, potentially an error in the down-sampling to the FFT's 2 MPt maximum?

Not necessarily an error. If there exist harmonics beyond 75 MHz then it is expected that they will fold into the first Nyquist zone, and if the carrier deviates only slightly from 50 MHz, the folded harmonics can indeed form a comb in the neighborhood the fundamental frequency.

--- End quote ---

Those spurious, that of course do not actually exist, appear with many FFT point size settings , as already mentioned to boredom it's all up to RBW setting toward CF value.

markone:

--- Quote from: Martin72 on May 11, 2024, 01:01:43 pm ---
--- Quote from: markone ---could you try again with also 50KHz and 1MHz SPAN ?
--- End quote ---

Same results, but won´t post it until the "new" thread exist..

--- End quote ---

are you sure that it will never exist ... ?

Martin72:
I have now pressed the "Moderator Button" again, we are already way over here, but the topic itself is highly interesting, which is why I am in favor of it being outsourced.
It's also not just the SDS800X HD, other scopes have the same behavior, so it's time it got its own thread.

markone:

--- Quote from: ebastler on May 11, 2024, 11:09:22 am ---Screenshots of the full 1 GHz spectrum and of my recreation of markone's settings are attached. The input signal was nominal 50 MHz from an (upgraded) SDG2042X.

The highest spurious peak in the "folded back" setting is -30 dBm as well, as in markone's measurement. But in the full 1 GHz spectrum, all harmonics are below -50 dBm. What's happening there?

--- End quote ---

Let's put it simple : the selected RBW values are not working correctly, it's useless to make comparison between utterly wrong spectrum images, if you want to get something similar to reality @ 50MHz CF / 1MHz SPAN, set RBW minimum at 3.36KHz (as I did in the "clean" spectrum that I posted).

markone:

--- Quote from: Martin72 on May 11, 2024, 01:23:41 pm ---I have now pressed the "Moderator Button" again, we are already way over here, but the topic itself is highly interesting, which is why I am in favor of it being outsourced.
It's also not just the SDS800X HD, other scopes have the same behavior, so it's time it got its own thread.

--- End quote ---

Do you know a "fast" way to dump the complete sampling buffer ?

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod