开云体育

On a Z405 SW ver. 1.4-156 will saving the config (Menu: Storage|Save Config) also save all calibration data including the tables and individual level corrections? At the same token, when "resetting" the calibration, will that reinstate the factory defaults on this category and delete all additional level correction entrees (if indeed they are all saved as taken?) ??

?

A second question: In the "expert config" are undocumented entries: "direct mode" and "linear averaging", what are they, where can I find a complete manual to do my RTFM?

I would imagine that the linear averaging means that when enabled (default ?) it will blend the level calibration (table) pieces where they are concatenated?

?

Still being new to this and currently trying to calibrate a little better (or perhaps purpose oriented) than the internal automatic calibration is doing.

?

Thanks in advance

Lutz

On Thu, Jan 23, 2025 at 10:05 AM, <lschafer@...> wrote:

On a Z405 SW ver. 1.4-156 will saving the config (Menu: Storage|Save Config) also save all calibration data including the tables and individual level corrections?

?

Yes

At the same token, when "resetting" the calibration, will that reinstate the factory defaults on this category and delete all additional level correction entrees (if indeed they are all saved as taken?) ??

?

You should never "RESET" a calibration as the current implementation of RESET is undefined and not updated for some time. I will remove that menu item

?

?

A second question: In the "expert config" are undocumented entries: "direct mode" and "linear averaging", what are they, where can I find a complete manual to do my RTFM?

I would imagine that the linear averaging means that when enabled (default ?) it will blend the level calibration (table) pieces where they are concatenated?

?

I've updated the WiKi menu tree for you:

?

--

Designer of the tinySA
For more info go to

Thank you Erik.

Regarding the level calibration for the higher part >5.34GHz, would a square signal with lower fundamental (e.g. 1.78GHz) containing a strong harmonic at exactly 5.34GHz be sufficient, or do we need a clean tone (mayby suppress the fundamental with a highpass)?

?

Thank you for the "reset" advice. Well unknowingly I used it a couple times... Alternatively using "clear config", will this for sure delete everything related to calibration towards factory defaults?

?

Thanks again, I admire your good work!

Greetings from Canada

Lutz

On Thu, Jan 23, 2025 at 11:10 AM, <lschafer@...> wrote:

Regarding the level calibration for the higher part >5.34GHz, would a square signal with lower fundamental (e.g. 1.78GHz) containing a strong harmonic at exactly 5.34GHz be sufficient, or do we need a clean tone (mayby suppress the fundamental with a highpass)?

?

Harmonic is sufficient

?

?

Thank you for the "reset" advice. Well unknowingly I used it a couple times... Alternatively using "clear config", will this for sure delete everything related to calibration towards factory defaults?

?

Clear Config resets everything back to factory defaults.

?

?

?

--

Designer of the tinySA
For more info go to

Erik, thanks again. I do have one more question, regarding the harmonic mixing part above 5.34GHz. I used the NanoVNA's generator's third harmonic (of 1.78GHz) at around -40dBm to run the calibration mentioned before. It succeeds but does not seem to be correct as there is a noticable jump in the noise floor and the amplitude of the tone jumps with it when going over this transition. So I manually corrected this (see corresponding attached plot). Shifting this second part of the spectrum by about -12dBm seemed to have fixed the problem, also for the LNA case seperately. Calibrations seems correct over at least a 50dBm dynamic range.

?

While understanding a certain amount of noise increase at higher frequencies, I am quite puzzled why it is so much. It seems to be rising with 10dBm/GHz until levelling out at around 9GHz with -40dBm in the hand-calibrated case. Is this much noise typical for the TinySA - or have I gotten a bad one?

?

I also would like to have a more smooth noise floor over the entire spectrum without these bumps. Can this be accomplished using the input level table?

?

?

Thanks in advance

Lutz

Lutz,

The purpose of the calibration? indeed to remove the jump in the signal when it switches to harmonic mode.

A jump in the noise is unavoidable as the signal path loss increases due to harmonic mixing.

The bumps in the noise floor are unavoidable as the internal signal path is far away from 50 ohm at these high frequencies.

?

--

Designer of the tinySA
For more info go to