Hello eevbloggers, here is a frustrating issue which has kept us going for some days on which any comments would be greatly appreciated.
Recently purchased and received a SSA3021X Plus + RB3X25, our first Siglent SA. Due to previous modding successes with Siglent scopes, this unit was foolishly not even turned on, but instead immediately opened (tautech's comment about rushing in without exploring features refers) in order to add the missing RF connectors for SVA1032X improvement (complete careful disassembly incl. the gazillion screws), as per Attenbach & mandor.
Upon turning on the unit after careful reassembly, there seemed to be an issue. The Spectrum Analyzer mode showed an initial response with nothing connected to any inputs very similar to the following:

AutoTune yielded:

Zooming in showed that the raised low-F response consisted of ~171 kHz harmonics (some more prevalent than others) out to nearly 50 MHz. Instrument still had original F/W:
Model: SSA3021X Plus
Software: 6.2R10
Uboot-OS: 3.2.2
FPGA: 20240828-1001
CPLD: 000000D4
HW: 02.00.00
Proceeded to cross-flash to SVA1032X (first downgrading to 3.2.2.5.1R1, etc, following the guidelines in these forums) in the hope that a firmware update would correct / calibrate the issue, but to no avail. After initial 3.2.2.5.1R1 flash, unit reported versions:
SW1: 3.2.2.5.1R1
SW2: 20211011-1001
SW3: 000000D4
HW: 02.00.00
After options, realizing that 3.2.2.6.0.R10 is not yet available, subsequently flashed latest SVA1032X firmware 3.2.2.6.0.R5, which was downloaded via an earlier tautech forum direct link (siglent china 3.2.2.6.0.R5 firmware version seems to be the same, by the way).
Did 1-port calibration using Siglent Cal-Kit F503ME, but response remains seemingly unchanged.

The SVA1032X teardown seems to show different paths possibly handling the various spectral portions. Could one of these portions be damaged?
The double shielded RF section was reopened and carefully examined under a microscope. This showed several solder balls prevalent in the original Siglent PCB Assembly. Most or all of these were removed with careful subsequent reassembly, but the unit still gives the same response.
Does anyone have any advice or recomendations, or something we could try to improve the situation? Any help would be appreciated!