You find pictures on high-amp.de
Not of the builds internally it’s just photos of the cases unless I’m missing something.
The schematics posted are not looking good to me at all.
Here… on the Sirius V4.5 at first glance.
The power supply section appears to lack any fuses or inrush current limiting. The input stage around T51 and SA970 is directly coupled to the high voltage supply rail. Some form of input protection, like series resistors or clamping diodes, is needed. Bias voltages for the output stage don't have much filtering/decoupling. really need some capacitance to help stabilize the bias points and reduce noise. The output coupling capacitors C66 look quite small in value (10uF) - might influence low frequency response.
KgssHV in comparison has legit input protection - diodes (D11, D12) and series resistors (R46, R47) - so real protect the input stage from overvoltage conditions. This was completely absent in the Sirius design.
Kgsshv also has much Improved power supply decoupling. There are several LED-resistor combinations (like D1/R5, D2/R8 etc.) which act as voltage drops and provide better supply rail decoupling and voltage distribution.
Current sources for front-end: Q19A/B along with R48/49 act as constant current sources to bias the input differential pair Q17/18. This should give better linearity than the simpler resistor biasing in Sirius.
Cascoded input stage: Q15/16 act as cascode devices for the input diff pair, which improves bandwidth and reduces Miller capacitance effects compared to the simpler differential pair in Sirius.
Kgsshv has more extensive use of local feedback: Several transistor combinations like Q11/Q12/R17 employ local feedback to stabilize gain and bias points.
(also for builder purposes kgsshv has adj bias controls because the potentiometers RV1-4 allow for fine tuning of bias voltages.)
Yeah… people need to be careful of just going off a random review of an amp sounding better. These high voltage amps have legitimate risks.