From
John Atkinson's measurements of the SLP-05:
The SLP 05 offered slightly different maximum gains from the figures specified on Cary's website. However, the specified gain is different on the Specifications page10dB single-ended, 8dB, balancedfrom what it is on the Literature page12dB. In a telephone conversation AD had with Cary's Dennis Had, Dennis said that the 10dB/8dB numbers were a misprint, the real numbers were 16dB unbalanced and 12dB balanced. With both input-level controls and the volume control at their maximum settings, I measured 15.3dB for unbalanced input to unbalanced output and 13.8dB for balanced input/output. The difference, perhaps, is due to inconsistencies in the tube gains. Usually, such differences would have been minimized by negative feedback, but the SLP 05 apparently doesn't use any negative feedback.
This suggests two things to me:
1)You may be able to change the gain significantly by tube rolling.
2)However, there is no point in doing so because you most likely can already drive the amplifier to the clipping point, considering that the present gain is most likely several db higher than the 10db I used in my earlier calculation (which reflected 1.9V into the power amp).
Also, consider that the SLP-05 has a rated nominal output level of 2V, and is presumably designed to be compatible with your CAD120S. That would suggest that the unspecified input sensitivity of the CAD120S is significantly less than 2V, to assure that the SLP-05 can drive it to max power.
All of which seems to reinforce the comments Herman and Greg offered about needing more power (which would also provide more gain, assuming input sensitivity remained comparable). A step-up transformer obviously will not help in that regard; might result in overload problems as has been indicated; might degrade the sound quality that you are presently happy with; and would result in your having to address and experiment with a lot of complex interrelated variables that sut's inevitably involve, if sound quality is to be fully optimized.
But there is another significant constraint involving increasing power. Your Harbeth's are
rated to handle a maximum of 200W on program peaks (their continuous power rating is probably considerably less), and 200W is only about 2db louder than 120W.
Note to John: Thanks for pointing out that the balanced output gain of the SLP-05 is spec'd at 2db less than for the unbalanced output. Actually, I had noticed that, but I was envisioning that the power amp's balanced input MIGHT have a 6db higher sensitivity (lower sensitivity number) than the unbalanced input.
Markus: So I'm not sure what the answer is, but those are the thoughts that I can offer at this point. Ultimately, you are constrained by the peak spl capability of the speakers.
Regards,
-- Al