sreten wrote:
(An option is changing the MC setting to only a little more gain.)
Er, no. That won't help SNR a bit since both the input signal and the
input stage's equivalent input noise remain the same.
Not wanting to steal this thread, but methinks a couple of simple hacks can improve the Cyrus One in MC mode:
1) replace or bridge R17 with something like 10 to 22 Ohms. The thermal noise of that 470 Ohms is huge!!! If you stick to MC, solder the proper loading resistor on top of R19.
2) leave the NE5534 in there. Contrary to popular belief it really isn't that bad. If you really want to replace, think OPA227 or 228, and if funds allow, LT1028.
3) replace C17 (1u) with the best-sounding cap you can fit in there. Probably increase value to 1.5u or 2.2u.
4) replace OA5 with something like the OPA2134, a FET opamp with low offset, allowing DC-coupling henceforth (really!)
5) bridge C23 and even C25 (but measure offset before you do so)(*)
6) remove C27
7) solder a 100nF ceramic cap between the V++ and V-- supply of each opamp

have a look at the power supply:

Replace C71 and C72 with 100-220uF quality elcaps with a fairly *high* ESR, such as Rubycon YXF or ZL.
Replace C69 and C70 with the largest-value low-ESR caps you can fit (100uF or so) and/or replace R115 and R116 with a 12V zener.
This reduces the output noise of the 317/337 and improves their stability.
Hope this helps...
(* Assuming that this doesn't throw off the low-end section of the RIAA too much ... have to simulate ...)