Just a few suggestions:
1. R1, R4 should preferably 1K/2W CFRs. Yours look like MOXs.
2. C4 should be an MKP or better, with short leads as mentioned earlier. K72s are great signal coupling caps, but not necessarily good for bypass.
3. The mute resistor should be 10k, if it isn't already.
4. A 10nF FKP2 is hard to beat at C7, though further experimentation is possible.
5. An FM at C9 is OK, but you should try a BG STD, BG PK, Silmic II or Nichicon KZ at that location.
6. C21 should be MKP/FKP or better - MKS/MKT will hurt sonics.
7. C6, C11 can be FMs, but try out Silmic IIs/Muse KZ/BG PK also.
8. C13 (if installed) should be a 1uF or higher polypropylene film/foil like a Sonicap Gen I or better. K72s will probably do well at this location.
9. The compensation network caps - C10, C32, etc. should be FKP2 or Silver Mica. Stay tuned for a possible upgrade from Dario.
Edit:
10. R3 should be non-magnetic/non-inductive. Go with Caddock (Ultimate BoM) or Fukushima MPC74 (my premium BoM).
Last edited:
hi siva most of what you mention have been implemented in my boards already. the symptom remains, though, which's why i'm so frustrated. weeks of dedication and still broken. on the flip side, sound staging is whole lot better than the gainclone or any other amp that i heard for that matter, though. if the drive matches up to it, i can see this being that giant killer of every diy-er's dreams.
any common issue that could make 3886 run cooler and dynamics muted? the sound feels like it's being bottle necked somewhere. detail and resolution is there, mids feel rich and harmonic but everything is subdued and closed in.
any common issue that could make 3886 run cooler and dynamics muted? the sound feels like it's being bottle necked somewhere. detail and resolution is there, mids feel rich and harmonic but everything is subdued and closed in.
Last edited:
The LM3886 runs very cool in a properly implemented MyRef, because it's very stable, with huge phase margin (90-100 degrees). If it runs cool, it probably means that your compensation network is OK.
Check your input source and attenuator - a 10k-25k pot is better for dynamics than higher impedance pots.
Check the rail voltages near the LM3886 - +/- 32V is close to optimal.
Recheck all component values anyway. The kit may have been correctly constituted, but parts may still have been incorrectly interchanged, etc., during assembly.
