I'm trying to improve this amp by installing better parts. All resistors were standard 5% tolerance - replacing them with 1% metalfilms greatly improved soundstaging, and the highs sound slightly smoother (s for sibilance 
). I ordered some silmic caps to replace the existing bypasses (standard Elnas), and BB2604 opamps to replace the JRCs in the preamp.
It appears that a lot of global negative feedback is used in the power amp stage. As a result, the amp sounds too controlled and cold. When paired with a good set of speakers that can go really low, the amp sounds hard in the low range and slightly aggressive in the upper mids. With sealed bookshelf speakers the bass lines totally disappear and it sounds sterile. I was wondering if adding emitter degeneration resistors to the input pair, thus decreasing the OL gain and NFB, would loosen up the sound a bit and maybe give it a touch of warmth?
The output bias current was originally set to less than 5 mA. Cranking it up served no purpose - another proof of excessive global NFB. The amp actually sounded worse with the increased bias, with even more emphasized highs.
The resistor at the input is definitely 150KR. It seems to me that R110, along with the 820R resistor in the NFB leg, is connected parallel to the input resistor to bring down the input resistance to a reasonable value. Or am I oversimplifying things?
). I ordered some silmic caps to replace the existing bypasses (standard Elnas), and BB2604 opamps to replace the JRCs in the preamp.It appears that a lot of global negative feedback is used in the power amp stage. As a result, the amp sounds too controlled and cold. When paired with a good set of speakers that can go really low, the amp sounds hard in the low range and slightly aggressive in the upper mids. With sealed bookshelf speakers the bass lines totally disappear and it sounds sterile. I was wondering if adding emitter degeneration resistors to the input pair, thus decreasing the OL gain and NFB, would loosen up the sound a bit and maybe give it a touch of warmth?
The output bias current was originally set to less than 5 mA. Cranking it up served no purpose - another proof of excessive global NFB. The amp actually sounded worse with the increased bias, with even more emphasized highs.
The resistor at the input is definitely 150KR. It seems to me that R110, along with the 820R resistor in the NFB leg, is connected parallel to the input resistor to bring down the input resistance to a reasonable value. Or am I oversimplifying things?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.