Strange problem with McIntosh C29 phono stage

Hi all, can I get a moment from the hivemind here? This one has me a bit mystified.

I've got a C29 with odd low frequency rolloff in one channel of the phono stage. In this model the gain stage can be set to phono 1, phono 2, or microphone. I've attached sweeps of all three modes, as well as a drawing. There's a significant dip around 30Hz in one channel but not the other. This dip exists in both phono 1 & 2 but not when set to mic.

Here's what I've done:
Replaced all the electrolytics with good parts, even went so far as to measure the new shunt feedback caps to make sure my new parts are OK.

Replaced the shunt feedback resistors with a hand picked matched pair of 51.1Ω resistors (4 wire ohmmeter).

Opamps are new 5534s, but the old ones showed the same behavior.

I've taken the opamps out of socket and verified the feedback network resistance from pin 6 to pin 2 - channels match.

Rails are fine at +/-18V

What am I missing? Thanks...
 

Attachments

  • C29 phono stage schematic.jpg
    C29 phono stage schematic.jpg
    67.4 KB · Views: 200
  • C29 phono 1 LF rolloff.png
    C29 phono 1 LF rolloff.png
    12.8 KB · Views: 176
  • C29 phono 2 LF rolloff.png
    C29 phono 2 LF rolloff.png
    12.9 KB · Views: 148
  • C29 mic input LF rolloff.png
    C29 mic input LF rolloff.png
    11.8 KB · Views: 160
The external feedback switching was the problem. The rotary source selector switch has a SPDT element to handle this; in mic mode pins 9 & 11 are shorted and in phono mode pins 10 & 11 are shorted. The bad crimp was on the wiper of the switch so in fault mode all 3 pins were open, an undefined state. Working well now.