Threshold S150 bias adjustment problem

I have an S-150 Series II that I've used in an Apogee Duetta Signature bi-amp configuration to power the MRT ribbons for a number of years now. A few years ago, I replaced/upgraded the large power supply electrolytic caps with a new pair of United Chemi-Com 22,000uF units and replace the two small 'lytics with Blackgates and the the two tants with WIMA film caps. Despite its small size, I've always found this amp to sound very good - and it still does.

The other day I went to tweak in the bias for the first time since the cap replacement. No problem bringing up the bias voltage to 145 mV across an emitter resistor on the left channel; however, when checking the voltage across any emitter resistor on the right channel, turning bias adjusting pot all the way clockwise would yield only about 65 mV.

Since 65 mV is way too low and the specs call for more the twice this voltage drop across an emitter resistor, does anyone have an idea where I should start checking to find cause of this problem? BTW, the bias pot measures out to be good and touching up the solder joints on the driver board made no difference as well.

Since the S-150, S-300, and S-500 all share the same topology, I figure that someone here may have run in to this problem at one time or another.
 
As it turns out, I ended up testing all the 26-or-so-year old A42 and A92 transistors on the driver board using my cheap Hfe tester and found the values to range from a low of 5 to a high of 183 with most being under 65. I ended up Hfe matching (Hfe=100 to 106) and installing a set of new MPSA42 and MPSA92. Low bias problem solved!

The S/150 is now biased droping about 200mV across an emitter resistor for a heatsink temp of about 43 to 45-degrees C. Sound is even better than before and you can warm coffee on it now. ;)

I know that Hfe matching isn't exactly the last word in transistor matching, but in this case it got me in the ball park.