Larry -- Connect a 91K across the Mag 1 input. When it is selected, it will then represent a 47K load to the cartridge connected to that imput, while its addition will have no effect on the Mag 2 input loading characteristics.
Regarding the noose, as I have studied it over the years, I have found that it addresses a number of areas that Fisher could have been concerned about:
1. Over-driving the output stage is one of them. The noose largely presents that and in so doing, protects the tubes, and could also act to limit blocking distortion as Gadget suggests by limiting the current required to produce the blocking action.
2. Properly bias the phase inverter section -- which I now believe was their primary goal:
Fisher was trying to have their cake and eat it to with the driver stage: Use relatively low value inverter splitting resistors to help emphasize a low output impedance on each side of the push-pull signal, AND maximize the gain from the overall driver stage -- and all at minimum cost. Maximizing the gain meant elevating the plate voltage of the AF Amplifier section, and increasing the value of that section's plate resistor. But the higher plate voltage of the AF Amplifier section in combination with the low value splitter resistance values then underbiased the splitter section: Without the noose, the bias of the inverter section would often either be 0 or even go positive! The noose then lowers the current flow through the inverter section (by shunting some current around it), when then allows it to properly bias itself. And, in some versions, connection the cathode side inverter resistor back to the NFB insertion point also provided a small degree of POSITIVE FB, which then increased gain even more.
It all "works" but:
A. Leaves so little current flow through the inverter section that it leaves its operating point right on the ragged edge of even allowing the tube to operate -- let alone linearly, or with enough distortion free output to fully drive the output stage.
B. Being on the ragged edge of the operating point, it means that some tubes will work OK, while others (most often the Russian tubes, but also some American examples as well) simply will not.
It is easy to see when it doesn't work by checking the power output of each channel with a scope. When the noose is rearing it's ugly head, the clipping gets very lopsided and weird well before full power output is developed. I have seen it happen way too many times on way too many units. At low or normal listening levels, you don't know its there. But push it and it shows up in a hurry.
I still believe that the best cure is to remove the noose, THEN source the B+ for the plate resistor of the AF Amplifier section to the next lower B+ supply point (that used for the line amp/tone control circuits), and then readjust the phase inverter control for balance using the procedure I developed. That approach is what I first used in the improved power amplifier section as published in the improving the 400 thread, and has worked perfectly every time since in any application using the noose.
Dave