Herman,
If there were not a separate buffer circuit in the preamp to keep the SE output electrically separate from the balanced output, you could not use both outputs at the same time. With Krell, Levinson, ARC, and most other premium preamps, you can use the SE and the Bal at the same time because one (or both) are buffered. If the Krell is Bal circuit topology to begin with, then it's likely only the SE output is buffered.
Optimum signal transfer in Bal mode (compared to SE) occurs because the amp is receiving (from the preamp) a full strength push-pull signal around ground. A SE input (at the amp) must first go through an inverter circuit which takes the SE input signal, divides it into two SE signals, inverts the polarity of one of them, then recombines both parts as a balanced signal which the amp circuit has to see. But that "manufactured" balanced signal won't have either as low a noise floor or as high a signal strength as an actual balanced input signal coming straight from the preamp.
So with a (stronger) Bal signal at the input, the amp uses less (possibly half as much) gain to produce the same output voltage as when using a SE input. That results in greater headroom (less distortion at a given ouput to the speakers) than when using a SE input. So it's not just about less noise but better quality as well.
Checking the output specs for preamps that offer both types of outputs, you will notice the Bal outputs are usually rated at around twice the voltage as the SE outputs.
.
If there were not a separate buffer circuit in the preamp to keep the SE output electrically separate from the balanced output, you could not use both outputs at the same time. With Krell, Levinson, ARC, and most other premium preamps, you can use the SE and the Bal at the same time because one (or both) are buffered. If the Krell is Bal circuit topology to begin with, then it's likely only the SE output is buffered.
Optimum signal transfer in Bal mode (compared to SE) occurs because the amp is receiving (from the preamp) a full strength push-pull signal around ground. A SE input (at the amp) must first go through an inverter circuit which takes the SE input signal, divides it into two SE signals, inverts the polarity of one of them, then recombines both parts as a balanced signal which the amp circuit has to see. But that "manufactured" balanced signal won't have either as low a noise floor or as high a signal strength as an actual balanced input signal coming straight from the preamp.
So with a (stronger) Bal signal at the input, the amp uses less (possibly half as much) gain to produce the same output voltage as when using a SE input. That results in greater headroom (less distortion at a given ouput to the speakers) than when using a SE input. So it's not just about less noise but better quality as well.
Checking the output specs for preamps that offer both types of outputs, you will notice the Bal outputs are usually rated at around twice the voltage as the SE outputs.
.