|The stock circuit was NEVER intended to drive a signal into "the outside world", but rather into faders, assign switches, etc., hence, the module's output can barely hit +18 dBu at clip because of the limitations of a single-ended 24 VDC power rail.
My client did have a pair of Neve LO-2567 output trannies, however, so I concocted a push-pull driver circuit using NE5534 opamps, and that combo resulted in a second, balanced output that could indeed push +24 dBm (yes, into 600 Ohms) from the balanced output connector. (That's why there is a TERM switch, to add an 820 load R across that balanced output when the preamp is NOT driving a true 600 Ohm destination).
As mentioned, I used a pair of NE5534 chips, both connected as inverters. Stage 1 drove one end of the Neve output transformer's primary *and* the other inverter (running at unity gain), which in turn drove the other end of the primary. Hence, "push pull" since the ends of the primary were driven 180 degrees apart. The opamps were powered from the single-ended 24 VDC supply with a bias circuit to make the output pins float at 6 VDC. Since both opamps floated at 6 VDC, there was no need to use a coupling capacitor to prevent a DC difference across the primary.
Whenever possible, I run opamps as inverters...lower distortion in most cases. But, I have to "keep track" of the signal path's polarity, and thus end up moving transformer inputs or XLR pins around to keep the polarity ("phase") correct.
Biasing the opamps was required in the Audix project since there was only a single-ended 24 VDC supply available, and not the +/- power expected by typical opamps.
For instance, this diagram shows two different methods to power an opamp:
The power connections at the left/bottom supply the "bias" for a single ended supply, like what Audix modules required In fact, in the Audix schemos for version 3, they did that exact same thing for the 5534 used as the mic preamp.
output generated using printer-friendly topic mod. All times are GMT - 4 Hours