Four amplifiers are swept with 6 second triangular sweeps to measure summing junction errors at the inverting input. The triangle wave is applied to the amplifiers with an inverting gain of -20. The supply voltage is +/-300V. The offset voltage is nulled out at the non-inverting input before collecting data. An inexpensive Data logging module DI-194RS was used to collect the data. These modules only cost $25 at www.dataq.com . A 20x attenuator and a 100x preamp scaled the voltage swings from the output and inverting input of the OPAMP under test. After data was collected from the various amplifiers, it was combined in Mathcad6.0 for display in these plots.
The Output voltage for one K2-W, one SK2-V, one K2-XA and a repaired and slightly modified K2-XA:
The inverting input voltage for one K2-W, one SK2-V, one K2-XA and a repaired and slightly modified K2-XA:
The SK2-V was the part of the last generation of tube opamps designed at Philbrick. It featured a fully balanced differential input triode Nuvistor pair. The nominal offset voltage for the SK2-V is 0V. But the nominal offset voltage for it's predecessors, the K2-W and K2-XA, is 1.5V. Typically, you need to apply +1.5V to the non-inverting input to null offset at the inverting input.
The plate voltages of the input 12AX7 triode differential pair differ by 150V. The intrinsic gain (mu) of these triodes is 100. The 150V imbalance at the plates, when referred to the input, causes the 1.5V offset.
This last parametric plot shows the transfer function for the four amplifiers. Note the non-monotonic transfer function for the K2-W. This non-monotonic transfer function is caused by the internal positive feedback. The K2-XA also has internal positive feedback, but, in these two examples, the positive feedback was not strong enough to cause the non-monotonic behavior.
The positive feedback in the K2-XA was manually calibrated at the factory for optimum open loop gain. My repair efforts succeeded in bringing a dead K2-XA back to life, but they did not bring this part to meet it's gain specification of 30k.
The SK2-V amplifier, like the UPA and USA amplifiers, does not seem to use positive feedback. The SK2-V also has the most linear and highest open loop gain, in excess of 100k. I have yet to trace out the SK2-V circuit. The SK2-V data sheet does not include a schematic.
The open loop gain spec for the K2-W is 15k, for the K2-XA is 30k and for the SK2-V, it is 100k.