Did a complete overhaul on this 333A. New Al- and Ta-electrolytics, many new transistors. The goal was to achieve like new performance. I have many of the HP special BJTs including the Germaniums from this era.
The 333A and 334A were popular distortion analyzers from 1968 to about 1980. They're actually very useful for measuring THD+N in tube amps but good luck trying to find one in good working order. They're straightforward to rework but getting the best out of them isn't so simple.
The key to getting these units to operating "like new" is the voltmeter circuit, it must measure accurately to the micro-Volts level and that's not so straightforward.
How does the analyzer work? A signal is applied to an amplifier to be tested for distortion, say a 1000Hz sine wave signal (the fundamental). The amplified signal is then sent to the analyzer where the operator "tunes" a Wien bridge to "notch out" the fundamental leaving the harmonics and noise for measurement by the meter circuit. The accuracy of the analyzer is governed by the accuracy of the voltmeter and the accuracy of the meter circuit over the bandwidth of the instrument (5Hz to 600kHz).
Below is the signal source, a 100uV, 400Hz sinewave signal which requires specialized instruments to source.
And the measurement of the 100uV sinewave at the 333A is below. The meter range knob (lower right) is set at .0003 which is 300uV or about 1/3 of a millivolt at full scale. The magnitude of the voltage is read at the 0-3 graduated scale. Note the pointer is at 1 or 100uV.
The 333A and 334A units are useful to about 0.03% THD+N which is about 30X higher than current state of the art units, like the Krohn-Hite shown above which is capable of 0.001% THD+N.