Page 73 - Analog Circuit Design Art, Science, and Personalities

Max Wien, Mr. He iwlett, and a Rainy Sunday Afternoon

Figure 7-17.
Common mode
suppression runs output lOV/DIV
distortion
analyzer into its Distortion
noise floor. (analyLer limited)
.0003%

A2 Output (AC
coupled) . 1 V/DIV

Horiz. =
200~secDIV

oscillator-pity poor Al, besieged by three masters) but does not adversely affect
its operation. With its positive input forced to virtual ground, A1 sees no common
mode swing. Williams’s Rule is satisfied, and ostensibly, good things should
happen.
To my utter amazement, this whole thing did not explode when I finally sum-
moned the nerve to turn it on. Even more astonishing was the distortion analyzer’s
0.0008% reading (Figure 7-17).
Its output showed only faint traces of the first harmonic outlined in noise. The
analyzer was indicating more than a factor of two beyond specification, which was
really asking a lot. While it’s unlikely that the oscillator and analyzer have compen-
satory errors, it’s dangerous to conclude anything. As such, I turned to some very
specialized equipment to get at the truth.
The Audio Precision System One will read distortion down to 0.0003% (3 ppm).
I was quite pleased to see that it couldn’t find anything above this level.
After Hewlett finished his oscillator, he and David Packard went into their
garage and built a few into boxes and then made some more kinds of instruments

Figure 7-18.
Bill Hewlett and
David Packard
building
oscillators at the
Hewlett-Packard
Company,
located in their
garage.
(Photo courtesy
Hewlett-Pac kard
Company)

54