Some of my worK benches at Dolby Labs.
 This bench was fairly well evolved,  with an Audio Precision System 1 Dual Domain, 4-Chan digital storage scope,
2-Chan analog scope, Ithaco filter, music center, PC and 2 lab supplies.

     The PC in the center of this picture was probably supplied
by our licensee with the sample to be tested.

     However the manufacturer of the processor may have sent
an add-in card and some demo control software to allow our
testing.
    
     When the PCs were supplied by our Japanese licensees,  
they most likely ran Windows in Japanese,  as not in English.  


     Windows makes it relatively easy because the format of the
error messages is very similar to the English version.

     We had an office in Tokyo,  so a screen capture of a
Windows Message Box in Japanese could be translated by
the time I got my email the next morning.

     I communicated directly with our licensees and would relay
my test results or problems from my day,  before I went home.  

     Japan being about 17 hours away meant that I usually got a
reply from the design engineer by the time I returned to work
the next day.
Circa '97
Same bench as above.
                    This is a Hitachi chip.

     This test setup is for the normal test equipment.

     This chip is what I called a "jungle chip" because it contains
more than one audio function block.

     This is most probably a cassette B/C-type noise reduction
processor for car audio products.

     It would contain the tape head switching,  preamp and playback
amp with tape EQ switching followed by something like a Bass Boost
circuit.  
Same bench as above.
         Same NR processor as above.

   Here the test setup is for the "special" test equipment,  my ears.

     The acid test was the listening tests.

    At the end of several days of bench testing,  I would spend
several more hours just listening to mostly music processed by
the IC.  
   
   I had collected a list of music cuts that I knew could cause
problems with  compander type noise reduction processors.  

   Passing the listening tests was just as important as any of the
bench tests.
Same bench as above.
 Another view of the listening test setup.  It
usually took me about 3 hours to connect and
align the listening test setup.  The level reference
for the IC was probably 245mV.  The reference
NR processor was a Model 422 running at
professional levels of about +4dBu.  I kept the
level matching and normalization to within 0.1dB.

 All of the signal switching and level matching
buffers used here were tools that I build up over
the years to reduce test time.
     This is a later evolution of my basic test
bench setup.  Here I have added an AP
System 2 Dual Domain along with a second PC
and another of my AP Switch Boxes.  
 I think the processor being tested is a
Mitsubishi Pro Logic decoder.  The left
screen, (AP Sys 1),  shows an auto balance
capture and correction ratio test.  The right
screen, (AP Sys 2),  shows a steering time
constant test.
    At the left  : Another later bench setup.

     The main addition to this setup was the addition of 2 of my bit stream switches.

     This gave my test programs the ability to select multiple digital signal sources and test points to be monitored.  
     Both monitors sit on 3-ring binders, old manuals.

     This was OK for the smaller 15" on the left.

     The newer monitor on the right was a tank and rather imposing from bench top level.

     Being San Francisco,  I secured the monitor's case in several locations with 1/8" nylon rope to the metal frame of the
bench.  It wasn't going anywhere.
 The left screen is showing a
command line driver used to control
licensee's software that only use a text
type command line.  My software
allowed assembling a text command
line by clicking text boxes or buttons
and concatenated a text command
line.  The text command was sent to
the OS via a Shell command.

 The right screen shows a dynamic
range compression test used for Dolby
AC-3, (Dolby Digital).
Same bench as above.
 Reverse angle of above.

 Far left are my 2 main test PCs.  Each contain a
24-bit Digital Audio sound card and are accessible
with the business end where is should be, in front.  

 Two bit
stream switches,  one just below the System
1 that are the sources from the two APs and sound
cards.  
 The other switch located between the two key
boards can select one of 4 optical or electrical
AES/EBU or S/PDIF bit streams and sends them to
both APs.

 At the upper right of the left side keyboard is a small
push-button switch.  I goes to a KMV switch and allows
me to use one track ball with both monitors.  This was
later expanded with a second switch so I could use
just one keyboard.

 As well as having program control of my bit stream
switch,  I used the AP Sys 1 PC to hold several Gig of
test bit streams of various formats.  My test program
running on my primary PC could control the playback
of bit streams from my second PC via a TCP/IP.  The
typical delay between the primary PC issuing a
playback request and the bit stream arriving at the AP
from the second PC was 140ms.  This allowed the
program to start both the playback and measurement
of a bit stream.

 The
bit stream switch allowed me to make 8 channel
FFTs in about 47 seconds.
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
100 Potrero
1st Floor Rear
100 Potrero
1st Floor Rear
100 Potrero
1st Floor Rear
100 Potrero
1st Floor Rear
100 Potrero
2nd Floor Rear
100 Potrero
2nd Floor Rear
100 Potrero
2nd Floor Rear
100 Potrero
2nd Floor Rear
100 Potrero
2nd Floor Front
100 Potrero
2nd Floor Rear
100 Potrero
2nd Floor Front
100 Potrero
2nd Floor Front
100 Potrero
2nd Floor Front
100 Potrero
2nd Floor Front
100 Potrero
2nd Floor Front
100 Potrero
2nd Floor Front
100 Potrero
2nd Floor Front
100 Potrero
2nd Floor Front
           This is an AP Switch Box
            (The backgrouond of this page is the schematic of this box.)

   I designed and built them to expand the Analog I/O of our 2 channel APs.  

   This switch expands each of the AP inputs to 9 and doubles the Aux inputs.

   It provides an "External Equipment Loop" with selectable sources and targets for the "sends" and
"returns".

   There are signal "Bypass" for both the inputs and outputs of the AP.

   It adds a standard audio system and ground to signal sources.

   Connections and switching for an oscilloscope and audio power amplification for monitoring
purposes.

   The AP signal generator monitor and sync signals are now available on the front panel.

       More details are to the right,      -> -> -> ->.

     Photographs of the fabrication of this box can be found here.

A similar device for expanding the Digital I/O can be found here.
      At the left, is an AP I/O Switch Box.
          (The backgrouond of this page is the schematic of this box.)
 I designed7 and built them to expand the analog I/O of our APs.  

 APs are basically 2 channel devices.  

 This switch expands each one of the 2 AP inputs to any one of the 8 input BNCs or
the return of and external equipment loop. The single AP Aux input is expanded to 2
BNCs.  

   The source for the external loop is selectable from either the input or output side of the AP. The loop
allows for the insertion of additional filters,  processes or test equipment into the measured signal path.  
The two rotary switches to the left of the "External Loop" BNCs allow selection of the signal being sent
out to the loop.

 There are signal "Bypass" switches for both the input and output of the AP so it can be easily removed
from the test signal path, (or even the bench), without having to disturb or change any of the wiring on the
test bench.  This makes a confidence check or service of the AP quick and easy,  as well as for
substituting  alternate equipment via the "External Loop".  Fast, quick clean and easy.

 The box also incorporates connections and switching for an oscilloscope monitor and a regular stereo
system, with a Tuner, and disk player as a source for normal audio program material as well as power
amplifiers, speakers and headphones.  

 Both the AP signal generator outputs and the stereo system signals are sent to the BNCs at the far left of
the box and selected by switch.  The source select switch is 3 position and can ground the input of the
DUT for critical noise
    

    
Photographs of the fabrication of this box can be found here.


  A similar device for expanding the Digital I/O can be found here.
Below is a schematic of the Digital
I/O expansion, signal routing and
switching.

The BitStream switches shown
below are pictured in photographs
at the left with the  label of
100 Potrero
2nd Floor Front
@ 721 Sansome

3rd Floor  1981  (start)
2nd Floor 1983


@ 100 Potrero

2nd Floor Rear  1985
3rd Floor Front
1st Floor Rear
1st Floor Front  
1997
2nd Floor Rear
2nd Floor Front  
2001
- 2004
(quit)