Doug Coulter has contributed a tremendous amount of documentation to this site. He lent me a Philbrick binder that he inherited from his father David Coulter. I scanned this binder into 512 scans and uploaded over 90% of these to the archive. The remaining 10% was redundant material. Of particular interest is the zipped file with journal article reprints showing early application of Philbrick opamps in science.

Doug Coulter (son) was exposed in his youth to the wonders of the operational amplifier, as he helped demonstrate with his father David, a speech synthesizer that employed Philbrick opamps at the 1964 world's fair in New York.


The following is an excerpt of an email interchange with Doug Coulter: on

Mon, 19 Mar 2007 12:37:42 -0400 

From: Doug Coulter to Joe Sousa

> I also plan to include a short reference to you and your father, as

> owners of the original material. The thought of drawing a line from

> early Philbrick sound processing gear through your father and through

> you has crossed my mind. I could attribute some of the deepest roots of

> your DSP work in Philbrick land.

I'm not real worried about credit, to be honest, but don't mind either. Like I said, the coolest thing I did with them [Philbrick opamps] was that playback device for analog signals hand drawn on velum with conductive ink.  And oh yeah, keep it all from going to the dumpster.  Some of the nifty stuff went to the NSA museum which could be linked in (including the playback chart recorder).

It might be even cooler to mention that I will be using them STILL IN THIS DAY AND AGE as the very best solution to certain classes of problem.  When I'm doing my very high energy Z-pinch sorts of things (kilo amps at kilo volts, billion watt peaks) I have no end of troubles with solid state stuff, even very robust stuff (think 40 lb lambda power supply), just going up in smoke, even though I try pretty hard to control/contain the emp and the SS stuff isn't even hooked in, just sitting nearby.

However, the philbrick stuff lives fine, it takes lots of energy to wipe out a tube grid.  Tiny parasitic inductances between any surge protector and any SS are enough to render the protection invalid.  And most surge protectors take some nanoseconds (at least) to work, and this EMP is lots faster than that (sub ns risetimes).  Just haven't come up with anything SS and even moderately sensitive that can match a 12ax7 for sheer robustness.


+++++++++++++ a few days later, Doug added this :

Date: Mon, 21 May 2007 14:42:49 -0400 

It's OK as is, but the analog memory/chart reader thing probably deserves a little more description.  I'm trying to get an original picture of it in high quality too.

In case I didn't describe it well enough, what is done is to draw your graphs on velum, with conductive ink.  Using the Philbrick stuff allowed use of aquadag instead of something costing much more.  The ends of your traces are brought to contacts at both ends of the chart (to be paralleled to reduce impedance).  The velum was stretched between these contacts in the version that worked the best, otherwise unsupported. 

You then roll a wirewound resistor along the chart, with some voltage impressed across it (I think we used about 20 volts).  We used the Philbrick stuff to buffer, level shift, and scale the results.  At first we tried regular power resistors modified so the windings were available, but this didn't work very well, so we wound our own out of thin copper magnet wire and just sanded the insulation off the outside only (so the turns didn't short).  Thinnest wire we could get wasn't really thin enough not to get warm in use, but hey, there was basically no other analog memory in many channels available that could be *edited*, at the time.  These you could edit with an X-acto knife and a paintbrush.  Try that with an FM instrumentation recorder!  For this use, voiceprints were read by eye and the important data transferred to the memory, also by eye, then by ear after listening to the synthesizer.

Eg it was a research tool.  Once you got the synth sounding very natural, you could go to work on the analyzer to get it to produce better data.  The chart that was used at the worlds fair might still be around, it was done with silver paint to last through the fair, and be a little better performing than the aquadag.

As I thought of the thing, and Dave worked for Melpar at the time (they were patent crazy), Dave had them put my name on the patent too -- at age 8.  Can't really take that much credit for it, as obviously at that age I was not the one who actually did most of the work to make it fly. At that point I was just up to building 5 tube radios from plans, and had just gotten my first transistor as a birthday gift (a ck722). This was sometime around 1961 or 2.  The fair wasn't until later on, by which time the thing worked pretty well.