Further Discussion of the Pneumatic-Electric Interface
I have had two requests for a more detailed discussion regarding the
pneumatic-electric interface that I described in last night's MMD. The
information that was requested is for the sources of the materials and
the mode of operation. I am happy to provide the desired information.
MATERIALS AND SUPPLIERS.
The optical switch that I used was HEI Part No. OS-521-200, which
I chose because of its unusually wide gap between the arms of 0.200
inches (5.08 mm). This spacing can accommodate the standard 5 mm
diameter NMR (Nuclear Magnetic Resonance) Spectroscopy tubes used by
organic chemists for determining the molecular structure of organic
materials. The address of HEI is
1495 Steiger Lake Lane
Victoria, MN 553386
I obtained a few sample NMR tubes from Wilmad Glass Company at no
charge for my investigation. The address of Wilmad is
Wilmad Glass Co.
Route 40 & Oak Rd.
Buena, NJ 08310
Both of these address are from my records, which are by now quite old,
so both the addresses and telephone numbers may have changed.
The HEI optical switch was somewhat expensive, so I had a discussion
with Wilmad about the possibility of their manufacturing special glass
tubes for my project. The goal was to have the tube fit within the
0.150 inch gap of an Optron OPB-804 optical switch, which I had used
successfully in the Boesendorfer SE project, and which I had in
quantity. After a lengthy discussion, Wilmad and I agreed on the
Inside Diameter (i.d.): 0.1205/0.1195 inch
Outside Diameter (o.d.): 0.1505/0.1495 inch
Length: 1.253/1.247 inch
Straightness: 0.001 inch
Chamfer i.d. and o.d. 0.010 inch x 45 degrees
These tolerances, although they appear to the novice to be quite tight,
are in fact common in the glass tube industry, and present no problem
in manufacture. The i.d. specification in particular is important so
that the clearance between the Delrin cylinder and the inside of the
tube will be well controlled. Wilmad never manufactured these tubes
for me because I abandoned the project before the tubes were required.
In testing the single model that I made, I used a Wheatstone Bridge
(also called an "H") configuration. Three of the arms of the bridge
consisted of ordinary cup bleeds, forced into a short length of 5/64
inch diameter rubber tubing ("tracker bar" tubing); the fourth arm was
connected to the tracker bar port. I used a vacuum supply of about 20
inches of water. I operated the tube horizontally, although this is a
minor point since the pneumatic forces are much greater than the weight
of the cylinder. The Teflon-filled Delrin was so well lubricated that
the cylinder would slide from one end of the tube to the other if the
tube were inclined at a very slight angle with the horizontal when the
vacuum supply was removed.
Note that all the forces on the cylinder are pneumatic forces. I set
out to build a transducer that would not use springs, for the simple
reason that springs require constant attention. They sometimes work-
harden and get stronger; other times they get weaker, apparently
because of local stresses. Anyone who has ever used a music roll
reader that incorporates springs can tell you that you must pay
constant attention to the spring forces, a maintenance procedure that
I wanted to avoid.
My long-term plan was to build a full reader that would make use of two
regulated pressures, one about 8 inches of water, and the other about
20 inches. The first pressure would be applied to one end of each of
the tubes. The second, higher pressure would be applied only to the
bleeds. My plan was to adjust the pressures such that the "on" and
"off" transit times of the unit would match. The relatively slow "off"
time that we see with conventional pneumatic devices is the limiting
factor on the speed at which rolls can be read pneumatically; I hoped
to improve the speed sharply by using asymmetrical pressures.
There is one other idea that was never tried. I planned to investigate
using "laminar flow" bleeds, made from a length of capillary tubing
with an inside diameter of 1/16 inch (about 1.5 mm) or more. Such
bleeds should pass any paper or dirt particles that can pass through
the tracker bar port itself, eliminating the possibility of clogging of
the bleeds. There is also reason to believe that there might be other
advantages to such bleeds. I never determined if this idea is work-
able; I hope that someone else will take up the investigation where
I left off.
As before, I hope this note is useful. With best regards,