Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!henry
From: henry@utzoo.UUCP (Henry Spencer)
Newsgroups: net.space
Subject: Re: spinoffs
Message-ID: <6169@utzoo.UUCP>
Date: Sat, 23-Nov-85 20:08:15 EST
Article-I.D.: utzoo.6169
Posted: Sat Nov 23 20:08:15 1985
Date-Received: Sat, 23-Nov-85 20:08:15 EST
References: <8511191259.AA24626@decwrl.DEC.COM>
Organization: U of Toronto Zoology
Lines: 79

> Frying pans are often cited as a trivial spinoff, but they are not even that.
> Teflon was actually developed by Du Pont in the Fifties, long 
> before Apollo needed tough plastics with high melting points.  

Quite correct, commercial Teflon was basically a spinoff from the Manhattan
Project, not the space program.  The Manhattan Project did the original
basic engineering needed to turn a laboratory curiosity	discovered in the
30's into a useful material.  DuPont finished it up for commercial use.
Possibly the space program may have done some work on one of the Teflon
variants (note that the original Teflon was not useful for frying pans,
since it wouldn't stick to the pan!) and hence gotten the story started.

> Surely, though, ICs are something major?  Yes, and in the extremely
> early days of the early sixties the space program did have an effect
> on their development.

Yes and no; ICs were a spinoff from ICBMs more than from the space program.

> They set the initial standards for temperature
> and mechanical stress that the IC makers had to meet, and also
> provided a small but steady market for them.

If you check, I think you will find that the original development of the
first practical ICs was military-funded work for the Minuteman ICBM.  So
the government role was a bit more central than just being a demanding
customer.

> ...  NASA landed men on the Moon using only
> resistor-transistor and diode-transistor logic, not even TTL.  Last I
> heard they still relied on DTL for their electronics.  Several
> generations have come and gone since DTL was introduced: regular TTL,
> PMOS, and enhancement-only NMOS. NASA is continually stuck with using
> obsolete parts because it takes so long to develop anything, and
> because its needs are so different from those of the mainstream
> markets. 

Don't forget a couple of other major reasons:  (1) NASA has been a little
bit short of funds lately, and (2) NASA's reliability requirements are
such that they can't risk using something hot out of the development groups.
The latter is not unique to NASA; you'll find the same phenomenon in any
environment where major mistakes are unacceptable.  The Bell System invented
the transistor, but it was a good many years before solid-state electronics
showed up in telephone switching systems.  When your equipment is supposed
to work for 40 years, you *can't* use a part in critical applications until
a good reliability database has been built up for it.

NASA landed men on the Moon using old IC technologies because most of the
Apollo hardware was designed in the early 60's, before TTL.  Remember the
lead times involved; Apollo incorporated new technologies only in areas
where it didn't require serious redesign.  Apollo hardware had to be ready
to *fly* in about 1967, which meant that a lot of decisions had to be made
very early indeed to permit adequate development and testing.  For example,
the VAB is bigger than it needed to be for the Saturn V, because the size
of the building had to be fixed before anyone was sure how big the booster
would be.  For another example, the Apollo SM engine was powerful enough to
lift the CSM off the surface of the moon, because its specs were fixed too
early for anyone to be sure that this would be unnecessary.  Remember that
NASA was running final tests for the first Apollo flight, with the lunar
landing (optimistically) hoped to follow in a year or so, a total of six
years after Apollo was ordered.  A rather tight schedule, given that nearly
every piece of hardware they used had to built from scratch.

> ...  A research 
> program might develop something unique and innovative, but it rarely 
> gains market acceptance because it is specialized to the needs of 
> that program...

But the underlying technology is another story.  Nobody is contending that
the precise parts used for Apollo were good for much else.

> The bottom line is that if you want your research to 
> be of commercial value, it must be directed to commercial needs.  
> Military and space work has commercial value largely by accident.

If so, then there have been a lot of rather lucrative accidents.  Enough
to make them fairly predictable, in fact.
-- 
				Henry Spencer @ U of Toronto Zoology
				{allegra,ihnp4,linus,decvax}!utzoo!henry