[Collins] Re: black beauty

[Dr. Gerald N. Johnson, electrical engineer]

Opinions vary about the benefit of the autotune transmitter for VOA.
Collins gave them away, then built only three more for Australia, to a
completely different design. Art decreed that the 821A-1 output network
would never be used again, and by the time I left, the only components
in common (other than wire and resistors) were the PA and modulator
tubes, the RF driver tube and maybe the RF driver tube socket. I suppose
we learned some in those three years of engineer time (while paying a
penalty per day for a year late delivery though VOA warehoused all their
transmitters for a few years because they didn't have the budget to make
room for the transmitters) to speed the 821A-2, that 821A-1 was an
expensive project. The kind of thing that would make the banks reluctant
to make loans.

A few years after I left, the BC and HF high power transmitter
department was sold to Continental.

The first big Collins computers were telecommunications projects,
routers and central offices, not to my knowledge IBM mainframe
competition. Though once you have a working CPU and memory, only the
peripherals and software need be different.

In a way it was fun designing at Collins (despite Art's close scrutiny),
if a circuit needed another part it could have it. Economics of design
was not practiced. Which takes away some of the engineering challenge to
create products in large numbers where saving a penny or two per product
could pay one's salary for the year. Right now that quest for the
cheapest miniature electrolytic (never very high quality parts) is
biting the hands that have bought them in rapidly failing assemblies,
sometimes explosively failing.

My first edition copy of SSB circuits, before Sabin, has two of the
three author's autographs. I have a similarly titled book from one of
the competitors. It has no useful information.

I rather think that Rockwell turned management over to the group heads
and department managers. I had expected massive replacement of
management because Art never let it make decisions, but Rockwell didn't
do that and the company progressed.

The key to successful DSP employment is achieving a wide enough dynamic
range in the A/D conversion to handle all the expected signals in the
front end pass band simultaneously. Its not quite possible to feed the
entire HF spectrum to one A/D at the moment and have that dynamic range.
DSPs exist, its the fast A/D that are the difficulty. From listening to
Rockwell Collins receiver designers speak, I'm not convinced the new
schemes receive as well as older quality analog techniques yet.

73, Jerry, K0CQ
-- 
Entire content copyright Dr. Gerald N. Johnson, electrical engineer.
Reproduction by permission only.