[time-nuts] HP Stories: Cesium Standards on Subs and Sperry

[jimlux]

On 2/24/19 5:37 AM, Rice, Hugh (IPH Writing Systems) wrote:
> While HP wasn't a direct defense contractor, we did sell a lot of test equipment to defense contractors.    The big American submarines had a Cesium Standard or two as part of their instrumentation systems.    I know little about the application, but heard it was part of the communication and/or navigation systems.    Maybe some of you have experience with this, and can add to the story.
> 
> None the less, Sperry corporation was a sub-contractor into the greater DOD eco-system, and integrated the 5061A into some larger system they sold to the submarine builders.    Sperry was a "real" defense contractor, and had to live by all the DOD rules.    There were a number of defense contractors in Silicon Valley, with Lockheed Missiles and Space Company being perhaps the largest employer in the area.   Both my father, and my wife's father were engineers at LSMC for their careers.   We used to joke at my High School that "everyone's dad worked for Lockheed."    Thus, DOD companies were not a foreign concept to me.   But they way the DOD procurement process worked was very unlike how HP worked and interfaced with our commercial customers.


Yes, and this how you wind up with $600 hammers/toilet seats/aircraft 
coffee makers.

It afflicts the space business too - We call it "heritage" - for 
instance, you're designing a new spacecraft.  A subsystem, or box, or 
anything, that already exists, you know the mass and power of, so you 
can put it into your "Master Equipment List" (MEL) and only add 5% 
margin.  Something new, might have to have significantly more margin, 
maybe 25% or 30%.

But on most spacecraft  you're severely mass and power constrained, 
because the rolled up budgets for those are often set by "what did the 
last one do"  - or maybe you've got the max mass you can send to Mars on 
the rockets that are available.

Even if the new design is infinitely superior in performance, the 
additional risk of "something new" gets in the way.  Most NASA 
Announcements of Opportunity (AO, similar to a RFP) say "Technology 
shall be at TRL 6, or a plan shall be presented to insure that it will 
reach TRL 6 by Preliminary Design Review" TRL 6 = "demonstrated in a 
relevant environment" - basically, you've built the box and run it 
through thermal vacuum and vibe testing, but maybe you used parts that 
aren't screened.

Notwithstanding that the parts list for that heritage box includes parts 
that are no longer available, or that have insanely high minimum order 
quantity (the mfrs way of telling you, "use another part, but if you 
insist") - that's what gets designed in.

And as Hugh points out, there's a whole team of people who look for ANY 
discrepancy from the original design. And those people aren't empowered 
to make judgement calls about whether the change is good or bad - it's a 
change.

One can, in fact, get waivers.  Or, you can, if you have sufficient 
budget (Hah!) you could say that your primary design alternative leading 
up to PDR is to use the Heritage Unit, but you'll be developing a new 
one to replace it, and if it's ready, you can pitch it at PDR (or maybe 
CDR).

However, it might be easier (from a risk assessment, budget, and 
schedule standpoint) to just use the old design. That is, the extra pain 
and cost for finding obsolete parts (Rochester Electronics, "the leader 
in the trailing edge of electronics" is your friend) is less than the 
pain and cost of getting waivers for a new design.