[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: [PSUBS-MAILIST] Through Hull Connections



Hi Jeff,

thx for this post.  the words themselves imparted some great common sense, 
but it was ur style that i liked :)

it had me laughing out loud - and Ive been down in the dumps the last few 
days, again thx.  gotta get myself one of those bill gates buttons :)

as i hope is being picked up from my posts - i will explore ANY idea, even 
if it appears at first to be crazy - some of the best ideas start that way 
(eg? light has a speed????)

The KISS principle is a favourite of mine.  i had originally planned a two 
belts and braces approach.  I figure manual controls and switches, also 
operable electronically and hydraulically - 3 completely separate, 
redundant and isolated systems.  that fitted with a safety factor of 3 that 
i was planning for the pressure hull.  I moved away from that a while back 
and felt that the decision many aviation designers have been going with to 
bite the bullet and run it all electronically was not so foolhardy.  Now 
these may be control structures that r manipulated at supersonic speeds - 
far beyond the capabilities of any human responses.  I likened the deep to 
a similar hostile environment.  I need to seriously rethink this approach.  thx

I intend putting a great deal into the research and design aspects and i am 
still way down on the learning curve.  I'll take this moment to express 
thanks for all the useful info and input everyone in this group has provided.

cheers
peter

At 08:20 5/1/03 -0700, you wrote:
>Hi Peter,
>
>I read Dan's response to this one, and it is right inline with my comments. I
>don't think you should complicate things with networks of computers and
>instruments. Dan is right to say that you would be best to manually control
>everything, at least to start. If you are a sophisticated electronics + 
>computer
>+ software person, you could expand later. You always need to worry about how
>you will operate when things fail. You alluded to this yourself with the air
>scrubber and a power failure.
>
>I have this vision of you being negatively bouyant, getting the blue screen of
>death, and it's one of those times when, somehow, the reset switch doesn't 
>seem
>to work. One of my favorite lines was someone saying he wanted a picture 
>of Bill
>Gates on the reset switch so he could punch him in the face every time 
>Windblows
>crashes!
>
>Basically, the simpler you can make things, the more likely you'll sucseed in
>building a psub project, and the more likely you'll be able to use and survive
>your project. It is very much like boats, the more crap you add on, the more
>time you'll spend installing and repairing, and the less time you'll spend 
>boating.
>
>
>Jeff
>
>
>Quoting mckellar@earthlink.net:
>
> >
> > Obviously I will need some power inside the hull, but I should be able to
> > keep the demand fairly low.  The thing that concerns me the most is the
> > requirement for over 100hrs life support.  without something to pump air
> > over the scrubber and monitor and control essential systems, any other
> > 'failsafes' (eg lots of air/oxygen) become next to useless.  do i put a
> > bicycle type driven fan or generator on-board? (i jest, but it has some
> > merit).
> >
> > Which brings me to another question.  I'm assuming with joystick control,
> > various sensors, cameras etc more than one computer is really required for
> > control etc.  Is the solution to make everything dumb except the laptop/PC
> > or do you run a lan?  if a lan, wireless or cabled?
> >
> > If cabled to things outside the hull, the same sealing problems occur.  if
> > wireless, that avoids thru hull connections, but how far is the signal
> > likely to transmit through saltwater (given that different manufacturers r
> > likely to have different broadcast strengths)?  then of course i have
> > to   have a client out-board to control thrusters, trim etc
> > <sigh>.  <warning: hopelessness and despondency follow :) > so many
> > problems at every turn :(  my speadsheet with specs is huge and still
> > nothing I'm really happy with.  The AutoCad 2000 solid modelling im
> > dabbling with is like a kids scribble :(
> >
> > cheers
> > peter