[alma-config] Forwarded mail bounced by email reflector

[Bryan Butler]

all,

> Simon Radford wrote:
>  >
>  > -- The engineers need a firm specification on the maximum fiber run. If
>  > the configurations were already known, this could be derived from the
>  > road layout, etc. But the configurations are still being designed while
>  > the engineers already embarked some time ago on development of the fiber
>  > links. Section 7.4.2 states "The longest distance to an antenna will be
>  > 25 km." Subsequent sections also mention 25 km. If this is the project
>  > specification, please elevate it to a more prominent place.
> 
> to which Larry D'Addario replied:
>
> My understanding is that the 25 km maximum cable run is already a hard
> specification, not something that we must derive from configuration
> discussions.  

is anything in this project a "hard specification"?  what is the difference
between a "hard specification" and a "soft specification"?  i'm being a bit
facetious here, but it *is* hard for me to tell the difference sometimes.
these are probably well-defined, i just don't know what the definitions are...

i do know that dick's FO group is carrying 25 km as the max allowable
fiber run length currently.  i do not know whether it is impossible for
that to be increased.  is 26 km any different than 25 km?  what is the
real cutoff?  IIRC, there is this "long-run regime" (i probably got the
nomenclature wrong, but i recall the discussion during ron beresford's
talk at the LO/IF/FO PDR in socorro) which we want to avoid, but i thought
that kicked in around 30 km?

> If the maximum baseline length is 10 km, then the 25 km
> includes considerable headroom, although it could all be used up if
> circuitous routing turns out to be needed.  It does *not* allow for
> the possibility of significant extensions to the maximum baseline
> length.

well, i think the approach should be that if the scientists want to
have a 14 km diameter configuration, then we either need to figure
out a way to cable it so that the max run is < 25 km, *or* we need
to make a good case that longer runs should be allowed.  the approach
shouldn't be to say that because the max allowable fiber run is 25 km,
then that necessarily precludes all configurations with diameter > 10 km.

> I agree with Simon that the 25 km spec should be made more prominent
> in the PB, and that the various other references to the array size
> should be made consistent and precise.  At this stage, any proposal to
> make the maximum baseline longer than 10 km should be treated as just
> that -- a proposal -- until its benefits can be carefully weighed
> against its costs.

i would state it as: any proposal to have fiber runs longer than 25 km
should be treated as just that - a proposal.  the spec isn't on the
max baseline length (is it?), but rather on the max allowable fiber run.

i do agree that the costs should be weighed against the benefits for the
most spread out configuration.  one problem is that we've had wildly
divergent numbers in the cost estimates (one big factor is whether we
are required to meet the chilean standard on the power distribution
conduit, and roads are another example).  the benefits, as in all cases
when considering the science output of the array, are hard to define
in a concrete fashion.  this is the situation we continually find ourselves
in.

	-bryan