[evla-sw-discuss] MIB Study - System

[Wayne Koski]

Kevin asked if we should define the System, rather then looking at
processors.

Because I'm a hardware type, I tend to look at hardware.  Also I thought
we are trying to locate a family of devices for use in the MIB, as
well as the Antenna Computer and Array Computer.

But OK, here is my ideas in easy terms of two possible systems:

1.  The MIB gest Ethernet messages (TCP/IP) and passes that information
    to the User Device.  Upon request or via polling it gets information
    from the User Device and sends it via Ethernet (TCP/IP) to the
    computer.
2.  The MIB is an intregal part of the User Device. It not only
    manages communications over Ethernet (TCP/IP), but the User Device
    operation as well.  It closes hardware loops, runs any User
    Software, Diagnostics, etc.

Currently the focus is on System 2 for the software group.  But the
hardware group may be looking at System 1.

My feeling is that the hardware group is going to steal from ALMA.
They are going to use whenever possible hardware designs, PCB's,
software, etc.  My understanding is this could be approximately
7-9 module types.

Now the way it works is thus:

ALMA uses Microchip PIC's for hardware control. They communicate via
the AMBSI1 or AMBSI2 to the computer. So the hardware folks would take
the PCB with the PIC, slap it into a module, and simply want the MIB
to replace and duplicate the AMBSI1 or AMBSI2 functionality.

This would keep costs down, provide hardware compatibility, etc.

New modules could be designed with System 2 in mind.  And I guess, we
could define System 1 to be a subset of a System 2.

Another reason I am open to multiple MIB types, and concerned about size.


-wayne-