[evla-sw-discuss] Re: A MIB Requirement

[Steven Durand]

Item #2 in the previous EMAIL states that the Device must power up in a
safe mode even if the MIB is missing or failed.  This forces device
designers to install a interface micro to set all registers to safe
values upon start-up. 

I understood that it might be possible for the MIB to set all the
registers to safe values upon start-up.  This would remove the
requirement for an additional micro processor. Has this changed??  




Wayne Koski wrote:
> 
> I think we should be very causious in this.  We have a situation here
> of power-on conditions which can be complex.  So here is my take on it.
> 
> 1.  The MIB and Device at power-on will have a reset applied to them.
> 
>     Note that the above resets could be produced by the same reset
>     device or different devices.  We have not imposed a coordinated
>     reset requirement.  Thus the problem is different reset periods for
>     the MIB and Device.
> 
> 2.  For the Device, its reset must place it into a SAFE mode of
>     operation.  It shouldn't harm itself or other Devices it may be
>     connected too.
> 
>     This immediately protects the hardware, and the MIB can actually
>     not be connected or the MIB can be in a failed state.  In other
>     words the Device should enter a SAFE mode of operation, with or
>     without the presence of the MIB. I think this is what Kevin was
>     trying in part to get across.
> 
> 3.  After the MIB resets, it can then load parameters, etc. in order
>     to place the Device into an INTIALIZED mode of operation.
> 
>     Obviously the INITIALIZED mode show also not cause damage to the
>     Device or any Devices connected to it.
> 
> 4.  A contentious possibility:  Once a particular module has been
>     INITIALIZED or in an OPERATIONAL mode, the MIB might signal a
>     downstream Device that it is safe to apply signals from the
>     upstream device. A method to use intermodule communications as
>     an additional layer of protection.
> 
> Example:  The 1st LO Synthesizer is set to a value and the output
>           power is determined to be in range.  The Synthesizer could
>           then signal the RF/LO/IF switch that the level could now
>           be applied to the Front-end, etc.
> 
> Just remember, I view the above as an additional layer of protection.
> Expensive items should be protected via hardware as much as possible.
> 
> -wayne-


-- 
Steven Durand
Member of Technical Staff (MTS)
National Radio Astronomy Observatory
PO Box 0 
1003 Lopezville Road
Socorro, NM 87801-0387
Office: (505)835-7103
Fax:    (505)835-7027