System information

17-Nov-2014 G2T Programmers Guide Rev-A
49
Note 6: When a CPU receives a UDP packet, either addressed to it or as a
broadcast on port 5417 of at least one character, the CPU responds with a UDP
packet with the format "PROJECT_ xxx" where PROJECT_ is the eight character
product code and xxx is the Network ID, with any leading zeros (Example: 11888101
007). The application for this is when DHCP is used. By sending out a broadcast
packet addressed to port 5417, one can discover which IP addresses were
assigned to what products by their DHCP server(s). If there are two systems with the
same product code, the keypad can be used to assign different Network IDs.
Note 7: Refer to the MAINFRAME section of the system manual to see additional
information on power supply monitoring.
Note 8: The following is effective on all revisions greater than 0502221100 regardless
of how many processors are installed.
• GET?/SET register 80 is defined as the "Dual Fault Condition Mask." It is
readable and writeable. The factory default is 0 but a different factory
default can be specified for each project.
• When the register is set to 0, it is "backwards compatible" and no
difference is apparent from previous versions.
• The register is a 4 bit binary mask to enable new functionality. When a bit is
set, its function ls enabled and when it's cleared, that function is disabled.
The function of each bit is:
♦ Bit 0 (weight 1): If the CPU resets and does not detect another CPU, it
logs fault 7900 if it's in the top slot and fault 8000 if it's in the bottom
slot.
♦ Bit 1 (weight 2): Assuming that two processors had been working, if a
CPU fails to get responses from the other CPU, it logs 8100 if it's in the
top slot and fault 8200 if it's in the bottom slot.
♦ Bit 2 (weight 4): If a CPU that has been working with a second CPU
detects that the other processor has reset for any reason including a
power-on reset, it logs fault 8300 if it's in the top slot and fault 8400 if
it's in the bottom slot. Note that in this case, it's to be assumed that
regardless of what state it was in, the processor that logged the fault
is now the "primary."
♦ Bit 4 (weight 8): If the top CPU resets and assumes "secondary" status,
it logs fault 8500.
Example: assuming GET? 80 responds with 15 (the sum of 1 + 2 + 4 + 8) and two
CPUs are installed and working (which you can verify with GET? 14 returning 2 from
one processor and 3 from the other):
• remove the top processor: fault 8200 is logged.
• reinsert the top processor: faults 8400 and 8500 is logged.
These features allow the user to tailor which incidents should log faults. Examples
(with the corresponding value to set in the register):
• A user with only one CPU probably does not want to be informed at every
reset that the "other" CPU is missing ( 0 ).
• A user might want to know if one of the two CPUs has a "hard" failure but
doesn't care if one temporarily "wigs out" ( 3 ).
A user might want to know if anything suspicious happened and may decide to
discuss it with the factory ( 15 ).
Note 9: Modules must be installed contiguously. Changes take effect after the next
self-test.