Hardware Manual for the PCD3 Series│Document 26/789; Version E 5│31.01.2005
Saia-Burgess
Controls Ltd.
Partitioning options for user memory
PCD3.Mxxx0 Classic CPUs and expansion housings
3-18
3
3.11
Partitioning options for user memory
In the PG5 hardware configuration, the user memory is partitioned by default into
lines of code and texts/DBs, in a way that suits most applications.
In the case of a large program with few texts/DBs or a very small program with
many texts/DBs, the user can partition the memory manually. In order to choose an
appropriate breakdown, the following should be noted:
the partitioning is into “kBytes lines of code” and “kBytes text/DBs”, where the
“kBytes lines of code” can only be changed in 4 kByte steps, as every line of code
occupies 4 bytes
the result of the formula (4 x “kBytes lines of code”) + “kBytes text/DBs” must
equal the available user memory, e.g. 4 x 24 32 kBytes = 128 kBytes
each character of a text occupies 1 byte
each 32-bit element of a DB occupies eight bytes in the address range 0..3999,
and the header of the DB takes up a further three bytes
We recommend that DBs with addresses ≥ 4000 should always be used. These
can hold more elements (16384 instead of 384), take up less space (only 4 bytes
instead of 8 bytes per element, but NB, 8 bytes instead of 3 for the header) and
the access time is substantially shorter.
Example of manual partitioning:
