
Section 2. Data Storage and Retrieval
The CR800 can store individual measurements or it may use its extensive processing
capabilities to calculate averages, maxima, minima, histograms, FFTs, etc., on periodic or
conditional intervals. Data are stored in tables. The number of tables and the values to
output in each table are selected when running Short Cut (Overview) or when writing a
datalogger program directly (Sections 4 – 9).
2.1 Data Storage in CR800
2.1.1 Internal SRAM
Internal SRAM is used as the storage area for a data table. The maximum
number of data tables that can be created is 30. Internal SRAM is battery
backed. Data remain in memory when the CR800 is powered down. Data in
SRAM are erased when a different program is loaded and run.
There are 4 Mbytes of SRAM. Some of this is used by the operating system
and for program storage. The rest is available for data storage. When a new
program is compiled, the CR800 checks that there is adequate space in SRAM
for the data tables; a program that requests more space than is available will
not run.
In September 2007, Campbell Scientific began increasing the
SRAM size from 2 MB to 4 MB. Dataloggers with a serial
number greater than or equal to 3605 will have a 4 MB SRAM.
The 4 MB dataloggers will also have a sticker stating “4M
Memory”.
NOTE
2.2 Internal Data Format
TABLE 2.2-1 CR800 DATA TYPES
Data Type Size Range Resolution
LONG 4 bytes -2,147,483,648 to +2,147,483,647 1 bit (1)
IEEE4 4 bytes 1.8 E -38 to 1.7 E 38 24 bits (about 7 digits)
FP2 2 bytes -7999 to +7999 13 bits (about 4 digits)
Boolean 4 bytes True/False —
String 1 byte per character
+ 1 byte, 16 byte minimum
— —
Data are stored internally in a binary format. Variables and calculations are
performed internally in IEEE 4 byte floating point with some operations
calculated in double precision. There are six data types used to store data. The
data format is selected in the instruction that outputs the data. Within the
CR800, time is stored as integer seconds and nanoseconds into the second
since midnight, the start of 1990. While IEEE 4 byte floating point is used for
variables and internal calculations, FP2 is adequate for most stored data.
Campbell Scientific 2 byte floating point provides 3 or 4 significant digits of
2-1
Kommentare zu diesen Handbüchern