SECTION 2. INTERNAL DATA STORAGE

NOTE: All memory pointers are set to the DSP location when the datalogger compiles a program. For this reason,

ALWAYS RETRIEVE UNCOLLECTED DATA BEFORE MAKING PROGRAM CHANGES. For example, assume the TPTR lags the DSP by less than 512 data points when the datalogger program is altered. On compiling, the TPTR is positioned with the DSP, losing reference to the data that was intended to be transferred to tape. The data is not automatically transferred and appears as a discontinuity in the data file. Until the ring memory wraps around and data overwrite occurs, the data may be recovered using the *8 Mode. This scenario is also true for the SPTR and data intended for a Storage Module.

output in high resolution or could be offset by 20 ft. (transforming the range to 30 to 50 ft.).

2.2.2INPUT AND INTERMEDIATE STORAGE DATA FORMAT

While output data have the limits described above, the computations performed in the CR10 are done in floating point arithmetic. In Input and Intermediate Storage, the numbers are stored and processed in a binary format with a 23 bit binary mantissa and a 6 bit binary exponent. The largest and smallest numbers that can be stored and processed are 9 x 1018 and 1 x 10-19, respectively. The size of the number determines the resolution of the arithmetic. A rough approximation of the resolution is that it is better than 1 in the seventh digit. For example, the resolution of 97,386,924 is better than 10. The resolution of 0.0086731924 is better than 0.000000001.

2.2DATA OUTPUT FORMAT AND RANGE LIMITS

Data is stored internally in Campbell Scientific's Binary Final Storage Format (Appendix C.2). Data may be sent to Final Storage in either LOW RESOLUTION or HIGH RESOLUTION format.

2.2.1 RESOLUTION AND RANGE LIMITS

Low resolution data is a 2 byte format with 4 significant digits and a maximum magnitude of +6999. High resolution data is a 4 byte format with 5 significant digits and a maximum possible output value of +99999 (see Table 2.2-

1below).

TABLE 2.2-1. Resolution Range Limits ofCR10 Data

A precise calculation of the resolution of a number may be determined by representing the number as a mantissa between .5 and 1 multiplied by 2 raised to some integer power. The resolution is the product of that power of 2 and 2-24. For example, representing 478 as

.9336 * 29, the resolution is 29 * 2-24= 2-15= 0.0000305. A description of Campbell Scientific's floating point format may be found in the description of the J and K Telecommunications Commands in Appendix C.

2.3DISPLAYING STORED DATA ON KEYBOARD/DISPLAY - *7 MODE

(Computer/terminal users refer to Section 5 for instructions on entering the Remote Keyboard State.)

 

 

MinimumMaximum
ResolutionZeroMagnitude

Magnitude

Low

0.000

+0.001

+6999.

High

0.0000

+ .00001

+99999.

The resolution of the low resolution format is reduced to 3 significant digits when the first (left most) digit is 7 or greater. Thus, it may be necessary to use high resolution output or an offset to maintain the desired resolution of a measurement. For example, if water level is to be measured and output to the nearest 0.01 ft., the level must be less than 70 ft. for low resolution output to display the

0.01ft. increment. If the water level was expected to range from 50 to 80 feet the data could either be

Final Storage may be displayed by using the *7 Mode. Key *7.

If you have allocated memory to Final Storage Area 2, the display will show:

07:00

Select which Storage Area you wish to view:

00 or 01 = Final Storage Area 1

02 = Final Storage Area 2

2-4