## Issue

What is the difference between an Internal Analog Input (AI) and Pulse Input (PI) when it comes to fractional digits?

## Environment

I/NET 7700

## Cause

It is not always clear the difference between Internal AIs and PIs when it comes to fractional digits. A clarification note will be added to the next release of the I/NET 7700 Operator Manual, but until then, please adhere to the following when you want to display fractional digits for either Internal point types.

## Resolution

**Internal Analog Inputs (AIs)**

Internal AIs are limited to a maximum value of 2^{16} (65535). Internal AIs fractional digits are a direct result of the M value in the conversion coefficient table. If you want to display a value in hundredths (x.xx) the M must be .01 and the B=0 and the highest number to be displayed is 655.35. If you want a number to display a value in tenths (x.x) the M must be .1 and the B=0 and the highest number possible is 6553.5.

**Internal Pulse Inputs (PIs)**

Internal accumulators can accumulate not only pulses but analog values as well, but when it comes to fractional digits Internal PIs are completely different from Internal AIs. PIs are a floating point type number. This means there is no limit to the magnitude or the number of fractional digits, except that only 6 significant digits may be displayed at any one time. Therefore, the fractional digits are not ruled by the number of decimal places of the M value in the conversion coefficient table. In fact, for a Internal PI to work properly, you must utilize a conversion coefficient pair of M=1 and B=0.

An Internal PI with a calculation of P0+P1, where P0 and P1 are External AIs with values of 34.1234 and 72.5678, would be able to display the 4 decimal places with no trouble. In fact, if you used a conversion coefficient of M=.0001 and B=0 for the Internal PI, and the AIs never changed their value, the Internal PI calculation would start rounding off the value every calculation. This rounding effect would in fact be shown by the value of the Internal PI getting smaller and smaller with every scan of the Internal PI.

Remember, for a Internal PI to work properly, you must utilize a conversion coefficient pair of M=1 and B=0.