TM 5-3895-374-24-1 Average Shift Value 3 (Parameter 43) - factory set at 1000 counts. This parameter should be set for approximately 2 times the value set for Average Shift 2. Average Register Size (Parameter 44) - factory set at 1 (10) In the previous examples the averaging register has been shown as containing 10 readings.  The size can be changed for certain applications by changing this parameter. Parameter 44: 1 = 10 shifts 2 = 20 shifts 5 = 50 shifts 10 = 100 shifts Changing the size of the averaging register will affect the speed of the averaging system when a single register is being loaded -i.e..  when the change in raw counts is  between Average  Shift Value 1 and Average Shift Value 2.  For example.  if the scale is subject to repeated small vibrations, their effect can be “filtered” out by increasing the size of the register and perhaps slightly increasing the value for Average Shift 1.  This should result in a more stable weight display. If the register size is met too large there may however be an unacceptably long delay in the indicator's response to a small change in weight.  Use trial and error to find the optimum settings. Hold off Delay 1 and 2 (Parameters 45, 46) The Hold Off Delay feature 1s designed to delay the start of the averaging system so that the indicator display will read the target weight sooner. Without any Hold Off  Delay, the averaging system (loading one register per cycle) is engaged when the rate of change of the internal raw counts falls below  500 (Average Shift 2).  This would then require 10 or more cycles for the average to finally represent the correct weight (as in the example illustrated for Average Shift  2).  Thus it would require 2 or more seconds for the display to read correctly after the weight had been applied. The Hold Off delays the averaging system for a specified number of cycles. speeding up the response of the scale display. Hold off Delay 1 holds the indicator in full update (load all registers) for a specified number of cycles. Hold off delay 2 holds the indicator in 1/2 update (load 1/2 of all registers) for a specified number of cycles. The following example will illustrate the effect of the Hold Off delay: Note : : In the following example: : Hold Off delays 1 and 2 set to 3 cycles. : Average Shift Values 1, 2 and 3 set to 3, 500 and 1000 counts. : Indicator initially stable at 10,000 counts, then a weight representing 5100 counts is applied. Internal Update without Update with Cycle#      Counts     Change         Hold Off         Avg.      Hold off Delay        Avg. 1 10000 0 None 10000 None 10000 2 15000 5,000 Load All 15000 Load All 15000 3 15050 50 Load 1 15005 Load All 15050?Hold 4 15100 50 Load 1 15015 Load All 15100 off 5 15101 1 Load 1 15025 Load All 15101?Delay 1 6 15100 1 Load 1 15035 Load 1/2 15101?Hold 7 15102 2 Load 1 15045 Load 1/2 15101 off 8 15103 1 Load 1 15055 Load 1/2 15102?Delay 2 9 15101 2 Load 1 15065 None 15102 10 15102 1 Load 1 15075 None 15102 11 15103 1 Load 1 15085 None 15102 In the above example' without a hold off delay,  the average reading has still not reached the correct value after 10 cycles (2 seconds).  with a hold delay the average reading approximates the correct value within 5 cycles.  In summary,  the greater the number of hold off delay cycles.  The faster the indicator display reaches the target weight. (page 3-77)