Detail for my sketch, Mk2_multiLoad_wired_7a.ino
 *
 * This sketch is for diverting suplus PV power using multiple hard-wired loads.  
 * An external switch allows either load 0 or Load 1 to have the highest 
 * priority.  Any number of loads can be supported by the logic, a dedicated 
 * IO pin being required for each one. 
 *
 * This sketch is intended for use with my PCB-based hardware for the Mk2 PV Router.  
 * The selector switch, as mentioned above, connects to the "mode" port which has been 
 * re-assigned for priority selection. "Normal" mode can be achieved by setting the 
 * anti-flicker offset prameter to zero at compile-time.
 *
 * The integral voltage sensor is fed from one of the secondary coils of the transformer. 
 * Current is measured via Current Transformers at the CT1 and CT1 ports.  
 * CT1 is for 'grid' current, to be measured at the grid supply point.
 * CT2 is for the 'diverted' current, so that energy which is diverted via the primary
 *  dump-load can be recorded and displayed locally.
 *
 * A persistence-based 4-digit display is supported. To free up the necessary IO pins
 * for driving multiple loads, the pin-saving hardware needs to be in place.  
 * These extra logic chips (ICs 3 and 4) reduce the number of IO pins 
 * that are needed to drive the display.  The freed-up pins are available at the 
 * J1-5 connector.  The uppermost position has been assigned to drive Load 1,
 * the next one down is for Load 2, and the lowest one is for Load 5.  The control signal
 * for Load 0 is available at the "trigger" connector.  
 *
 * With the green (rev 2.1) version of my PCB, each of the additional outputs has an 
 * associated ground pin.  It is therefore more sensible for those outputs to be active-high
 * rather than active-low.  With a 5V regulator rather than the normal 3.3V one, these 
 * outputs are able to drive an SSR directly. energymonitor.org/emon/node/1757
 *
 * September 2014: renamed as Mk2_multiLoad_CAT5_3, with these changes:
 * - reimplementation of cycleCount, as it could have overflowed with unpredictable results;
 * - the functions increaseLoadIfPossible() and decreaseLoadIfPossible() have been tidied;
 * - energyThreshold_long has been renamed as midPointOfEnergyBucket_long.
 *
 * December 2014: renamed as Mk2_multiLoad_CAT5_4, with these changes:
 * - persistence check added for zero-crossing detection (polarityConfirmed);
 * - lowestNoOfSampleSetsPerMainsCycle added, to check for any disturbances;
 * - the logic for each of the 5 additional loads has been inverted by use of the '!' character.
 *     these outputs are now active-high rather than active-low
 *
 * November 2015: renamed as Mk2_multiLoad_wired_5, with these changes:
 * - the original twin-threshold algorithm for energy state management has been reinstated;
 * - improved mechanism for controlling multiple loads (faster and more accurate);
 * - the phaseCal mechanism has been reinstated;
 * - SWEETZONE_IN_JOULES has been replaced by WORKING_RANGE_IN_JOULES.
 *
 * January 2015: renamed as Mk2_multiLoad_wired_5a, with this change:
 * - minor bug-fix in allGeneralProcessing() which affects how the energy thresholds are adjusted immediately
 *     after a change of load-state has tqaken place. 
 *
 * January 2015: renamed as Mk2_multiLoad_wired_6, with this change:
 * - reinstatement of min & max limits for the energy bucket's level.  This section was lost during the 
 *   conversion from version 4 to version 5.  The absence of this section prevents diversion from starting 
 *   in the correct manner.  Versions 5 and 5a should therefore not be used.  Version 6 is believed to be
 *   a correct implementation of the improved mechanism for controlling multiple loads.  
 *
 * January 2016: renamed as Mk2_multiLoad_wired_6a, with a minor change in the ISR to 
 *   remove a timing uncertainty.
 *
 * January 2016: updated to Mk2_multiLoad_wired_6b:
 *   The variables to store the ADC results are now declared as "volatile" to remove 
 *   any possibility of incorrect operation due to optimisation by the compiler.
 *
 * February 2016: updated to Mk2_multiLoad_wired_7, with these changes:
 * - improvements to the start-up logic.  The start of normal operation is now 
 *    synchronised with the start of a new mains cycle.
 * - reduce the amount of feedback in the Low Pass Filter for removing the DC content
 *     from the Vsample stream. This resolves an anomaly which has been present since 
 *     the start of this project.  Although the amount of feedback has previously been 
 *     excessive, this anomaly has had minimal effect on the system's overall behaviour.
 * - tidying of the "confirmPolarity" logic to make its behaviour more clear
 *   
 * February 2020: updated to Mk2_multiLoad_wired_7a with these changes:
 * - removal of some redundant code in the logic for determining the next load state.
 * 
 *
 *      Robin Emley
 *      www.Mk2PVrouter.co.uk
 */