Preliminary thoughts about a project: smart coexisting irrigation controller

[Back2Volts]

We have a 16-zone programmable irrigation controller .   It is from a reputable brand, multiple programs, remote control, etc... but it lacks flexibility for the "on demand" watering.   The user interface at the control unit is a decade old technology: selector knob, 4 push buttons and small dim custom LCD.   The remote works, but only allows to interact with a zone at a time, with to time feedback.   The bottom line is that it would be nice to have a modern flexible interface, logging/reporting, mobile access, etc...   

So looking at options:

1)  Replace controller with one of the few mobile access irrigation control units on the market.   Pros:  probably does some of what we would like, better interface.   Cons:  a) The local servicing company (winter prep, spring start...) would not be familiar with it.   b) It may not have all the whistles I could implement if I developed the software myself. c) There is no fun to it.

2) Replace controller with my own design:   Pros: many.   Cons: a) same above servicing issue.   b)  Wife is a "practicing" Master Gardener, so the controller is "mission critical".   Too risky!

3) Hybrid system, with the existing controller as "dumb master" and my own designed as "smart slave", where the only prerogative of the master is to activate/override any one of the ports at any time.   The new controller with full functionality as desired, plus awareness to relinquish control if it senses (rare) activity from the master.   Old master remains as full online backup, alleviating stress from rework lapses on new controller.

Option 3) is what I am exploring.   I think of it as:

1) Run the lines from master and slave through a bank of DPDT relays, with normally closed on the master side.   I say relays because I thing it would be problematic to have two sets of TRIACs driving the same lines. 

2) New controller includes monitoring of the status of the master's ports to provide logging and turn off any other active line if required.   I think of AC voltage sensing on all ports (OPTO ?) plus possibly a current sense transformer on the common line to valves from master side.

3) Drive both units from same 24V AC class 2 transformer, replacing existing 24VAC/1A unit with 24VAC/1.5-2A to allow for new circuit load.   Isolated DC PS for new digital side as well as isolation with the 24VAC lines.    I am a bit worried about high voltage spikes on the valve lines due to storms, but at the end there is probably not much that can be done about it other than adding some TVSs.

4) New system based on PSoC4/5 with 5" TFT/touch (keypad?).   Possibly combined with RPi for local GUI and mobile support if doing this on PSoC becomes a bit challenging).

All of this is very preliminary thinking.    I will appreciate any comments and ideas.

Tony