...The timer will run +/- 30 minutes every time it runs....
If I remember correctly, with the old software "randomized" the time by adding 37 min to the previous time but always staying the same hour. The formula for the min time was to use the remainder of the (number of the day since the last download X 37 + the starting minute) divided by 60. This meant if you had light A start at 1:00 and light B start at 1:05 the "random" time for B was always 5 min after the "random" time for A. I use to have a chart of the times that my "random" lights would come on. It worked like this. Assuming you started on the first day with a security time of 1:00 for A and 1:05 for B, this is how the times would go.
| Day 1 | Day 2 | Day 3 | Day 4 | Day 5 | Day 6 | ...and so on. |
Light A | 1:00 | 1:37 | 1:14 | 1:51 | 1:28 | 1:05 |
Light B | 1:05 | 1:42 | 1:19 | 1:56 | 1:33 | 1:10 |
I used a third party software back then to program times for my CM11A. The software allowed some interesting features such as odd or even day settings, updated sunrise/sunset settings, even used some of the otherwise unused memory space of the CM11A for data storage. (I don't think it allowed the stored data to be used as interactive flags, however I remember finding some use for it.) The main catch is that you needed to download to the CM11A every week. This meant with my automatic Sun download my random times for each day was constant.
I haven't tested the CM15A to see if it's random times are predictable.
[Note: I may have this confused this with the system that preceded the CM11A as I seem to remember having a DOS system back then. How old is the CM11A and what preceded it?]
steven_r:
I think you must be referring to the old X10 CP290 interface.
The CM11A came out around 1997 and a few of the things you describe don't apply to the CM11A:
1. There's no built-in Odd/Even day feature in the CM11A. The software could specifically program odd and even days but this will chew up a huge amount of EEPROM memory. (It does have built-in support for specific days of the week.)
2. Although one can write data into the CM11A's unused EEPROM memory space, there's no way to read it back out.
3. The ActiveHome software creates a stair-step approximation to the daily Sunrise/Sunset times over the year and uses this for timers downloaded to the CM11A EEPROM. So the downloaded schedule can nominally run for a whole year before it needs to be redone. The CP290 used a constant value so the schedule needed to be re-downloaded frequently, especially in the Spring and Fall when Sunset and Sunset changed more rapidly.
What I'm seeing with the security feature in the CM11A is that the "random" time just increments by 10 minutes every day after a starting increment the first day. When the total of the increments exceeds one hour, it wraps around. (The ActiveHome software subtracts 30 minutes from the programmed time so the variation is +/- 30 minutes.)