Just an idea

[djulien]

I can address any of the 72 lights with a combination of turning three specific channels on or off, but an important disadvantage is that I can't turn all 72 lights on at the same time, only blocks of 24.

An interesting feature about driving incandescent lights with charlieplexed SSRs is that there is usually enough current drawn to latch the Triacs, so you actually can have more than one row turned on simultaneously. For example, if the controller turns on the first row of optos for, say, 10 usec, then you can move on to the next row and turn them on for 10 usec, next row for 10 usec, etc. (I'm using 10 usec as an example - I've read posts that the MOC3023s can be turned on faster than that, but it depends on how much current you give them). Even though there can only be one row of IR LEDs (input side of the optos) turned on at any one time by the charlieplexing circuit, the Triacs on the output side can stay turned on longer than that (if there is enough current drawn by the lights), which basicaly means that all the lights can be on at the same time. This is actually an advantage that AC SSR charlieplexing has over direct DC drive of LEDs. For example, I have 4 ArchFans controlled by charlieplexed SSRs (total of 64 channels, 56 of which come from one cat5 cable), and they are all on at the same time at some points in the sequence.

All this to say that you can actually get better performance from charlieplexed SSRs than what you would first expect (IMHO). However, it depends on your controller hardware + firmware, whether you want exact timing or can live with +/- a few dimming levels, etc.

EDIT: NOTE: my comment about all rows on applies to ch'plexing, not to the row/column multiplexing which is quoted above. Sorry for straying off topic from this post.

don

[Skunberg]

Don thank you for jumping in. I get the idea of this I just can't seem to explain it. I had planed to use it's once but the added complexity to sequence I decided not to use it.

Sent from my SCH-I500 using Tapatalk