Just looking at my power supply PCB, there really is not much filtering. I would have expected some large filter caps after the AC to DC conversion. It would be interesting to measure the ripple and see what it is. Just for grins, I'd temporarily add a large electrolytic cap across the DC supply to the SWG and see if that improves anything. My thought is that the IC60 pulls more current, causing the ripple to increase and maybe introducing enough noise on the comm line to cause intermittent communication failures.
Just a WAG!!!
Jim R.
Ha! My previous post didn't take because I used the alternative to this next statement, so here it is again.
Pretty DARN good idea!
I actually ordered a 10000mfd cap a while back when I thought the issue could be power supply sag. Once that was ruled out, I sort of dropped that. But PS junk coupling into the signal? Capital idea!
Also, from e-mailing with the Pentair rep, there -may- and I say -may- be a correlation of resetting with the end of a chlorine dosing cycle.
From what I have learned:
Cycle is a 5 min cycle. 4:45 [285 seconds] of that is the time when the plates are actually energized, creating chlorine, when set to 100%. This "sort of" correlates to my experiments in that short (10%) doses seem to reset in about 30 seconds [or is it 28.5?]. And longer doses [16%], if memory serves me, seem to take longer to reset, say 40 seconds [or is it 44.45?].
So it may be that once you get past the 50 something mark, by the time the failure mode decides to reset, it is back in dosing mode. And if this means it is pulling more current, that may help smooth out a noisy power line.
Well, now I'm going to go back to my cap order and see where that thing is.