I feel Fred's pain.
I have the same setup. The only semblance of success I'm seeing in having the system work closed loop based on ORP, is to set the intellichem to dose ORP by time, not to setpoint.
For my pool, if the intellichlor is left on 100% of the time, it will drive ORP down so far (about -150mv delta) it will stay on basically forever, burn through acid at about a gallon/24 hours, and never reach ORP setpoint. Its does eventually bottom at some number in the low 500s but never rises and so the FC goes off the chart, rising about 7ppm in 24 hours. After manually turning off the intellichlor by lowering the ORP setpoint, ORP will rise almost immediately and continue to rise over the next couple of days into the high 700s or low 800s usually. Then it will slowly drop proportionally to the FC all while cycling up and down due to the sun. I wish I had a schematic for the probe circuitry in the intellichem. I'd love to know what kind of isolation (if its isolated), amplification, and filtering are on the probe input. Whats the response time constant?
Some say this is due to hydrogen gas production and then off-gas, which I'm sure is at least partly to mostly right, based on months of experiments and reading papers, even with a gold tip probe which Sensorex and others recommend for salt water. (Is there a way to "getter" or remove the hydrogen gas in the flow cell?) I think for my pool's installation (which is not exactly per the manual for intellichem) and based on a few articles and posts I've read, a component of the negative offset might also be electrical current in the pool water but I haven't tested/proven that yet. I've looked for errors in grounding and bonding creating ground loops but nothing obvious has popped out at me. You can see noise (about 1mv) in the ORP when the cell is energized though.
So far by experimentation, I've determined that by setting the intellichem to dose intellichlor for 45 minutes on and 15 minutes off, I can keep the system from entering a permanent on runaway condition. The ORP still drops, BUT it rises over time so it eventually will reach setpoint and turn off. This is highly dependent on the on/off ratio and I don't know how stable it will be as we heat up here into summer. I may have to trim it.
I wish the intellichem would allow the cell to be turned on to a percentage value. It seems it only bangs the cell on or off at 100%. I'm pretty sure if I could tone it down enough to stay below the critical "hydrogen limit" the system would be more workable.
Confounding me a bit is my CYA is 50 and so I get a big swing in ORP just from the sun during the lengthening clear days in Arizona. I'd like to see what happens when CYA drops to about 30. I'm hoping that will happen during the summer under the intense sun. I so far am unwilling to drain the pool to test this.
If you want to experiment, I highly recommend you keep intellichem in charge of the pH. It works fantastic for that and a stable pH removes a HUGE variable in ORP level. If my pH also moved around a lot it would be impossible to develop any sort of work around.
I'm an engineer and this bothers me no end. Especially since there are anecdotes and videos from professional pool people out of Hawaii using intellichem all over with salt cells. I can't see these folks running around all the time manually adjusting ORP set points all the time. Is it something about the water in the southwest? Mineral content? I know folks will recommend we just abandon ORP control, but I'm not there yet.