Hi Jason,
Andy described the PRISM, but let me get a bit more detailed in regards to other CCRs.
The Classic Inspo has two controllers for the setpoint that display the pO2. Each is powered by its own battery, both of which are housed in one compartment. The two controllers are connected to one another, thus not independend. The first one to power up becomes the "master", the other one the "slave". The latter is supposed to take over setpoint control and monitoring when the former fails.
The Vision equiped AP units have the same "master/slave" layout, as well as two batteries in one compartment. The batteries however are shared, either can be used by either controller, the "master" powering the solenoid gets power from the stronger battery until finally power from both is used to supply setpoint control. That can be either a step forward or back, depending on how you look at it. On one side the setpoint control will get available as long as possible, which is good, on the other hand the batteries are now also wired together, which may not be. Failure such as a short in one controller and/or battery could influence the other.
The part I don't like is that you have to trust one controller when the other screws up. Both versions had software problems initially, and with the fairly new Vision some are still emerging. While that is to be expected, it sucks if its all you got. You got problems in one controller, the other one will be plagued too. No independent redundancy.
I'm sure when AP decided to use multiple controllers they were influenced by the design of the Cis-Lunar MK-5p. That unit had an enormous amount of redundancy, three setpoint controllers, 4 or 5 batteries, 4 displays, all spread to different location in and on the rebreathers. Blackbox with dual controllers, each with their own battery in its own compartment, the main one with a backup battery. DOS-HUD display on the BOV (signaling status of Deco, Operation and Setpoint) and a buddy display (setpoint) on the back. The big brick-like display had yet another setpoint controller and with battery in its own compartment, and the main display for sepoint, operation, onboard (and optional 2 tank offboard) tank pressure and deco computing. On the backside was the fourth display, LEDs that read the pO2 of the three cells, also powered by a separate battery in a separate compartment. So even with all the setpoint controllers and displays Dr. Stone believed that a simple, low power, independent pO2 monitor was a necessity to safely dive a rebreather. I firmly believe that decision was correct. The drawback of all the redundancy of course are complexity and expense.
If you look at the other big kahuna of rebreathers, also available at the time, the MK15.5, the unit is the perfect exercise in simplicity. A single setpoint controller with a battery in a separate compartment. A wrist (or HUD) LED display that conveyes the setpoint. An analog display that reads the three cells (and the battery when the electronics are on). A power switch to physically disconnect and cut the power. If there is a short somewhere, no battery power can be fed back to the sensors or analog display and fudge the readings. Since the the MK uses high output cells that are measured directly by the analog gauge (mV meter with a pO2 scale) there are no batteries to fail in that part of the unit. The horseshoe board inside the center section separating the signals can be a pita to work on, but that aside the unit is as simple as it gets. And for many people, there is safety in simplicity. Passive safety as there are less parts to break or fail, active safety as there are less mistakes to be made when maintaining, assembling and diving the unit.
Due to his long experience with the MK series, from the Carmellan rebreathers used by Rob Palmer to explore the Bahamas to the SM1600 he build under agreement with Carleton, he decided on that route for the PRISM. Same kind of electronic simplicity. Bit more advanced of course, the original MKs were analog, his are digital. The HUD conveys battery status, single or multiple cell failures on top of the setpoint. The analog gauge has improved, too, with a dual jeweled movement over the origianl single one, and also conveys setpoint in addition to pO2 readings and battery status. To this point the SMI electronics are still the only ones tracking cells, thus being able to distinguish between one or two cells going bad and voting out the correct one. (The Vision was supposed to have that, too, and initially did, but the feature was thrown out in one of the early revisions). As for power management, the PRISM's battery operates the setpoint computer, HUD and solenoid. Should the voltage drop below 7.6V, the solenoid gets voted out. The diver retains the HUD display (that signals the situation) and controls the setpoint manually. Even with a complete electronic shut down, the analog gauge is still usable. Again, simple.
And I firmly believe that when things go wrong, simple becomes important. Neither PRISM nor MK have menus to scroll through and choose from. No multiple buttons and combinations of buttons to push to make some selection or other. A simple flick on the power switch and the electronics are off. A gauge that tells you where you're at. Simple, easy, quick.
The one major drawback of analog gauges are that they damage easily. Drop it and you're most likely in line for a new one. Drop it on your week long liveaboard trip and you better have a spare or you'll be spending the rest of the trip at the bar mourning your display and dives. There is one alternative from the aftermarket, a digital, battery powered secondary. Based on the one used in the Colkan 155 (MK15.5) it displays all three cells, both batteries and setpoint. Made by Subsea Systems in Australia, it's a replacement plug-and-play for the analog gauge, and a perfect backup as far as I'm concerned.
Back to the electronics, the APECS Meg has dual custom battery packs inside the head, each in its own box. One supplies the setpoint computer, primary wrist display and solenoid. The second one supplies the independent pO2 monitor and the setpoint repeating HUD.
On the Hammerhead (used in the Optima) you have the batteries in the handsets. The primary supplying power to the setpoint controller, deco computer and solenoid and the display. The secondary supplying the independent pO2 monitor, dive gauge, display as well as the HUD.
Last but not least the Ouroboros. It uses three batteries in the main electronics pod. Two are the main batteries, one is a backup (providing about half the operating time). These power the setpoint controller, solenoid, deco computer, HUD (which includes setpoint, deco, solenoid and general warnings) and buddy display (includes pO2 readout and 4 LEDs mirroring the HUD). The digital secondary pO2 display has its own independent battery in the handset.
IMHO each and everyone a better solution than AP's dual controller without independend redundancy.
Finally, the AP units have the batteries (now sealed against water) at ambient pressure, all other units above have them sealed at 1ata (well, surface pressure, anyway). AP and Meg have the batteries inside the loop, the rest outside. Cis, MK, PRISM and Ouroboros also have the solenoid outside the loop.
