First of all, I would like to clarify that I'm only indicating a short wait to see if there is a continuing problem or if the error was triggered by some minor . I never meant to imply that the dive should continue normally. Moreover, I said it is perfectly acceptable to bailout and end the dive. If you read the first few posts this is indeed what the student was told. The instructor even asked the student if he wanted to call the dive when the error occurred in the water ("(Presumably) thinking that this may have rattled me, he asked if I wanted to ascend and end the dive" post 2).
I would agree that a 2-cell voting logic system would be a horrible idea. It's important to keep in mind that the discovery's electronics do not use simple voting logic and that the electronics test the cells "throughout the dive" to monitor cell reliability (see manual excerpt below). That is to say the cells can be assumed sufficiently reliable unless there is an active error. This is very different from typical rebreather were the only cell tests may be calibration before entering the water and a linearity test at 20 feet with no testing over the next several hours or more.
I find it unlikely that the PPO2 could change enough in a minute of to go from normal operation to a hypoxic or hyperoxic condition. If there was an O2 valve stuck open, possibly, but that should be recognized by a competent diver as a different issue (see " capable of understanding the conditions" above) As the C1 error is not a total loss of confidence it is also possible that the unit would remain fully operation during a short waiting period. Keep in mind that both cells passed confidence testing during earlier built-in tests during the dive. The manual states that the reliability of the cells is monitored "throughout the dive."
While I'm sure you check your cells for high PPO2 readings regularly, I'm pretty sure you don't do so "throughout the dive", nor would you necessarily check the response time of your cells as this is difficult to gauge unless it is much longer than acceptable. Do you know, to the second, how long it takes from solenoid or manual injection for the gas to mix and get to your sensors?
When comparing a loss of confidence to voting logic what I was trying to imply is that voting logic should keep you safe if one cell is faulty and two cells are reliable. Even if your monitor doesn't actively warn you that one cell reading was dropped, you should still know that one cell is dropped from looking at the tree readings on your monitor.
I don't personally know what conditions would lead to a C1 error, but I suspect the instructor above either knows or has a good idea about one way the error can occur. Since the C1 is not a total loss of confidence, it may be reasonable to that one cell is still at least partly reliable. It could be that a C1 error is triggered by situations that other RB divers/systems would not even consider during a dive, such as the cell response time. Thus it is possible that the unit may be able to function normally during a short wait. I would never recommend continuing a dive normally on only one reliable cell. If the unit re-tests the cells and the condition clears, then both cells are now considered reliable by Poseidon standards, thus continuing the dive after an error clears seems like a acceptable choice. I know of no other rebreather where the cells are compared to manufacturer specs so often including during the dive.
If the error re-occurs during the dive, then calling the dive seems like the best option. The cell(s) may be close to permanent failure then.
I certainly agree that having 2 cells out of 4 giving faulty readings is time to call a dive on most rebreathers because the reliability of the remaining sensors is an unknown. I think a better comparison for most RBs would be losing one cell out of three. If one of three cells were different enough to be voted out (electronics looses confidence in that cell) would you bailout? What if, after 20 seconds, all three cells were reading the same reasonable PPO2 and none were voted out(electronics regains confidence in all cells)? Keep in mind the fact that one cell lost and one cell remaining in the Discovery means that one cell passed the reliability tests, whereas the two (of three) cells remaining in any other unit are not tested, so there is less reason to trust them ( unless you actively test them).
From the Discovery user Manual
I would agree that a 2-cell voting logic system would be a horrible idea. It's important to keep in mind that the discovery's electronics do not use simple voting logic and that the electronics test the cells "throughout the dive" to monitor cell reliability (see manual excerpt below). That is to say the cells can be assumed sufficiently reliable unless there is an active error. This is very different from typical rebreather were the only cell tests may be calibration before entering the water and a linearity test at 20 feet with no testing over the next several hours or more.
I find it unlikely that the PPO2 could change enough in a minute of to go from normal operation to a hypoxic or hyperoxic condition. If there was an O2 valve stuck open, possibly, but that should be recognized by a competent diver as a different issue (see " capable of understanding the conditions" above) As the C1 error is not a total loss of confidence it is also possible that the unit would remain fully operation during a short waiting period. Keep in mind that both cells passed confidence testing during earlier built-in tests during the dive. The manual states that the reliability of the cells is monitored "throughout the dive."
While I'm sure you check your cells for high PPO2 readings regularly, I'm pretty sure you don't do so "throughout the dive", nor would you necessarily check the response time of your cells as this is difficult to gauge unless it is much longer than acceptable. Do you know, to the second, how long it takes from solenoid or manual injection for the gas to mix and get to your sensors?
When comparing a loss of confidence to voting logic what I was trying to imply is that voting logic should keep you safe if one cell is faulty and two cells are reliable. Even if your monitor doesn't actively warn you that one cell reading was dropped, you should still know that one cell is dropped from looking at the tree readings on your monitor.
I don't personally know what conditions would lead to a C1 error, but I suspect the instructor above either knows or has a good idea about one way the error can occur. Since the C1 is not a total loss of confidence, it may be reasonable to that one cell is still at least partly reliable. It could be that a C1 error is triggered by situations that other RB divers/systems would not even consider during a dive, such as the cell response time. Thus it is possible that the unit may be able to function normally during a short wait. I would never recommend continuing a dive normally on only one reliable cell. If the unit re-tests the cells and the condition clears, then both cells are now considered reliable by Poseidon standards, thus continuing the dive after an error clears seems like a acceptable choice. I know of no other rebreather where the cells are compared to manufacturer specs so often including during the dive.
If the error re-occurs during the dive, then calling the dive seems like the best option. The cell(s) may be close to permanent failure then.
I certainly agree that having 2 cells out of 4 giving faulty readings is time to call a dive on most rebreathers because the reliability of the remaining sensors is an unknown. I think a better comparison for most RBs would be losing one cell out of three. If one of three cells were different enough to be voted out (electronics looses confidence in that cell) would you bailout? What if, after 20 seconds, all three cells were reading the same reasonable PPO2 and none were voted out(electronics regains confidence in all cells)? Keep in mind the fact that one cell lost and one cell remaining in the Discovery means that one cell passed the reliability tests, whereas the two (of three) cells remaining in any other unit are not tested, so there is less reason to trust them ( unless you actively test them).
From the Discovery user Manual
Oxygen Sensor Confidence
One of the most sophisticated features of the MkVI Discovery is the automatic oxygen sensor vali- dation system, which monitors the reliability of the oxygen sensors throughout the dive. Through a series of algorithms, the system assigns a confidence rating to current oxygen sensor readings, based on several factors including primary sensor validation, dynamic response of sensors, and a comparison between primary and secondary sensor values. If, for some reason, the system loses confidence in the oxygen sensors, then every
few seconds an error will be displayed momen- tarily on the Primary Display where the PO2 value is normally displayed – in a manner similar to how the PO2 Setpoint is displayed. If there is no confidence in the oxygen sensors, then “C0” is displayed. Other levels of confidence based on various factors include “C1”, “C2”, and “C3”. The last of these (“C3” is normal, and means the system has high confidence in the sensors. The other levels (“C0”, “C1” & “C2” generate errors, and will trigger appropriate alarms.