Is the Mk VI / SE7EN really that dangerous

[Superlyte27]

It has 2, 3 is only useful if voting logic is implemented. There are some fundamental reasons why 2 can be better than 3, especially in a calibrated and monitored scenario.

No way, lol

 

[JohnnyC]

Think about inline valves and think about using camstraps as mounting equipment. JJ's, AP's and I'm guessing Poseidon's love inline valves, our SF2's don't.

I was under the impression that Poseidon required a factory configuration during training, in which case I've never seen anyone use inline valves in a valves-up configuration before. I guess it's possible, but I don't know why you would.....

 

[RainPilot]

No way, lol

I will look up the article, it came up during my test pilot course. It’s the reason why many aircraft have two systems for redundancy not 3.

In the Poseidon, think of the computer as the third cell. It runs a model of a theoretical cell in the current conditions and compares the readings it’s getting to that ideal cell

 

[JohnnyC]

Are you still required to buy a separate set of lungs to get manual adds on the Seven? I know that was a thing when it first came out that made me go cross-eyed for a minute. Although I'm not sure if they ever actually sold any of them that way....

Or was that the VI? I don't remember.

 

[Superlyte27]

I will look up the article, it came up during my test pilot course. It’s the reason why many aircraft have two systems for redundancy not 3.

In the Poseidon, think of the computer as the third cell. It runs a model of a theoretical cell in the current conditions and compares the readings it’s getting to that ideal cell

I don’t know that I trust any computer that much.
Maybe my thinking is flawed, but take most rebreathers...
If I have 4 cells and all 4 cells and two computers all agree, the math is pretty significant that everything is okay. If I have 3 cells that agree, and two computers agree that the 3 cells are right and one cell is off, more than likely, that cell is dead, or just wet, and will likely come back, but I can continue my dive. In the unlikely scenario (but it has happened) that two cells read one thing, and the other two read something else, based on depth and gas, I can determine which cells are right. I’ll end my dive, but will likely stay on the loop, following the two proper cells. Obviously, this could be an issue with only 3 cells. What if the two cells reading together are actually the bad cells. This has also happened to me. The one cell reading off was the good cell. I again was using the depth and DIL to determine which cell was accurate, but I can see how complacency could have caused bad problems.

So, moving to the Seven... you’re trusting the computer to verify two cells. What happens if the computer screws up, and you don’t know it. What if two cells are reading exactly wrong, on both cells and the computer misses it. We have dive computers all the time with issues. I either have to verify mentally the two cells with depth and DIL or blindly trust in only two cells and a computer that could be flawed. It is after all, man made.

I think I’ll stick with 4 cells and have that much more data to pull from rather than to blindly trust that a computer is properly computing.

 

[jfe]

..........So, moving to the Seven... you’re trusting the computer to verify two cells. What happens if the computer screws up, and you don’t know it. What if two cells are reading exactly wrong, on both cells and the computer misses it........

The reason for asking my initial question, are there recorded evidence of a VI or & having both cells fail, or are we talking about possibilities. There are many scenarios that can be made but I'm after sound proof that these possibilities are real life threads. Will two cell fail at ones or do they fail intermediatly?
Speaking in rec terms I believe the Poseidon's doesn't allow for manual adding dil and O2 hence the early warning to bail out rather than trying to correct at depth. Moving onto tec terms it changes with the addition of the MAV's, but I stand to be corrected.
Currently I'm only interested in rec with limited deco should it become a factor, MOD1, but please continue the discussion it's very interesting.

 

[Superlyte27]

Yes, two cells will fail at once, absolutely, without a doubt. It’s not a matter of IF, but WHEN.

 

[Superlyte27]

@jfe if you’re thinking about a first time purchase, there are 5+ rebreathers that would be a better first purchase. And all 5 wouldn’t require you to purchase another expensive part to keep you from outgrowing it.

Off the top of my head, and purposely in no particular order...

JJ
Optima
Meg
Liberty
SF2
Kiss

 

[RainPilot]

Are you still required to buy a separate set of lungs to get manual adds on the Seven? I know that was a thing when it first came out that made me go cross-eyed for a minute. Although I'm not sure if they ever actually sold any of them that way....

Or was that the VI? I don't remember.

Two different size lungs. The larger ones came with MI ports and you could buy the MAV kit. On the smaller ones the MAV kit goes on the t-pieces with an adapter.

 

[taimen]

The issue with two cells is that the only calibration and monitoring is against a single other point of failure.

AFAIK, in Poseidons system the two cells are not calibrated or monitored against each other. There is also no voting logic between them.
The primary cell is actively validated by readings it gives when O2 or dil with known composition is injected on it and referenced to ambient pressure of course. If it gives correct readings, it is assumed to be working and used. When it fails, the other cell is validated and used. As a measurement system, this is better validated and monitored than a 3 cell voting logic, possibly also more redundant.
When both fail, both fail. But again, it should't happen unnoticed. It also has an optional 3rd cell for even more redundancy.
Bashing Poseidons oxygen measurement/validation system simply by counting cells and saying more is better is unfair.
@RainPilot can correct me if I am totally wrong.

I dive a JJ, but I wouldn't have any problem with Poseidon style cell validation system either.