Current State Of Rebreather Electronics

[silent running]

Well, a dive computer that gives 25-30 hours between charges is pretty good. If the SSS drops that by 50% (much more than observed by users so far) then I would have 12-15 hours between charges. Thats enough for a couple of days of the dives I do, or i can recharge it overnight and do whatever dive I may want to. The charge percentage and remaining battery life is very well calibrated and shown, just like on SW computers etc.

Battery capacity claims are often overstated, I’ve never used a battery that actually lasted as long as claimed and low ambient temperatures adversely affect battery life. It’s one thing to use/observe and have confidence in the battery life indicator when the battery is in the top 1/3 of its capacity, but has anyone run a Poseidon battery all the way down, say on the surface, to verify how they behave at the end of their capacity? I don’t worry about battery life because I don’t need one to know my PO2, but if I was doing a big dive on a fly by wire ECCR I would need to know to high degree of probability that the battery capacity as stated was correct and that there was no chance of an accelerated discharge below 50% capacity.

Those are due to people removing the batteries while the units are still wet and allowing water to run into the connector. Same as people getting 1st stage issues from allowing salt to get into the inlet etc etc.

An O2 first stage issue is likely to manifest as a slow decrease in performance over time, a battery intermittently failing to deliver current to the electronics could result in a sudden electronics shutdown. I had a solenoid clog up with corrosion from wet O2 (Indonesia, probably a failure to vacuum the supply cylinder before refilling after leaving the valve open), it took days for the solenoid to clog.

Water can easily get on battery contacts even after a unit has dried. Rough seas/spray, wet gear hanging above the tank racks, spray from the crew washing down the deck, I’ve never seen a dry dive deck. The camera table is dry for periods of time, but I don’t think even a photographer would leave their camera gear out of their housings on a dive deck camera table for any length of time.

If the Poseidon battery contacts are on the outside of the unit and thus exposed to the elements in between charges, this is a real
vulnerability. I wonder if an induction type battery connection to the head could be made to work in a future design?

Again, I don’t worry about battery contacts because I don’t need a battery to finish a dive, and my battery harness is disposable in the event of exposure to water. I’ve flooded my battery compartment twice, it was a minor inconvenience; dispose of the battery and harness, clean contacts, replace battery and harness = $10. IMHO, the less reliance on batteries the better, they really don’t function well in wet places.

One of the big reasons is about the concept of "independent" sensors, which we do not actually have with dual or more galvanic sensors (watch this for some background on the Poseidon cell philosophy
Short version, if you have two cells with same mechanism, they cannot be truly "independent" since they are exposed to the same conditions. If you have one SSS and one galvanic as a backup, you have a genuinely independent , redundant system.

So every system needs two different measuring devices to be considered independent/redundant? How did anyone dive CCR until now? The USN must have been very reckless to trust the lives of their highly/expensively trained EOD divers to galvanic sensor voting logic for 5 decades. Requiring two different measuring devices for accuracy would also invalidate an awful lot of science over the last 500 years. The two different measuring device argument proves a bit too much and in my mind, leaves open to question the confidence Poseidon have in their sss.

Again, we all know galvanic sensors are not very reliable, but their failure modes are known and when used in multiples with simple voting logic and that onboard computer known as the human brain for backup, we have something simple that works predictably, and without the many vulnerabilities of more complicated systems...

 

[RainPilot]

but has anyone run a Poseidon battery all the way down, say on the surface, to verify how they behave at the end of their capacity?

When you place the battery in the smart charger, it will do a learning cycle if it has been more than 30 days since the last one, or if there are any discrepancies in the battery discharge log. This involves the battery being fully charged, then put under a representative load until fully discharged. The calibrated information is then stored in the computer located inside the battery case and used for the resource algorithm.

A lot of people have issue with eCCR in general, I understand the "I don't trust electronics, I will control my dive" mentality, have at it by all means. My personal philosophy is that an eCRR, coupled with close monitoring and understanding of what the system is doing and how to ID and fix errors, is safer than a system that relies only on a human monitoring and operation system. This has proved true in aviation and many other HRE operations over the years.

A false sense of security from an eCCR ie not monitoring it and assuming that the machine will always work, is a training and attitude issue, not an inherent problem with the system itself.

 

[njdiverjoe]

Well, that was a big honkin’ load of Kool-Aid.

Yes, I watched the whole video. Yes, I majored in Mathematics and Logic & Computability in college.

I well give the presenter a kudo, though. He had enough integrity to say in his Summary that 3 sensors in a voting configuration are better than a single sensor. He just said it in such a way as to absolutely minimize the real superiority of the 3 sensor voting configuration.

He said “unless very careful steps are taken, three ... are little better than a single sensor.” He totally redirects from considering how much better 3 are when you ARE very careful. At least he acknowledges that even if you are not very careful 3 are still a little better than a single sensor.

Hi Stuart, can you explain in more detail what you mean by this?

 

[stuartv]

Hi Stuart, can you explain in more detail what you mean by this?

I will. But, I have a shload of stuff going on today and leaving this afternoon to dive in NC for 3 days. So, it might be a few days.

 

[silent running]

When you place the battery in the smart charger, it will do a learning cycle if it has been more than 30 days since the last one, or if there are any discrepancies in the battery discharge log. This involves the battery being fully charged, then put under a representative load until fully discharged. The calibrated information is then stored in the computer located inside the battery case and used for the resource algorithm.

A lot of people have issue with eCCR in general, I understand the "I don't trust electronics, I will control my dive" mentality, have at it by all means. My personal philosophy is that an eCRR, coupled with close monitoring and understanding of what the system is doing and how to ID and fix errors, is safer than a system that relies only on a human monitoring and operation system. This has proved true in aviation and many other HRE operations over the years.

A false sense of security from an eCCR ie not monitoring it and assuming that the machine will always work, is a training and attitude issue, not an inherent problem with the system itself.

I’m glad to hear Poseidon take battery diagnostics seriously, it is imperative considering their entire life support system relies on them.

It’s not that I don’t trust ECCRs, I dive one. It just happens to be the simplest kind and only maintains SP. It also has sophisticated cell tracking/validation and a very sensible, simple HUD which informs me by colored LEDs whether I’m going hypoxic, hyperoxic, low battery and whether 1 or 2 sensors have been voted out. In case you missed it, it also has a passive needle gage secondary which is driven directly off the high output sensors and requires no battery or electronics to show PO2. I treat electronics as a convenience, not as a stand in for my basic competence as a CCR diver.

I’m not trying to be argumentative for the sake of it. My point is that after all the fatalities and near misses of the last 20 years, are the hazards of CCR diving going to be significantly reduced by increased diver reliance on technology, or is the better path to examine mindset, training and ways to make existing technology safer and more reliable? I think we’re a long way from 100m dives with hypoxic gases being a walk in the park, no matter how sophisticated the technology.

Deco theory and the effects of lengthy high PO2 exposure are still largely unvalidated and unknown. We are all pioneers in hyperbaric medicine and experimenting on ourselves every dive. And as such I think it’s misguided to promote a high tech CCR as a path towards “easy” or “safe” diving. The less involved you are in the workings of your CCR and the more reliant you are on it, the less aware you are of the fact that you alone are ultimately responsible for your own life. The more complicated the machine the more hardware and software there is to go wrong. Human being make mistakes, and that includes software coders, technicians and engineers, none of whom will be in the water with you when something goes wrong.

If Poseidon want to develop tech that will be truly invaluable and revolutionize CCR diving, they should partner with a medical device company to come up with real time monitoring of blood gas composition and Doppler monitoring for input into a deco computer, that would be a true game changer...

 

[tbone1004]

If Poseidon want to develop tech that will be truly invaluable and revolutionize CCR diving, they should partner with a medical device company to come up with real time monitoring of blood gas composition and Doppler monitoring for input into a deco computer, that would be a true game changer...

that implies that someone who know what to do with that data. Science still hasn't cracked that one yet.

 

[KevinNM]

They know what to do with blood gas results. In an ICU, with a team of experienced experts.

 

[silent running]

If oxygen % can be measured optically maybe other gases can be as well. Even if you could optically measure 2 of the 3 trimix gases, you can extrapolate the % of the third for a high probability blood gas composition in real time. Add real time Doppler info and maybe there’s a way to non invasively extrapolate the rate at which He and N2 are diffusing from tissue back into the blood, and thus compute a much more accurate estimation of the time and depths necessary to off-gas properly.

Deco computers do this based on models, I’m talking about giving the models actual real time information of that which they model.

Is there a non invasive/optical/acoustic way to determine blood gas composition and bubble size, and can this be done underwater? If so, shouldn’t it be then possible to get much more accurate real time deco information?

 

[KevinNM]

IIRC, their is only some reasonable correlation between bubble numbers and DCS. People with lots of bubbles usually don’t get bent and people with few bubbles sometimes get bent, but it’s more common for people who got bent had lots of bubbles.

 

[Jeremy Williams]

If so, shouldn’t it be then possible to get much more accurate real time deco information?

If you are diving with a single diluent (most people in CCR), then you can calculate the inspired gas fractions in real time from just the O2 sensors we already have. And we do. The blood gas fractions closely trail the inspired fractions to the point that they are practically the same over any reasonable timescale. Measuring the fraction of helium in the inspired mix doesn’t really add any additional data. I don’t believe that measuring anything in the blood gas mix will substantially improve the data given to the deco algorithm either.

There is no place in ZHL-16 to integrate doppler data even if it existed in real time and we knew what it meant when we got it. The best we could do is increase the gradient factors in some prearranged way.

What you are really proposing is new fundamental technology that has nothing to do with recreational rebreathers but could be used by rebreather divers if it existed. Like most fundamental advancement it will probably need to come from the military (I.e. NEDU) or commercial operations. No rebreather electronics manufacturers are going to do that kind of expensive and speculative work.