Split sensors or not (rEvo, 5 cells)?

[lermontov]

you cant have 2 controllers regardless of how you have the cells hooked up - i cant see any issues with the nerd being on a split cell and used as a monitor

 

[Wibble]

you cant have 2 controllers regardless of how you have the cells hooked up - i cant see any issues with the nerd being on a split cell and used as a monitor

Interesting point. There’s one controller and one backup passive monitor. As above, a splitter connects two independent systems together, thus risking the failure of both under certain limited circumstances.

In the standard configuration, should the controller fail, e.g. dead battery, you’ve still got the two cells on the backup to manually fly the unit home.

If the controller failure happened at the end of the bottom phase you could have a long decompression obligation to run on two cells. If a splitter was installed then you’d have three cells, but another possible failure point.

How many simultaneous failures to plan for?

 

[broncobowsher]

Isolation is in case one system has a catastrophic failure and back feeds the signal wire. Might be a high voltage spike (and high voltage is a relative term since the system normally reads fractions of a millivolt normally) or some random voltage oscillation. The couple of resistors in the splitter are to guard against one of the two systems shorting the output to ground, but that isn't the only possible failure.

Why do you need the same information twice? I would rather use that extra slot in the displays for other useful information.

 

[lermontov]

Interesting point. There’s one controller and one backup passive monitor. As above, a splitter connects two independent systems together, thus risking the failure of both under certain limited circumstances.

In the standard configuration, should the controller fail, e.g. dead battery, you’ve still got the two cells on the backup to manually fly the unit home.

If the controller failure happened at the end of the bottom phase you could have a long decompression obligation to run on two cells. If a splitter was installed then you’d have three cells, but another possible failure point.

How many simultaneous failures to plan for?

i cant see how you can lose both read outs just because theyre sharing one cell- if that happened it would be caused by something else and the odds of that happening are remote

why would you have a long decompression just because you only have 2 cells? the deco would be the same regardless

well you have 2 readouts and BO and if you have to you can also use your buddies BO and his spare DC- that seems plenty too me

 

[beldridg]

why would you have a long decompression just because you only have 2 cells? the deco would be the same regardless

I think what he said was that you could have a long decompression schedule and be reliant upon a computer with just two cells.

Let's assume, in this one case, that your Petrel controller fails (for whatever reason) at the end of your scheduled bottom time and you have a two hour decompression obligation.

You are now relying on two cells on the NERD2 for your decompression. I would absolutely not want to bailout in that case. I would if I had to, but that would probably be my last resort.

In this specific case, would you want to have a third cell that was split between the NERD and the Petrel?

Note that in this failure scenario, it assumes the Petrel itself has failed (e.g., battery dies, computer malfunctions, etc.) and NOT that the cells connected to the Petrel died.

- brett

 

[beldridg]

I know people who slip a 6th cell into the center of the tray.

Another reason for not running a splitter, you have to mentally acknowledge that you are seeing one piece of data twice. If that data is bad you have to mentally process that you are looking at a single bit of bad data displayed across multiple displays. But that only applies for a single sensor which makes it even more confusing because you have to remember which on cell on each display is the shared data.

I think this is a pretty big factor to consider.

There are certainly ways to mitigate some of it. E.e., the left most cell on both Petrel & NERD is the one that is split. But, the mental gymnastics seem like a potential downside.

This has factored into my decision to not install a splitter (more on that in another post).

In terms of the 6th cell, I've heard about that as well. In my system, the potted RMS electronics are in the middle of the tray. Do you know if they have that same setup? It seems like it would cause the 6th cell to stick up too high and not allow the scrubber canister to fit? I personally haven't tried it.

- brett

 

[broncobowsher]

I have not tried the 6th cell, also have not looked at my friend's who has it. I have no potted electronics, guessing you are running RMS? That might be the difference.

If you had an airplane that had 5 gas tanks, but room for 6 gauges on the dash. Would you put two of those gauges on a single tank? Or remove/disable the extra gauge to avoid confusion?

 

[beldridg]

I have not tried the 6th cell, also have not looked at my friend's who has it. I have no potted electronics, guessing you are running RMS? That might be the difference.

Yes, I have RMS and that is what the potted electronics are. I'm guessing it would force a 6th cell to stick up too much.

If you had an airplane that had 5 gas tanks, but room for 6 gauges on the dash. Would you put two of those gauges on a single tank? Or remove/disable the extra gauge to avoid confusion?

Yes, this is a good way to think about it.

- brett

 

[Divermikey]

I don't see a need for it. For me, I like the 3rd display for the AI. The mini display for T1 and T2.
Don't need the same information repeated again, no need for a splitter.

I ran a Nerd 1 for 3+ years with a splitter and had no problems and it wasn't an issue to remember which cell was shared.
I recently replaced my Nerd 1 with a Nerd 2 and I also like the idea of displaying the 2 transmitters information rather than the 3rd cell. Gonna try it in a couple months down in the Caribbean.

 

[beldridg]

First, let me say "thanks" for the constructive dialog.

This is a rather long post but hopefully the people who dive a rEvo will find it worthwhile.

I find it helpful to get other viewpoints and opinions on topics like this. I don't think there is a "right or wrong" answer which is what makes it interesting. As I said in my original post, I would post my opinion and thoughts. I've got a "bonus" at the end as well.

IMHO, it doesn't make sense under normal circumstances to split a cell when you have a 5 cell configuration with 3 on the controller/primary and 2 on the backup/monitor. Here is my list of pros/cons:

Pros / Benefits

• Could help with a scenario when the monitor has two connected cells and the controller with three cells fails completely. In this case, the monitor has three cells.
• Could help with a scenario when one of the two cells on the monitor fails (i.e., the monitor could vote out the bad cell and still average the remaining two).

Cons / Costs

• Additional point of failure (splitter)
• Potentially an issue with "connecting" the two completely independent systems and it goes counter to the concept of independent systems
• Additional complexity (remembering which cell is split on each display, thinking about what it means if they are different, etc.)
  • There are ways to alleviate this such as putting the shared cell for both on the left, but it still involves "mental math" when the sensors start disagreeing
• Lose the ability to have AI info in the third PO2 "slot" in the middle row
  • Doesn't affect people not using AI. Also, on a rebreather, I really don't need the AI info all the time anyway and I can always just use a few right button pushes to have it displayed on the bottom row.

My conclusion a few years back was to not split a cell to put a third cell on the NERD display. I'm not claiming that it is the "right" decision, just the one that I have gone with.

However, I do have some additional thoughts.

These are based primarily on "expedition" style trips to remote places like Bikini or Chuuk, etc. I used to do those trips often but haven't in a while due to COVID but hope to get back to them.

Thoughts

• Carry a splitter in your bag on remote trips
  • Note that rEvo sells these with and without isolators (e.g., R436 - Cell Splitter 2)
  • There are failure cases on a remote trip where it can really make sense to split cells. I carry extra cells to avoid that but the "cost" of carrying a splitter is basically nothing

Now, another interesting topic (the "bonus") I mentioned above.

Let's assume that you are in a very remote location like Bikini Atoll with your rEvo. Your primary computer (controller) on the 5 pin wet-mate DiveCan bus takes a crap and is no longer functioning.

Most people would take the third molex connector on the hard-wired monitor cable and connect it to one of the other three cells that was connected to the controller. Now you have a computer monitoring PO2 with three cells. Not a controller, but no big problem either.

I also have a Teric that I use with an internal set point of whatever I'm diving so I have a backup but not connected to the system.

When I started diving in really remote places, I wanted to have the ability to connect a backup computer in case my primary computer failed. I originally had a NERD but upgraded to a NERD2 to get AI and other features (as others have done).

I kept the original NERD and I take it on "expedition" style trips. However, it cannot be used on a DiveCan bus with a 5 pin wet-mate since it is a 4 pin analog wet-mate connector.

Hmmmm...what to do? This was my solution:

“Expedition” readiness with my rEvo

I'm not saying it is right for everybody. In fact, for most people, it just adds complexity and doesn't make sense. However, if you are doing expedition style trips and depending upon your gear, it might make sense.

Regards,

- brett