Cobalt Deep Stops

[Akimbo]

I have a Cobalt 1 and Gap Dive-Planner with RGBM. I have noticed that the Cobalt calls for a lot of short deep stops compared to the RGBM implementation on Gap. Two others onboard during a recent trip noticed similar use of short deep stops on their Cobalts. All the other different brand computers onboard called for their deepest stop at 20'.

For example, on a dive with a max depth of 165' the Cobalt’s first stop was at 80' for a minute and every 10' for 1-5 minutes until hitting 20'. Gap’s first stop was at 20'. The total decompression time was in the same ball park.

Don’t take this as a complaint. I am just curious to know more about different RGBM implementations.

 

[RonR]

Not sure how I missed this post for a month, but it’s a good question, sorry it took so long to respond.


On dives deeper than 150’ the Cobalt switches over to a fully iterative RGBM algorithm. RGBM implementations in other dive computers are actually adding factors to a Haldanian model to attempt to reduce bubble formation. This “folded” method is much less computationally intensive than the fully iterative calculation, and makes more sense for shallower dives- where typically divers are concerned about no-deco limits. RGBM actually doesn't generate NDL's. The Cobalt was the first computer to offer a “full” RGBM for deeper dives, running the actual iterative calculations in real time- though a VPM algorithm had been available on Liquivision computers for a while, and they now have a full RGBM version available as well.


I’m not sure why the GAP implementation would be dramatically different- I don’t think it should be overall. However the use of 10’ intervals for stops on the Cobalt is somewhat arbitrary- the algorithm could generate a schedule at about any reasonable interval. A “full” RGBM schedule will typically start decompression at about half maximum depth with what is essentially a deep stop (on the Cobalt, for simplicity’s sake we keep the deco stop label), and then proceed in a series of short stops to the surface. The goal is to keep bubbles from forming in the first place, not to reabsorb them after they form. but the use of 10’ or 3 m intervals is an interface decision, not dictated by the algorithm. We could use 15’, or 20’, with longer times at each stop.


You can use the Cobalt’s built in simulator to compare profiles and schedules with desktop software or other computers. If you create the profile in the Cobalt’s simulator, then press the BACK button when starting the ascent and select “Save”, the simulated dive will be saved in the Sim Dive Log. You can look at the schedule by selecting the View Saved Dives option and then the Show Deco button. I just did a 165’ for 10 minute bottom time, and yes, stops start at 80’ and proceed every 10’ in gradually increasing lengths with the last one at 15’ for 6 minutes and a total ascent of 24 minutes. The same dive to only 145’- which doesn’t quite trigger full RGBM- yields a stop at 80, but then just at 50, 30, and 15, and a 10 minute ascent time.

Ron

 

[Akimbo]

Thanks for the detailed reply. To see if I understand, does RGBM basically call for a continual ascent and progressively slower rates and the Cobalt converts it to stops in 10' increments for practicality? Again, no complaint. The only way you could do such slow ascents (like the last 20') is in a chamber.

I am glad to hear that RGBM is available on the Liquivision now. Is it the same implementation as Atomic's?

 

[RonR]

Yes, that's basically it. The rate of ascent and not violating parameters that would cause bubbles to grow is what is important, the exact stop intervals are arbitrary and intended to be reasonably easy to follow.

The RGBM implementation for Liquivision was done by a 3rd party, who I believe sells it independently through Liquivision. It's based on exactly the same underlying code as the implementation we did, so it should end up being very similar, given similar choices for conservatism.

Ron