Hi, we are the original developers of the Cobalt's user interface, electronics, and firmware, and implemented the RGBM algorithm that it runs. The Cobalt's RGBM implementation is in the middle of the road as far as conservatism is concerned, neither the most conservative or liberal for typical profiles. At 45-50M you are getting into the depths where the Cobalt's algorithm migrates to a fully iterative RGBM calculation. That could result in different sorts of ascent profiles, but shouldn't show dramatic differences in total times. There is no minimum deco time in the Cobalt's algorithm, you could get deco stop totals as short as a minute.
We would really like to look at your profiles to see why you are experiencing such a difference with other computers.
It would also help to know to know what model of Cobalt (1 or 2) you have, and what firmware version is installed.
I will send you a message ("start a conversation" in ScubaBoard speak) with links to software and instructions that will allow for the dive log memory to be sent to us. That's valuable for analysis because it lets us see all of the settings and individual tissue groups.
Just ran a couple of simulations, and if you were at 45M for the entire 15 minutes, you could end up with an 18 minute ascent time. If you went a few meters deeper that could increase due to the differences between the "folded" and "fully iterative" calculation.
To answer your question about Buhlmann, there isn't any way to incorporate another algorithm into the Cobalt- it's just not like a phone or other device that runs an OS. However, we have developed a new Cobalt version that allows the user to switch, on the fly, between the Atomic RGBM implementation and Buhlmann. It's available to Atomic if they choose to go that route.
We have recently started a ScubaBoard Manufacturer's forum under Ocean Concepts to support the Cobalt computers.
Ocean Concepts
-Ron