....(I had a number of them fail on the boat, usually because of operator error, but they shouldn't be so easy to make fail, IMO)...
I’d be very interested in hearing what the “operator error” fails were. We try to make the Cobalt interface work in such a way that users can’t cause any problems by doing something wrong. Of course, if they enter an incorrect O2 %, we can’t guard against that. But there shouldn’t be anything a diver could do that would make the Cobalt actually malfunction. What kind of failure are you referring to? As interface designers this is important feedback.
I don't like AI computers with QDs. I don't like them on the Cobalt, or the Smartcom, or on the Suunto. The failure point is that if a drop of water gets in the QD, they fail. And many folks want to take their QD apart on every dive.
At least for the Cobalt, this is not the case. You can dive it without the QD attached and water won’t hurt anything. The HP sensor face is an inert ceramic. We recommend keeping salt deposits from building up, and rinsing if salt water gets into the HP opening. Water is not a problem.
It seems to me that the PR disaster is because folks like the ones in this thread can't seem to get their computers fixed even on the third try. Atomic needs to make this right and come up with a final solution.
I can’t dispute that this has been a PR disaster- particularly the depth sensor issue that coincided with the Cobalt 2 release, because that halted production for months until the issue was fixed by the sensor manufacturer. That meant that some customers waited a
long time for repair or replacement. This same sensor is used by many other dive computer manufacturers (including Shearwater). Atomic- and Cobalt purchasers- just had the misfortune to be on the receiving end of a packaging change from the Swiss manufacturer that created a very difficult-to-test-for failure.
But for some perspective- I’ve been involved with diving product development for over 30 years, and I’ve been very familiar with a large number of product launches. I’ve seen on the other side of the curtain for some dive computers that are well known, some that had early failure rates around what the Cobalt did initially. One difference is that Atomic has been very open and public about any problems. That has not always been the case with other manufacturers. Another is that Atomic has a wide and global distribution- you have to take absolute numbers sold into account when assessing internet feedback. Another is that the Cobalt, because of its unique market position, cost, and user interface, attracted a lot of attention. None of that is an excuse for problems, but it’s important to bear in mind when looking at these kinds of threads.
Some people experienced multiple failures- though not always the same thing. The major issues with the Cobalt were water leaks, HP O-ring leaks, rusting button magnets, and sensor failures. While the majority of users had none of these issues, some had all of them. I don’t blame users who experienced multiple issues for losing confidence.
As to fixing the problems, from what I have seen- and this is borne out by actual numbers from the last couple of years- Atomic moved aggressively to alter the physical design and the manufacturing process to prevent any failures that turned up in earlier production. They moved to very extensive wet testing of every computer prior to shipping after the sensor problems. But the fact that many people dive infrequently, and that there are a lot of Cobalts in the field, means that earlier problem units still turn up. I’m guessing that had Atomic repackaged the Cobalt so it looked different, and given it a new name, we would be seeing threads about how much more reliable the “new” computer was. Unfortunately I'm afraid the reputation from early issues will persist long after they have been fixed.
Ron
