Any reported cases of Ox Tox between 1.4 and 1.6?

[AeroSpartacus]

Some of these will start oxygen flushing at 30ft / 9m because "it's easier to hit 1.6" when in reality they will be off-gassing better at 10 ft / 3 metres at a ppO2 of ~1.3

Not your point, but I will start flushing at ~30 ft if I remember for verifying cells but I don't hang out there for deco

 

[Nick_Radov]

uints should be abolished from all „control“ software tbh, not like memory is such a bottleneck nowadays (or a really specific usecase)

Wait this is _still_ out there? Such an overflow surely must have been fixed already

Unfortunately it's only partially fixed. The FIT file format defines the various message types for CNS% as uint8 which would be difficult to change at this point. Garmin did change the CNS% calculation algorithm in 2022 so that if you exceed 1.6atm for a few seconds it no longer goes totally wacky and instead now basically copies Shearwater's algorithm, which although still arbitrary at least seems more reasonable. So in practice the risk of an overflow is greatly reduced. But since the CNS% algorithm is worthless anyway I just ignore the computer warnings about exceeding 100% and stick within the commonly accepted tech diving oxygen exposure guidelines. Garmin makes excellent hardware but some of their software design decisions leave me scratching my head and wondering if their employees do any real diving?

Shearwater does have a good article "Oxygen Seizures at PO2 ≤ 1.6 Bar: How Rare?" which is relevant to the original question on this thread and summarizes the available research as of 2017.

 

[Kendall Raine]

Curiosity pulled the cats tail, and I found myself here: https://cumbria-hyperbaric.org/wp-content/uploads/2019/09/repex0.pdf
I will dig in now that I realize repex is an algorithm

See also this article (it gets the exponent sign correct...)

 

[Kendall Raine]

See also this article (it gets the exponent sign correct...)

Here is a link to the original 1988 Hamilton, et. al. source document. Too big to upload.

https://www.google.com/url?sa=t&source=web&rct=j&opi=89978449&url=https://repository.library.noaa.gov/view/noaa/60673/noaa_60673_DS1.pdf&ved=2ahUKEwiznvaelM2JAxUIIUQIHfZ-E3kQFnoECCAQAQ&usg=AOvVaw0ZXdlbqc1iXg7FtAJDE7qR

 

[JMarc]

Unfortunately it's only partially fixed. The FIT file format defines the various message types for CNS% as uint8 which would be difficult to change at this point.

It should still be possible to compute a "correct" value and use a saturated one when protocols and files formats mandate the use of a field unable to represent it.