As Doppler is saying, it is clear that the algorithm your dive computer is using to calculate your no-deco status is unhappy with the time you are spending deeper. Some are; some will give credit for the compartments which are offgassing during deeper stops. A lot also depends on the ascent rate you use to GET to the deeper stops; many people go much slower than they think they do, and if you review Marroni's papers on ascent strategy, a very slow ascent from the bottom was one of the worst for bubble generation.
But the bottom line is that, when you use a particular model, you need to comply with the assumptions and constraints of the model. You are using a brain-derived adjustment to your ascent strategy that is not compatible with the computer you are using. Change your ascent strategy, or change your computer.
But the bottom line is that, when you use a particular model, you need to comply with the assumptions and constraints of the model. You are using a brain-derived adjustment to your ascent strategy that is not compatible with the computer you are using. Change your ascent strategy, or change your computer.