Dana Myers wrote:
> The problem is that knock sensors wait for knock and then react.
> At that time more retard is necessary (Flying Miata give a number
> of 30%) than if the timing was set to prevent knock in the first > place.
I haven't measured the timing retard, that's one of the things I'd very much to do. I think an Auterra OBD-II kit would do the trick, but I need to hold off on spending another $200 until after the New Year :-).
I guess the problem is that the glowing deposits caused by the
> initial knock require more timing retard until they cool down.
I had not thought of that. However, it doesn't *seem* to be the case that the Mazda ECU is backing *way* off and staying there; it recovers quite quickly. I must admit I like the improved mid-range torque gained using WI instead of backing off on the ignition advance.
[...]
>>and one of them
>>>is new. Does the Link ECU run closed-loop under cruise?
>
>
> It is supposed to do lambda l3 tuning. I think the idea is
> to rapidly vary the mixture about 14.7 to allow the cat to
> do its work.
>
> I think the problem started when I replaced the link chip
> with the latest, but I am not sure. I did so much work on
> the car recently. Maybe I should not just have dumped the
> settings of the old chip onto the new.
Ah. More than one change at a time, eh? That can make diagnosis tough, but you know that already. Good luck.
Dana