1996 was the first year that Mazda installed a crankshaft wheel on the Miata, though it was used purely for OBD-II misfire detection. The '96 and '97 models do have a raw VR sensor on the crank for this purpose. However, they also still have the same 4/2 (4G63) CAS as the older models, which continues to be the primary trigger for spark and fuel. These cars could still be (and usually are) wired and configured identically to the '90-'95 models. Alternately, if you want the extra precision offered by the crankwheel, it would be possible to do a hybrid input system on these two years, with the VR input driving the primary trigger as raw VR and the 2 hole portion of the CAS driving the secondary trigger as open-collector. I've not tried this configuration, however my instinct is that the '99 software settings would still work in that mode, assuming the second trigger doesn't mind being single-single as opposed to double-single. Someone with a '96 or '97 would need to verify the tooth orientation on the wheel, as I'm not sure they conform to the same degree standard as the later cars. Certainly however, the most common configuration on '96 and '97 is to wire and configure them exactly as a '90-'95, and ignore the crankwheel.
Ok, enough speculation.
In '99 (there was no '98 in the US, and in other countries the '98 was a held-over '97), when they restyled the car, they also changed the trigger mechanism. The 4G63 CAS was eliminated, and the new-style crank and cam sensors were installed. Both of these sensors have open-collector outputs- they are
not VR sensors.
There's been ongoing debate on the MiataTurbo.net forum as to the best way of dealing with these sensors. Initially, most '99+ owners elected to install a CAS from an earlier car- there is still a provision for it on the head and the cam- and configure it just like a '90-'97 car. Abe Mira was one of the early pioneers of installing an MS2 on an NB ('99-'05) Miata. Initially, he built input circuits for it that were just like those specified in DIYAutoTune's writeup for the earlier cars, which can be seen here:
http://www.diyautotune.com/tech_article ... _miata.htm Basically, just the stock hall effect input and a simple resistor/capacitor secondary input, unbuffered, straight into the CPU. He had considerable problems with this setup, misfires and lost sync. He took it upon himself to reverse-engineer the inputs on the OEM ECU, and came up with a circuit that worked perfectly using a dual JFET op-amp. I have reproduced that circuit in clean schematic form with the appropriate pinout, and it's linked to below. Feel free to copy this and use it in the documentation.
One last area of puzzlement for me is that throughout the MS2Extra documentation, as well as within Megatune, the later Miata is referred to as "99-00" only, to the exclusion of the '01-'05 cars. I'm not absolutely 100% certain on this point as I've never personally scoped an '01+ Miata, however I strongly believe that they use the exact same sensors and trigger layout as the '99-'00 cars. I wonder if someone knows why this is stated the way it is? Certainly the '01+ cars have the additional complication of variable intake cam timing not present on the '99-'00, however I suspect that this could be easily overcome, either by pinning the came in place (this is known to work) or by using the EBC driver in open-loop mode to control the VVT solenoid. That's a job for someone else, as I don't have an '01+ at my disposal to experiment with. Pretty much everything else about those cars is identical to their earlier ('99-'00) brethren.
For reference, here is the original thread where we've been hashing this out:
http://www.miataturbo.net/forum/showthread.php?t=22530 Note that many of the drawings in that thread are speculative and/or just plain wrong.
