Resurrecting this thread, as there are some significant updates, and the thread comes up awfully quickly when you Google "miata vvt swap".
The car now runs, and for the small area I've tuned (<4k rpm), runs well. I haven't gotten the over-run tuned at all yet, VVT isn't enabled, and the idle still hunts a smidge, but overall, success.
There's not much mechanically to update on: I reused my 94 intake tube, with the GM IAT sensor stuffed into the old IAC tube connection. I believe Keith wrote about using this location during his Targa build, way back when that car was a 4-cylinder. I cobbled together a connection between the 2.5" (more or less) intake tube and the 3" (more or so) Prelude filter I've been using. I also sourced a 2001 throttle cable and fixed that issue. Other than that and an EGR block-off, everything else is off the 94.
I had a one minor wiring mistake, plus one bad connection, and one mis-adjusted resistor. I believe the car actually fired on the first try, but I was too busy worrying about other things to pay attention.
Here is the description of the modifications I had to make in the engine bay.
The "sensor" column is mislabeled, it should be just "object". Whatever.
The coloring only applies to my particular combination of engine and chassis, you'll have to do your own work in linking my names to your harnesses.
As I said earlier, the "Pressure Regulator Control Valve" is the odd doohickey that sits in the valley between the intake manifold and the head on the 94. I believe it has something to do with fuel pressure and vapor lock? Anyhow, it's quite obviously perfectly placed to be used for the VVT solenoid.
And here is the wiring at the MegaSquirt. This is almost completely in line with what Frank recommends on his WestfieldMX5 page.
I believe I may have reversed the two fans. I'll have to check my work and update.
The car ran great immediately, but refused to rev over 2250/2500rpm. It sounded bad when I eased up there, yet my wideband wasn't showing lean. To cut a long story short, I used the MegaSquirt's logging capabilities and noticed that I was losing synch there. Then I watched the "sync loss counter" on MegaLogViewer, saw it incrementing, and saw the "sync reason" value point to loss of cam signal. After some fiddling, I checked the voltage of the pull-up (I think that's what it is) on the cam line and discovered it was out of spec. This is R11 trimpot on the MS3X board. It's supposed to give a value of between 2.5 and 3v at a specific pad, and I had dialed this in on the bench, but the value when attached to the car was 5v! With that, no more sync loss and I can rev!
The LC-1 wideband was another issue. Though the instructions say (counterintuitively) that the power and signal grounds from the LC-1 should be attached at the same place, this is not what worked for me. I got all sorts of noise on my TPS and CLT measurements (though oddly enough, not on the IAT) when I had the LC-1 wired like that. What has worked now was to wire the damn thing in a sensible manner, using the stock O2 sensor connector. The coloring is problematic, as there are two black wires on that connector, but it's possible. Sensor ground to sensor ground, heater ground to heater ground, done. Ignore Innovate's instructions. I could explain how in reality the sensor is actually grounded at the same place as the heater, but this post is long enough already!
Where does that leave us?
Street tuning. I've had a devil of a time finding good basemaps for a NA VVT engine. I finally settled on one and found that I was regularly going into the 17:1 AFR range while tuning. Not liking this, I asked around here for advice, then settled on a tuning method that leaves me on the rich side of the equation as things are tuned. This place is great.
VVT was definitely throwing me off, you'll find my post at the end of the thread above describing how I should have approached setting those values. I'll get around to that next time I'm out at the garage.
