I've got a 1994 miata with a 2000 motor, running all 94 wiring / sensors, but with 2000 injectors.
Got my megasquirt installed last weekend. www.spitfireefi.com
did the build and install work. I'm full standalone with this build. Since I didn't build the board, I don't know if it follows any of the known how-to's, or if they derived everything themselves (they do this sort of thing for a living). If anyone has specific questions, I can try and get answers.
1) Cooling fan relay is strange in the 94. It's always got +12v to one side of the relay input, and to trigger the relay, you have to put +12v to the other side, so there's 0V across it, and it'll flip. I don't have it working yet, but I have a circuit to try that looks surprisingly similar (if not identical) to the PWM idle mod.
2) Not controlling idle yet. I'm going to try the idle mod at some point.
3) Idle does surprisingly not-bad for having no idle valve control, and not having my map fully tuned yet.
4) The tach is strange. While the wiring diagrams make it look like the tach has nothing to do with the ECU, it still didn't want to work without the stock ECU plugged in. I don't remember the details, but sticking a resister in the diagnostic connector (probably between tach and ground) made it come back. More investigation is warranted here.
5) I cut the wires off a stock O2 sensor, and used the heater circuit +12v / gnd for my Zeitronix Wideband, but the readings the megasquirt saw were off from what the Zeitronix showed. I switched the ground to one shared with the MS, and that helped, but the MS still reads about .2 on the lean side. I'll probably switch the +12v over as well. Then I'll use one of the wires to run simulated narrowband into the stock harness so that if I put the stock ECU back in, it'll get an O2 signal.
6) We ran the IAT back through the stock wiring to the ECU connector using one of the a/c control lines from the big fat gray connector near the passenger headlight. There's a total of 3 wires there that go back to the ECU connector, so I should be able to run my boost control and VICS out through there later. One other setups, they have used the MAF lines to run the signal, but the location is better for IAT on the A/C harness connector, and it makes it that much easier to swap back to the stock ECU if I need to.
7) I'm ditching EGR, Evap Purge, and the fuel pressure control solenoid. I may try to use the fuel pressure control solenoid to control the VICS. But VICS will probably become moot when I add turbo later.
8) spitfire made me a nice little adapter harness that goes from the stock ECU connector to the megasquirt, and the MS is going to sit in the stock location behind the seat.
Somehow my spark dwell settings got pooched (set to 75% duty cycle) between first start and when I was actually working on tuning. It ran fine for a while (a couple hundred miles of driving), but last night my coils died. Hitting the "Dec" button in the warmup wizard too many times in rapid succession may or may not have also had some play in this (because it seems to burn settings when you hit dec, which could maybe overheat the coils like when you leave them connected while doing a firmware flash). So I'm going to use 2000 coils and forego the tach until my new-to-me 94-97 ebay coils come in.
In spite of the coil issue, the car feels like it's running stronger than it ever has since I put in the 2000 motor.
I have only a small pang of regret about not building everything myself. But when I consider the fact that it was only about 21 hours elapsed from unhooking the stock ECU to driving home on the megasquirt, I'm all smiles. Now that I have an up-and-running setup, I'm really tempted to start building an MS-II to replace it (because my friend with a 94 miata wants MS, so I could sell him mine), for higher resolution and C source code understandability.