I recently took ownership of another Miata (possible build thread to follow) and I'm trying to get it running. The car is a 1990 with a 1.8 swap (don't know the year of the motor), still using the 1.6 ECU/AFM/etc.
The PO was driving down the road and the car just died with no warning. Now it will start, run for a second, then stall. The tachometer does not move when the car starts.
I swapped the CAS with the one from my fully functioning 1996 Miata, which made zero difference. Per the PO's diagnosis the AFM is still working, but I will hopefully be borrowing a known good one to try.
OBS codes: there are two longer flashes followed by size short flashes, which I believe makes that code 26. "Solenoid valve (Evaporative canister purge".
If it's a 94 motor the tach signal comes from the coils.
Sorry was in a rush earlier. I guess the way you can tell a 94 is the extra pin on the coil pack.
I believe you can look up the computer number also to see what year it is.
Runs for a second then stops is characteristic of a missing "keepalive" signal from the AFM. Use a paperclip to jump from FP to GND in the diagnostic box, see if that solves the problem. It's not a permanent fix, but it will diagnose it.
If you are swapping the AFM for some reason, do NOT disconnect the two tempting screws beside the connector. If you do that and pull, you'll break the fragile wiring inside. Usually this results in a missing keepalive signal from the AFM...
The proper way to disconnect the AFM is to remove the big square C clip holding the connector on.
The coils will still be 1990 coils. If the car starts, it's getting a good CAS signal.
Interesting on the keepalive.
The car died while driving down a dirt road, and has some kind of lowering Springs and Konis, is that wire fragile enough to break from vibrations? This car has definitely be ridden hard to say the least.
Will try jumping thone pins tomorrow and report back.
A modified car that's 27 years old? Anything's possible. That AFM may have been molested before.
A good point 
Jumped F/P to GND, no change.
Good to know. It's an easy test at least.
How long does it run? Just a quick burst and die (ie, about one or two ignition events per cylinder), or sustained combustion for a second or two?
It runs for a second or two.
Runs smooth, no funky noises or anything, then dies.
About that FP jumper test - when you turn the key on, do you hear the pump running? Because this sounds exactly like the keepalive problem. Weird.
I'll poll the other guys in the office and see who can come up with something.
Yes, I do hear the fuel pump running with the key on.
Somehow you need to find a way to figure out if it's fuel, air, or spark.
I'd doubt that it's spark related, but you can check that by seeing if the spark continues as the engine runs down for the stall.
I also doubt it's air FLOW related- as the original failure happened while the throttle was open- so even if the idle motor slammed shut, it would work. It could be air-measurement related. Most engines have starting fuel that does not use whatever is measuring air flow (MAF or MAP).
You are already on the fuel part. The question I have- does the pump turn off when the engine is running?
With that jumper in place, that should force the pump to run with ignition no matter what. But I would like to know what the fuel pressure is during all of this. It should stay steady. If the pressure drops with the jumper in place, you know the pump's having trouble.
Here's a thought - if the flapper in the AFM doesn't move, the car won't get any fuel past the initial prime squirt. Check for a big air leak, probably between the AFM and the throttle body. Also check under the throttle body to see if there's a leak on the inlet hose for the IAC. It's about 3/4".
A thought from Bill - if there's an intermittent connection problem at the CAS or coils, it could be pulling loose when the car fires and rocks over. A wiggle test at the wires would be in order, but that also would have happened when you swapped the CAS.
The fuel pump is continuing to run after the car shuts off, I can't hear it over the engine when it's running though. Will try to come up with a fuel pressure tester.
I don't see any obvious major air leaks. Will grab a second pair of eyes to see if the flapper is moving (no air box or filter on it right now). Would it be a worthwhile test to prop it open with something and see what the car does?
Pulled out the ECU, no obvious issues with the connectors or anything like that. The chassis ground under the brake booster was disgusting so I cleaned that. No dice yet.
Edit: everything appears to be securely plugged in, but I will do a more thorough wriggle test back there. The coil mounting situation is pretty minimal...
Messing with the MAF flapper would not be a bad thing, but I have a hunch that the fuel pressure test will say a lot. (high pressure- in the injector circuits, low- probably fuel pump)
But one thing to note if I get these systems right ( they are different than ours) IF the pump is going, then the maf is at least partially working.
The jumper bypasses the fuel pump circuit in the AFM. Normally, the pump comes on for a prime shot, during cranking and whenever airflow is detected. With the jumper in place, it's tied to the ignition and it's a very simple circuit.
The air leak will be very major if there is one. The IAC connection is hard to see, and with a conversion like this there might be a certain level of jank in the pre-TB intake tract.
If it has a flapper, see if it moves smoothly and returns to its stop. I bought an MR2 Turbo for $1200 that was doing this. It would start, run a few seconds and flood out. Towed it home, found a spare AFM in the pile of parts that came with it, swapped it out and it ran great.
The flapper door was stuck about 1/2" open, so it would start on the cold start injector and then flood out.
If an MR2 has no AFM connection, it will start, run 2 seconds and die. Rinse / repeat. I think it's getting fuel and spark.
I assume there is no aftermarket alarm or immobilizer that just came out of dormancy?
The level of jank is pretty high throughout the vehicle. Nothing would surprise me.
What confuses me is where the tach not working plays into the running issue.
Edit: the AFM flapper does move and return smoothly. I don't see any evidence of an alarm or anything.
More years ago than I care to remember a guy had his Vega towed to my grandpa's mechanic shop.
It was doing similar things. After hours of frustration someone said "hey, I wonder if the timing belt slipped". That was really genius in the mid-70's since most cars had timing chains. That is what it turned out to be. We proved it by trying to line up the timing marks and found it was not possible.
I Second the air leak. I just had that problem aver changing a bad alternator on a Focus. It took me a while with the smoke tester but I finally found a large vacuum line off. Try starting it at 1/2 throttle. If it runs longer or just really rough it is probably a vacuum leak.
Throttle position makes no difference in how or how long it runs. It will rev up and dit just as quickly as it idles and dies.
IMHO you really need to get a fuel pressure gauge to eliminate that as a possibility.
