EDIT: RESOLVED. Turned out that I had a '93 California-model ECU which had sequential injection by default, so my MS3 set up by default for batch mode was only firing two cylinders. Turned on sequential injection and everything was

Hey dudes,

I've been banging my head against this dead cylinder for a while now, have looked through lots of intertubes and forumz and yuhtubs, and now come in shame to y'all for some help.

Here's the sitch: '93 1.6L Miata, installed an MSPnPPro (MS3 Pnp)--including an AFM delete and an IAT install--and now cylinder #4 is dead. I'm running the base map provided with my MSPnPPro + setup steps called out in instructions (e.g., set timing). Innovate MTX-L wideband. NGK Iridium plugs. Magnecor wires. Racing Beat full exhaust. No other remotely engine-related modifications.

The car seemed to be running fine on all 4 cylinders before I did the MS-install + AFM-delete/IAT-install + spark-wire-swap (did all three at once).

How I know it's #4 that's dead, specifically:

1. I have exhaust headers installed and the #4 header can be held with bare hands while the engine's running, while the #1,2,3 headers all instantly boil off sprayed water, and
2. I went through and pulled spark plug wires while the engine was running and #4 wire changed nothing when it was pulled, where #1,2,3 all killed the engine.

What I've done so far:

1. Verified MS injector and spark wiring: used Tuner Studio injector/spark test mode to confirm that the right coils fire for each coil output and the right injectors fire for each injector output.
   
2. Verified ignition components are not the problem (note, didn't say "spark is not the problem"): Popped into injector/spark test mode in Tuner Studio, manually triggered each coil, confirmed that they were all triggering with a timing light on each wire. Then pulled the #4 spark plug out of the car, grounded it, and triggered it via TunerStudio and confirmed that it is in fact visibly generating spark (when outside of the car). So the coil, wire, and plug are all doing their jobs, though I can't confirm that the spark is actually firing inside the engine.
   
3. Cold compression test: I can't get the engine up to operating temperature on only 3 cylinders, so I had to settle for a slightly warm compression test. Everything seems hunky-dory. All cylinders clock in right around 190 psi.
   
4. Impedence tested the injectors: tested resistance with a multimeter and all four were within spec.
   
5. Listened to injectors, both while running and in test mode: I'm absolutely sure #1,2,3, are firing (very loud click through a screwdriver to my ear). #4 sounds like it's firing, but it's harder to get clear access and it does seem to be a little quieter, so since it's still batch injection it might just be the click of the #1 injector vibrating through the fuel rail. I have not removed the injectors or tried to analyze their flow.
   
6. Confirmed reasonable MAP, CLT, O2, and IAT readings: I'm getting right around MAP 100-101 kPa with ignition on but engine not started, and about 95 kPa while cranking, low on idle, etc.. IAT reads at room temp before start, and higher after engine starts up. CLT reading is within reason, though no matter how long I ran the engine on three cylinders, it never got fully up to temp (capped out at ~165, not sure if that's reasonable for one dead cylinder). O2 wideband sensor reads exactly the same in MegaSquirt/TunerStudio as what shows on the gauge, and the readings are within reason (I can't know if they're right).
   
7. Checked over all of the vacuum line changes: visually inspected for cracks/leaks, put clamps on the T joint that I used to T into the FPR vacuum line from the rear of the intake manifold. Again, MAP reading seems right so not too worried about it.
   
8. Checked fuel pressure: with pump on engine off (~40 PSI), at idle (~35 PSI), and with bumps of medium throttle (~35-45 PSI).

Total list of everything I've done that might possibly affect things:

• Installed '90-93 Miata MS3/MSPnPPro, running a slightly changed version of the base map that came loaded (added timing adjustment as per instructions, GM IAT calibration, etc.)
• Installed Magnecor competition wires
• Installed NGK Iridium IX spark plugs
• Installed GM IAT sensor, drilled and tapped into my DIY air intake
• Installed Innovate MTX-L wideband sensor and guage
• Installed 949 Racing / Supermiata QMAX coolant reroute
• Replaced CAS sensor o-ring
• Removed valve cover to paint it, clean it, and replaced valve cover gasket and re-sealed corners upon re-installation
• Removed OE airbox and air flow meter
• Removed A/C, heat, fan, stereo, power windows, door sensors, seat belt sensors, etc. and removed all of the wiring for those things, so maybe screwed up some other wire?
• Installed Racing Beat exhaust headers, midpipe, muffler

Current tune and composite log attached.

Any suggestions for what to learn / where to look next? I'm honestly not sure what to do next to figure out what's wrong, and don't really want to throw money at replacement parts if I don't know.

 

Fuel, air, spark, compression, timing.

Generally you want to run cheap sparkplugs (NGK BRK5E is stock iirc) when tuning so you can take them out and read them. Is your car california spec? Those were sequential injected, the rest were batch fired and the wires have a joint about 3 inches from the ECU plug if memory serve.. MS appears to be set up for batch. Wondering if you're only injecting on cylinders 1 and 2, but there's enough suction to pull fuel to cylinder 3 but not 4. Might try disconnecting injector 3 and see if it behaves the same.

3.0ms is a little low for stock coil dwell, wants to be closer to 4. 5ms/5.5ms for cranking.

 

I can't remember who it was, but there was someone with the exact same problem. Spark confirmed out of the engine and in test mode, but when the engine was "running", it only fired on 2 cylinders (that was a wasted spark setup, IIRC).
Maybe worth a deep search...

 

Check wiring! I had a Hydra installation that would intermittently drop one of the injectors due to a poorly soldered patch harness.
Wiggling the wires at the connection to the stock wiring harness confirmed the problem.
Sounds like you have lack of fuel.
You could pull plugs to verify. #4 will look NEW and the others will look a little bit dirty.
You could swap injectors around and see if the problem followed the injector.
My money would be on wiring problem.
Did you change injectors at the same time?
If so injector could be issue.
If not look to wiring.

 

HOLY FRIGGIN ***** BATMAN, YOU ARE MY HERO! I never would have tracked this down.

I don't live in CA, so just to double-check I did some Googling and it looks like the part number for the 93 CA ECU was "B6AW..." and guess what it says on my old ECU? Looks like I have a Cali ECU indeed. Man I never would have found this. Thank you, thank you!

Popped into my MS3, set up the jumpers as described in the instructions for sequential injection, set up sequential in tuner studio, and VOILA, running like a dream.

Seriously man. My hero. Can't thank you enough. I wasted weeks on this.

Alright I'll check this out. Thanks.

YO WTF why didn't I think of this?! I did the same for plugs, wires, and coils and for some reason didn't occur to me to try with the injectors. Good tip. Will try if I run into things like this in the future.

 

Just in case other people come along and find this, here's a list of part numbers for the 1993 Miata ECUs. If your ECU says "B6AW," you've got a Cali model.

Edit: image didn't link properly. Re-added it.

 

Props for updating the first post and explaining the solution in detail for future searchers.