Rebuilt 3-cylinder Miata motor (long)
So my oil pump failure last fall killed a couple of rod bearings. I pulled the motor and dropped it off with John Day for a rebuild. Bored .020 over, new pistons, rings, bearings, exhintake cam swap, 5-angle valve job, block and head shaved to ensure flat surfaces, etc...
While he had the motor, I replaced all the water hoses, the front trans seal, the clutch, and a few other sundry items. I installed a new header - and have a 3 month-old Magnaflow cat.
I re-installed the motor and fired it up. It seemed to run well, but had way too much HLA noise. I tried idling it, driving it gently around the neighborhood, nothing helped. I talked with John and we decided that re-using the old lifters may not have been the best choice - so I bought a new set of lifters and he installed them.
After he installed the lifters, he started the car idled it a bit, then took it out for a short drive. When he first started it, he says it was running smoothly - and he did let it idle until it was warm and the idle dropped to it's usual 800-900 rpm. After a short drive, it started idling very roughly. As it was late on a Saturday and he had birthday dinner plans with his wife, I agreed to get the car home and we'd tackle it again on Monday.
I gently drove the car home (3 miles) and when I got there it was running perfectly. It sat for a couple hours, then I fired it up again, set the ignition timing, and took it out for a spin. It seemed to be running great, but after driving it for 20 minutes or so, it was running very roughly again, and has ever since. (A week and a half) So here's the diagnostic work done so far:
Air:
* New filter, OEM intake system
* Problem seems to be confined to cylinders 3 & 4, with 4 being much worse
* Valvetrain is now quiet, and when it was running smoothly - it seemed to be running quite well
* Gary and I checked leakdown - 8% to 10% across all 4 cylinders
* Gary and I checked compression - #1/175, #2/165, #3/165, #4/150
* At idle, vacuum gauge reads ~15 in. Hg - vibrating +/- ~1"
* Checked for vacuum leaks and found none
* Intake manifold bolts are tight and I used a new gasket
Fuel:
* Certainly have fuel pressure, though I haven't measured it
* No visible (or sniffable) leaks at rail
*Using my stethoscope, I think I can hear all four injectors clicking, even when it's idling poorly
*When the car is barely idling, I can unplug the #4 injector and there's no change. If I unplug #3 injector, there's little to no change. Unplug #1 or #2, the engine will die
*I've pulled the injectors, cleaned them, replaced the o-rings and the "insulators" that seal them to the head, and intalled them back in different holes - the problem stayed on cylinder 4
*No obvious visible damage to injector wiring harness - wiggling, pushing, pulling on the connectors and harness itself while the car is running doesn't seem to make any difference
*I checked the wiring to the injectors from the ECU - it's continuous
*I also checked to ensure the switched ground lead wasn't continuously grounded
Spark:
*Used a "spark checker" and verified spark at all four cylinders
*Plugs were new with the rebuild, have only 50 miles or so on them. John swapped 2 of them trying to diagnose the problem on when it first occurred, but the problem stayed with #4
*Pulled the plugs and 1, 2, & 3 look like new plugs 4 was BLACK, like it was fouled
*Relatively new Magnecore wires, correctly installed
*Swapped in a set of NGK blues (with under 1000 miles) from a friend's car, no change
*Gapped and installed another set of new plugs - no change
*Swapped #1 & #4 wires at the coil - no change
Timing
*John installed the (new) timing belt after the lifter replacement - feels confident of valve timing.
*Yes, it has the exhintake cam swap, but it HAS run smoothly since the swap.
*After I brought the car home, during the time it was idling smoothly, I adjusted the timing with my timing light to ~12 deg
*I have tried adjusting timing over a pretty wide range while it's idling poorly and while I can tell I'm making a change, nothing seems to help the 3-cylinder thing
*I've checked the wiring to the CAS: +12, +5, +5, GND
*I've also verified the air gap on the crank angle sensor, FWIW
OBD2
* I have had a CEL a couple of times during this fiasco -- code reader said P0300 - random multiple misfire
* CEL will clear itself if left alone
* O2 sensor readings are jumping around a lot, like incomplete (or not at all) combustion in at least one cylinder (tell me something I don't know)
* new Bosch O2 sensor in the header - older one (don't know if it's orignal or not) post cat
Other:
Gary and I put a thermocouple on the primaries of the header and found that #1 & #2 are hot, #3 cooler, #4 cold (relatively)
Anyone got any bright ideas?
