I have a 1998 Altima SE with a 5-speed

I was bug bombing my house on a hot day, so I had the dog and cat in the car, driving to my brothers house. I was sitting at a stop light, and the engine RPM dropped down to between 200-500 for about 5-10 seconds, and then died. Not the most convenient situation for that to happen...

Not sure if this makes a difference, but note this: I had the a/c on. When the a/c compressor engages, it causes a little belt squeal for a split second-it had been running like this for a month and I hadn't got around to troubleshoot it.

My brother and I checked that it had spark. We heard the fuel pump turn on for 5 seconds when the key is moved to "on", we even put some gas in the tank just in case. We verified that the camshafts were moving while cranking. Later that day I had it towed back to my place. I checked everything again. Put it on the charger overnight.

The next day, I put a fuel pressure gauge in the line just upstream of the fuel pressure regulator. I got 45-50 psi if I remember correctly. I also noticed that the pressure would have slight, quick downward spikes of about 1 psi, and would recover. I took that as indicative of the injectors working correctly.

I pulled out the fuel pump fuse and sprayed some carb cleaner into the air intake, just downstream of the maf sensor, and it started for a little bit. I put the fuse back in, and it started and idled fine. Now I was confused.

I knew there was something wrong, so I pulled the codes:
P0325 - Knock sensor
P0400 - EGR system
P0440 - Evap small leak
P1448 - vent control valve
P1706 - P/N pos switch

I tested and verified the knock sensor was bad, so I replaced the knock sensor.
I don't think the P/N switch is bad, since it wont crank when the clutch pedal isn't down.
I didn't investigate the EGR system
I didn't try to find an evap system leak
I didn't look at the evap vent control valve

I tried to start it after that, and it didn't start. I put carb cleaner back into the air intake, and it still didn't start. Though I think I didn't take the fuel pump fuse out that time.

I thought maybe a timing chain skipped, so I took off the valve cover and verified the upper chain is aligned, and the lower chain does line up with the upper gear now and then, but that doesn't verify it's good. The tensioners are good, and the chains are tight.

I removed all spark plugs, put a little oil in the cylinders, and a little carb cleaner in each cylinder. Cranked it, and it sputtered a little.

I hooked up the OBD2 reader again, and tried to get sensor data such as RPM. I could read the ECT was at 60F which was right, and throttle position sensor was good, but when I crank, all data went away, and one of the obd2 programs I'm using says there is some sort of OBD bus short.

I put in a new battery, and that problem went away, Now I can get sensor data while cranking. The RPM is about 200, MAF sensor value changes from .17 g/s before cranking to about 4.1 g/s while cranking with pedal down, Barometric pressure sensor looks good at 98 kPa, Timing advance is at 3 degrees before cranking, but within about 1-2 seconds of cranking it zooms up to 20 degrees and stays there.

A few weeks went by, because I had to go on a business trip. When I came back, I did a compression test:
Cylinder 1 : 137 psi
Cylinder 2 : 150 psi
Cylinder 3 : 145 psi
Cylinder 4 : 120 psi, 140 psi after about 1 tsp oil

Checked spark again. All cylinders have spark

Sprayed carb cleaner into the vacuum line on brake booster, downstream of check valve. Cranked it and nothing.

While cranking during the various tests, pretty regularly it will do a slight "puff" backfire right when I stop cranking when I have the pedal down.

I checked injectors just for kicks. They are all at 11 Ohms, which is in spec. I checked #3 with a incandescent bulb tester; when key is on, I get a light on one of the terminals. While cranking, I still get a light, but it's dim, and I can't see pulses, probably because I was using a bulb and not an led.

I got a Hantek 6022BE pc oscilloscope, and made a cylinder pressure sensor from an a/c pressure transducer and a trip to ace hardware. I'll plot cylinder pressure vs. the spark igniter signal today and definitively know if the timing is off.

I was able to plot the cylinder pressure vs igniter signal


I output the data, and analyzed in excel


I found that the time between peaks is .6738 sec, and the time between when the spark fires and TDC is .02 sec. Noting that the crankshaft makes 720 degrees of rotation between peaks, I find that the spark is firing 21 degrees before TDC

Could the 20 degree timing advance that I noticed in the sensor data via OBD2 account for this?

After talking with my dad, 20 degrees BTDC might be normal. Is that right? anybody know?

I tried using starting fluid in the intake rather than carb cleaner. Still nothing happened. I pulled out a plug, and had spark, and sprayed a little starting fluid in before putting the plug back in. Cranked it and got more of a response. Does this mean that there might be a huge vacuum leak?

I had it towed to Pep Boys. It was funny that while I was waiting for the tow truck I tried starting it one last time, and it started. It ran for about 15 minutes, and I could rev it and everything. When the tow truck showed up, it died. Pep boys replaced the fuel pump after determining that it wasn't energizing. I don't know what that's all about, I've always heard it run for 5 seconds when the key is brought to the ON position, and I tested that it had fuel pressure. They were able to test drive it for a mile, and it died. Then they figured it was the distributor. Even though it had a spark, it must have had a weak spark. They replaced the distributor, and it has been running fine since.

Conclusion: Even if you have spark, you may have a weak spark. If you have tested everything, and can't figure out why your car wont start, consider replacing the ignition coil(s). The theory is that it may spark at atmospheric pressure, but at 150 psi, the air is 10 times more dense, and has a larger dielectric breakdown voltage than atmospheric air as described by Paschen's Law http://en.wikipedia.org/wiki/Paschen%27s_Law.