Rough start and abrupt recovery



Hey -
I have an '06 Nissan Xterra off-road that came used and we’ve done a lot of work on, which I’ll list below. After all that, we’re still having a very persistent problem. Sometimes when you start it, it runs extremely rough. Any throttle will kill it right away. After doing this for a minute or two it very abruptly recovers to a normal idle and then drives mostly fine.

We’ve noticed that this seems to happen most when the car has only been off for a fairly short period of time and so is somewhat warm, and it also only seems to happen when the A/C is turned on. But both of these are not high-confidence, it really doesn’t do this very often either way.

Possibly related symptoms:

  • We had the SES light turn on with an O2 sensor, bank 1 sensor 1, out of range code on the second sensor. After a while it cleared on its own and hasn’t come back on since.
  • Sometimes at high-ish RPMs, say 3k and up, power gets surgey and the SES will flash which I believe indicates a misfire. More often it surges a bit without any flashing SES indication.
  • Sometimes right after pumping gas it won’t turn over. If you wait fifteen seconds and try a few times it starts up and runs fine.

Things we’ve done that are related:

  • A mechanic replaced all four O2 sensors as part of replacing the primary cats (burned out due to pre-existing MAF problem, see below), and when the SES came on for bank 1 sensor 1 after the replacement he replaced that sensor again. SES came on about this sensor again after that, but then cleared itself and hasn’t returned for a while. I still wonder if there could be a problem there, perhaps wiring or an exhaust leak or something.
  • MAF was bad and air intake manifold was leaking slightly when we purchased it, replaced both. Also replaced spark plugs as they were both not OEM-recommended model (NGK Laser) and fowled up, probably due to bad combustion due to bad MAF. Wires looked OK.
  • A/C was completely depressurized when we got it, we replaced all O-rings, pumped down, and recharged. But in the process of doing so we found a bit of metal mesh in one of the AC lines, which we think is from an internal failure of the expansion valve. A/C works fine now but we are worried about possible debris in the compressor. Compressor sounds okay, but as said, it’s possible this problem is related to the A/C.
  • We suspected the problem starting after pumping gas was related to the canister purge valve, it seemed fine but we “fired the parts canon” and replaced it anyway. That symptom still happens occasionally though.

In almost certainly unrelated service, we’ve also replaced shocks all around and replaced limit switches on the transfer case (but it’s still acting up with slightly different symptoms, possibly wiring? haven’t looked at this yet since it still actually works fine, the dashboard indicator is just wrong. Could be as simple as a loose connector).

I don’t have a ton of theories right now. It seems like the temperature sensor is possibly a suspect if it’s causing weird cold/warm idle detection behavior (would this show up as odd behavior on the temp gauge?), maybe it’s something about the EGR somethings, possibly a lingering problem with the O2 sensor, or something about fuel delivery?

At this point I’m thinking about paying the Nissan dealer for diagnostics to see if their CONSULT II box will tell them anything useful, I’m not seeing anything that strikes me as odd looking at realtime data in Torque but I also haven’t been able to get important things like the fuel pressure. I do have the short-term and long-term trim numbers but to be honest I don’t know enough about fuel injection and ECMs to make much of these. I’d love any ideas.


Hmm, 12 year old car running rich? And it still seems to run rich. You posted a long story but didn’t say how many miles on the engine. Have you replaced the thermostat, or temp sensor? Does it get up to temp quickly?

Are the short and long fuel trims high or low? Post them please. And any data you collected on coolant temp sensors and the fuel pressures you measured. Otherwise, we’ll be just taking a stab at it.

So stab I will.

Flashing SES at over 3K is bad. You should pull over if it continues to flash. It sounds like you are running out of fuel flow. That says clogged fuel filter, bad fuel pump or damaged fuel lines or a bad fuel pressure regulator. If the car doesn’t have a fuel pressure sensor or Torque can’t “see” it, get a mechanical one and tape it to the outside of the windshield and take it for a drive. It will tell you a LOT. If it drops low over 3K, change the filter and re-test. If it still drops, I’d suspect the pump. It might vary based on manifold pressure IF it is referenced to that. If the pressure varies all over the map regardless of RPM, suspect the regulator.

Post back and we’ll keep trying to help.


Whoops, long post but missed a few useful facts. The mileage is about 140k, it’s certainly not young. Have not replaced thermostat or temp sensor. It seems to get up to temp just fine, but it’s a bit hard to tell on that for sure as, as I said, we’re getting 90+ weather here and so the temp gauge is often off the bottom peg already when you start it just from sun.

I’ll collect the trim numbers next time I drive it, but the big trick is that it doesn’t do this very often and I haven’t found a way to reliably reproduce it, so I haven’t been able to get trim numbers for when it’s idling rough. I haven’t seen the OBDII reported temperature do anything strange either, I can record that as well next drive. I haven’t been able to get any kind of fuel pressure measurement unfortunately and I don’t have a gauge but I think I’ll see if I can borrow one, because I also suspect that fuel pressure might be an issue. Unfortunately the filter is integral to the pump assembly and the pump assembly can’t be accessed without dropping the tank, so I’ve been reticent to do that.


Bummer! No access panel? Can you cut your own? I did on my GM SUV. It made fuel pump and integral filter replacement easy.


It sounds like one of the problems (there may be other) is the fuel to air mixture. The idea above to secure a fuel trim measurement is where to start on that. You don’t have to do that necessarily when the problem is occurring, as whatever’s causing it will likely still be causing an abnormal fuel trim. The car engine will run pretty good with a fuel trim off by 20% either way, but a 20% fuel trim means something is definitely amiss.

fyi, a +20% fuel trim means the computer has to inject 20% more gasoline to satisfy the pre-cat O2 sensor, when otherwise it would be injecting less gasoline if based on the amount of airflow into the engine (the maf sensor alone). The maf sensor and the O2 sensors are supposed to match up more or less, and when they don’t a fuel trim reading results.