I pulled the plugs on cylinders A2 (always a misfire code) and A3 (never a misfire code). They only have about 25 miles on them but I can't see any differences. In the photo #3 is on the left and #2 on the right. With the right lighting they look identical.

 

Did you ever check your fuel pressure, as suggested in post #5?
Can't seem to find your reply.
I can only think about this, or a problem with knock sensors and/or timing for above symptoms.

 

I posted fuel pressure back in post #13 (I had to go back and look myself, I'd forgotten if I'd posted it or not, too many posts!), pressure was 38psi or 44psi depending on whether the vacuum line to the regulator was attached. When revving the engine it dropped to about 36 but that was in the driveway. I'm going to check the pressure again while driving up a hill and see if maybe it's dropping when driving.

For timing, cams on both banks seem to be in synch based on position of the flat spots on the cams. I haven't done anything yet with knock sensors.

Thanks again to everyone for the great ideas, this has been very helpful.

 

OK, sorry, overlooked that one obviously.

Just to be clear, I meant ignition timing, not valve timing.
Was just thinking the old way of fuel, air and spark.

You really need to get some live readings on what is going on for various riding conditions.
With those, you will be able to check your trims (fuel), your MAF (air) and the spark advance, and might be able to focus in on one of them as the culprit.

 

Since you seem more than willing to pull the ignition cover off the cam cover, maybe time to really check the wiring continuity. It is beginning to sound like a pinched wire in the ignition bundle. Maybe a careful inch by inch check of each wire, a cleaning of the coils, and a careful assembly of the whole shebang.

A couple of years ago, on the XK8 forum, someone had the same problem, I and few others were positive it was a bad cam tensioner (all misfire codes on one bank); wound up being dirty coils.

Also, check the knock sensor on the right side as Eric suggested.

I sure wish Bob, Sean, or Brutal would toss a thought.

 

David,

Regarding your sense that the problem is load-related, can your scanner give you Live Data so you can watch your MAFS signal in real time? It would be interesting to know if its signal is in the right ballpark at different engine speeds and if it changes appropriately when you are ascending hills.

Here's another crazy idea: is it possible you have just enough water in the bottom of the fuel tank that when you start up a hill it spills over the rim of the surge pot and is inhaled by the fuel pump, but when you are on level ground the water is outside and below the rim of the surge pot? The actual surge pot is not as tall as it appears in this diagram from the Engine Management System manual:



I can't remember if you've changed the fuel filter?

Just trying to think of everything.

Cheers,

Don

 

ericjansen: based on earlier recommendations, I have ordered an ELM327 code reader and should have it Monday or Tuesday. My current reader is old school and only downloads the codes. With the new reader installed, I plan to hook up my fuel pressure gauge and take it for another drive on my hilly circuit.

jhartz: I have new coils all around, and have swapped them around trying to get the codes to move but so far that hasn't changed anything. I definitely plan to go over the wiring carefully if the new code reader doesn't show anything useful. I am definitely getting power to all coils but I know there are other signals that get transmitted to them. And I haven't touched knock sensors, so those get added to the list.

Don: draining the tank is also on my to-do list, along with a new filter, as intermittent water in the fuel could account for the symptoms.

Unfortunately, I leave this weekend for out-of-state work for the next week so not much will happen in the garage. But I've added all these great suggestions to the list. Thanks!

 

Back on the Jag after a week in Boise ID, a really cool city. I got my ELM327 and let the car idle in the driveway for several minutes. Got the coolant temp up to 190F. The short term fuel trim stayed in the negative range, running between -10 and -21. If I revved the engine, the STFT get up to 0, then fell back to negative numbers. Air/Fuel ratio was running about 11.9:1. I didn't need a gauge to know the engine was idling like crap.

I know the coils are getting power, but is there anything else to check in the wiring to know if they are firing? I assume there is some type of signal to tell them to fire, other than the constant 12 volts. I feel like the wiring may be an issue but cannot see anything wrong just by looking at them.

 

Forgot to mention, I drained the fuel tank, and added fresh fuel, along with an additive to help remove water just in case any was left.

MAFS were running 1.1 to 1.4, and MAP was 35-42 at idle, jumping up to 64 when revved.

 

That is WAY to low.

 

Hi David,

The STFTs are high rich, as is the AFR. What are the O2 sensors doing?

I assume the MAFS reading is in pounds/minute, and the MAP reading is in pounds per square inch? If so, I think they're OK if the MAFS reading was at idle. As Eric says, in grams/second the MAFS reading is much too low. It's helpful if you can post the units of these measurements so we know we're not looking at grams/second, Pascals, etc.

The coils are connected to ground by the ECM to fire them. With an inductive/capacitive sensor, you can test the coils in situ. Or if you have a Coil-On-Plug tester, you can remove a coi, plug in the COP tester, crank the engine and watch for spark. If you don't have a COP tester, you can remove a spark plug, insert it in the coil, and with a jumper cable connect the plug's threaded sleeve to a good ground, such as the engine ground strap. Crank the engine and watch for spark. It helps to do this in the shade or dark because it can be hard to see the spark in bright sunlight.

Cheers,

Don

 

I had an extra XJ8 MAF on the shelf in my garage for two years and I JUST got rid of it Friday! I have a couple of XJR MAFs but I guess they are different.

You can get a re-manufactured one for about $130 after your core refund.

 

LTFTs are what?

 

It should not drop to 36psi in the driveway no matter how much you rev.

Since you are also fading going up hills, the fuel pump and pressure regulator
are a good bet.

The reference level is 3 BAR, but I forget if this is against atmospheric,
or manifold pressure.

Come to think of it, probably manifold pressure since on the SC cars
we have to hold 3BAR above boost to be healthy.

If you pinch off the pressure regulator return and the problem goes
away, it is the pressure regulator.

 

I would get some time on the road before looking at fuel trims, since just idling in the driveway may not have the engine fully up to temperature - just cruising is fine. But those fuel trims are crazy-high - almost maxxed out. Related, those spark plugs look pretty black for being brand new, also suggesting a rich mixture. Plus the reported A:F!

36psi is low fuel pressure, but not criminally low. I would think the car would be fine as long as it doesn't drop lower than that.

The low MAF reading is double-weird, since I would expect a really low MAF signal to also be running lean, but yours isn't as evidenced by the plugs and the fuel trims.

Perhaps you have an obstruction on the intake or exhaust... that would result in a poor flow across the MAF, an apparent overly-rich mixture, poor idle, acceptable cruise, and a loss of power.

With the information at hand, I would try disconnecting the manifolds at the cats and see if things change. I'm assuming you've already thoroughly checked the intake system for a good air filter, etc.

 

You can also try disconnect the MAF.

It will probably stall the engine (as it will change to a default mode), but you can keep it running playing the accelerator, and see if the mixtures change.
If memory serves well, it will still run a bit rich, but should make a significant change over what you have now.

 

Out on a limb: I think the temperature sensor is bad. It keeps telling the ECU that the engine is cold and keeps it in starting mode.

 

Fuel trims should be closer to 0, fluctuating slightly, anything over 5 and you have an issue somewhere.




The fuel pressure should never go down when raising the RPM's, it should go up. Have you taken the vacuum hose of the fuel pressure regulator, is there any fuel coming out?




Definitely no vacuum leaks anywhere?

 

Thanks to everyone for the great suggestions, I'm slowly working my way through them. Because of work I haven't had as much time as needed in the garage but am doing a few things each night. Some of them are repeats just to be sure I haven't missed anything too simple. Last night I cleaned the MAF and connector again with CRC, and gave it several applications to be sure all the parts are shiny clean, followed by a hard reboot. This evening I used my coil-on-plug tester on Bank 1, and all are getting a good healthy spark.

Next I hooked up the ELM OBDII reader and attempted to drive around the neighborhood, but it's running so rough I couldn't go very far. I believe I got it close to operating temperature, so here are the readings (not sure what some of them mean yet but will research tonight):

MAF (lb/min): about 1.3
FRP (psi): N/A
MAP (kPa): about 52
Load %: 38
Timing (degrees): around 5 to 8 positive, varies
STFT%: -25.0 unless revved, then moves toward 0
LTFT%: -20.3 unless revved, then moves toward 0
Coolant Temp: 194F
Lambda: 0.75 (rich)
AFR: around 11.0:1
O2 Sensor voltage- N/A
Vacuum: average about -8psi
Intake temp: 95F

I think that's all I saw. Initially the only code was P0172 but it was running really rough so I'm sure there will be more.

 

Just to reiterate, if David's MAFS reading is in pounds/minute, it isn't low. 1.1 to 1.4 pounds/minute is 8.31 to 10.58 grams/second. That's actually on the high side at idle. This is why it is so important to know the units.