Hi all,

I'm at wit's end trying to get my EF XR6 to run properly, and need some help.

I'm running an AU block with an EF XR6 Tickford head and am running a TI Performance J3 chip. I am self tuning using the full-house tuning pack, and have disabled O2 sensor feedback, so engine operation is controlled completely by the tables. The transmission is automatic with a high-stall (2600-2800 RPM, depending on torque). Exhaust has no catalyst (yep, legal in NZ).

The problem is that the AFR in a couple areas will not respond to the changes I make in the VE table, so is effectively uncontrollable.
- at 1500-2300 RPM and MAP below about 10 in HgA - any degree of overrun or deceleration - the AFR goes super lean - over 19:1, right off the scale. Adding more fuel (VE) in multiple 10% increments makes no difference, possibly even makes it worse. Cruising down very slight slopes on the highway, it will misfire - surging and hesitating. Decelerating down long hills, the engine will flame out completely and stall.
- at idle (750 RPM and 10-11 HgA), the AFR is very rich - around 12.2 - causing a rich misfire and a very rough idle. Leaning out the fuel (reducing VE) multiple times made no difference.

Recent changes before these problems:
- Mildly ported the XR6 head and skimmed it for 10:1 compression
- Swapped my previous 195/192/114 cam for a 208/198/115 cam
- Replaced the factory Bosch 0 280 155 844 19.8lb/215cc injectors with genuine Bosch 0 280 155 931 24.7lb/267cc (GM LS1) - and changed the Injector Slope in the tune to match

Other work already done - most chasing this problem:
- replaced ECU with tested unit (tried two others, as well)
- new J3 chip
- new Goss fuel pump
- new fuel filter
- new Bosch 300kpa fuel pressure regulator
- cleaned fuel rail and flex hoses, checked for obstructions
- new Bosch coilpack
- new Autolite 985 spark plugs
- near new Bosch Motorsport ignition leads - resistance tested fine
- new MAP sensor
- new coolant temp sensor
- tested throttle position sensor voltage - all good
- tested compression - 165-190 psi (most at 170)
- replaced idle air control valve with good used one
- replaced cam angle sensor with good used one (to clear diagnostic error)
- checked timing marks on crank pulley - within 2-3 degrees
- checked base timing - all good
- checked diagnostic codes - all clear
- checked for vacuum leaks, spraying Brakleen on all joins between TB and head and on all vacuum hoses - no RPM increase
- burns no oil
- spark plugs have good colour
- new Bosch WBO2 sensor (for AFR monitoring), in case it was giving bogus AFR readings

Maybe a slight clue: The reason I changed to the larger fuel injectors is that with the previous head (ported EF non-Tickford) the tune kept drifting lean at WOT, so I had to keep richening it up multiple times until I maxed out at 100%. The larger injectors give me more duty cycle to play with.

It is also well down on power. It is at least half a second slower in the quarter than it was with the old oil-burning 350,000km EF bottom end, stock EF XR6 head, 9.2:1 compression and milder cam.

Has anyone on Fordmods experienced anything similar or have any suggestions? I'm just about ready to throw the SOHC engine out and go Barra. But if I don't know what's causing my problem, there's a chance I will still have it, even with a new engine and ECU. Any help would be hugely appreciated.

I haven't done any J3 chip tuning but I have one question. Why are you tuning without the O2 readings? As these change throughout the rev range, the ECU need them to determine the AFRs.

I haven't heard of anyone doing this method of programming before.

I'm effectively running in open loop all the time. As long as the tables are right in the tune, you don't gain anything from closed loop. Running solely from the tables gives you more control. For example, you can tune it to run leaner than factory at cruise to give better fuel economy.

I do use a wide-band O2 sensor, but only to log actual AFRs in order to check and adjust the tune.

I have been running off the tables, without 02 feedback, for about 6 years. It has worked great until this recent problem where the system is not injecting the fuel that I am commanding at certain setpoints. Something must be interfering with or overriding it.

Did you input all the injector data? Just changing the slopes isn't enough, it needs the full range of data - injector offset, low and high slope, breakpoint etc.