tedean

I have quite a mystery on my hands. Being one that makes every attempt to understand (and repair) things, I am at my wits end. I will be dragging the cup to the local expert next week, however I thought I might see if you guys had any ideas.

Here are the facts;

During 2011, ran several practice sessions and three club races with no issues.
Last race was Miller Sept 2011, car ran perfectly.
Engine has 70 hours, rebuilt by PMNA.
Highest rev 9235rpm
Ran VP 110 (RO 110) leaded fuel all 2011 season, 0% oxygen (why, don't ask..)
Drained fuel
Towed back to Denver, stored in heated garage, did not start for 2 months

Restarted engine to prep for 2012 season.
Very rough idle, sounded like one cylinder was missing.
After warmup, 60C, revving engine to 3k, sounds like one cylinder is not performing.
Proceeded with preventative maintenance
New plugs
New Coil Packs
Fuel for 2012 season is Sunoco 260SX (RO 98) Unleaded, 0% Oxygen (car has a fuel cell)
Still has a rough idle and is rough on run up to 3k
Did infrared scan, found #6 header pipe at lower temp.
Performed leak down test on #5 and #6, less than 5%, 6% respectively
With #6 plug in, coilpack out with another plug, verified spark did exist
With #5 plug in, coilpack out with another plug, verified spark was same intensity as #6
Pulled right injector bank, verified with 9V battery each injector solenoid functioned, injectors were not plugged.

At idle, moved #6 injector electrical socket to identify a possible loose connection or wire. Idle and at 3k remained the same, rough.

Readjusted idle to prevent stalling.

Took car to track for testing
Engine felt like it ran on 5 cylinders most of the time.
Engine would clear out and run correctly at times.
Engine did consistently stumbled above 7k rpm. Pumping the throttle twice seemed to cause the engine to clear and run correctly until redline.
Once, existing the pits, engine was stumbling, then cleared, only to stumble at times on track.

Modus (I have cable, software) reported Lambda error, then after clearing, error did not return.
Suspecting contaminated Lambda sensors due to lead fuel, both Lambdas were replaced.

Engine continued to run roughly.

Modas shows no errors.
At 2200rpm, turning off 1-6 cylinders individually does cause drop in rpm as expected, however it appears #6 causes less rpm drop.

Pulled plugs #5, #6, #5 black, #6 ceramic was tan see pics attached.

MoTec data shows no abnormalities in
fuel pressure
crankcase pressure
2011 data showed Lambda #1, #2 consistent, tracking together
2012 original Lambda #1, #2 not tracking consistently together. (replaced Lambdas as noted above).

Thermal scans clearly show #6 header considerably cooler than #1-5.

Quite a mystery if you ask me, but maybe is is something simple?

What do you think guys, any ideas?

amondc

Looks like #6 is not getting fuel. Have you checked fuel volume or just pressure? Had a pump going out, pressure was there but volume wasn't. You say you checked the injector, did you verify fuel was passing thru it? Or just listen to it click? Might try swapping injectors with #5 and see if the miss moves.

J richard

^^^+1 all signs point to a fuel issue. Given the sooty plug looks like its leaking too much fuel. Alternatively a broken or soft valve spring can give you some weird results...

cstreit

I'll confess that when the first thing you mentioned was a high rev count and rough running, I thought a dinged valve or bad spring... Failing that (have you done a compression/leakdown on that cylinder?) probably fuel... run test on the injector and on the injector signal iteself...

Dan Jacobs

Sure sounds like a stuck injector. Pull all 6 and send them to Marren www.injector.com or someone similar

tedean

I did perform a leak down test on 5 and 6. They were both around 5%, so all good there. I am told 9.2k max rev would not float the valves.
Regarding fuel flow, the cell has 4 pumps running to fill the reservoir, plus I changed the fuel filter last year. Also, the fuel feed for that bank intersects in the middle of the rail, so #6 is not, as one might say, at the end of the line.

I might switch injectors 5 and 6, but they are very hard to get to. Not impossible though.

tedean

I did pull that side, individually tested each injector with a 9v battery while blowing thru them. Of course that was not much pressure.

J richard

You can have a weak valve spring or even one that has snapped off a few coils that will hold the valve shut for a leakdown but will cause a cylinder to drop out at revs. Don't forget that as a motor sits several of the valvesprings are sitting completely compressed for extended length of time.

As far as the injectors it does not take much for an injector to wreak havoc, including not closing completely, poor spray pattern or not opening fully, any one of which can only be tested on a rig...

cscrogham

Not to add any misdirection to some excellent advice, but have you replaced the O2 sensor?
Leaded fuel will kill them off pretty quickly and often won't throw a code, the ECU thinks its getting good info and altering the map. Not sure if your car is running closed loop or open to be affected by the lambda.
Might be worth a shot...
Best of luck.

cscrogham

Duh, I guess I should read before posting. Even though you already replaced them, you might try disconnecting them and see how the car runs, then ck the connectors and reset that code again. I just can't help but think the fuel caused a problem there.

tedean

Correction, the fuel feed is at one end, so #6 is at the end of the line, however the rail is large enough to feed evenly.

I did swap #4 injector with #6 and had the same results. #6 is still a on go. At lease this eliminates the actual injector itself.

So it leaves ECU/wiring or valve spring?

turbochad

ECU/ wiring.

Cheyne

Put a lght on the #6 injector harness

Dan Jacobs

What he said Got any mice in your garage?

tedean

J Richard wins the free lunch, broken outer intake valve spring. Wow, weird results is correct, I would have never guessed that. Thanks for everyones thoughts!