An EFI (or other) issue that is driving me NUTS

[RickWI]

Recently I refreshed my motor and it now is exhibiting a driveability issue that I simply can not resolve. On the refresh I reringed it, rebuilt the shaft rockers, installed new bearings and resealed everything. So nothing major that should have affected the tune. I run a F.A.S.T. EFI system and have had this installed for the past 4 years.

What is happening is under most all light throttle accel conditions as well as under cruise conditions it surges/misfires/runs unstable. However you want to define it is subject to interpretation I guess. I suppose the best way to describe it would be a surge type feeling.

Doing datalogs I have noticed that my target A/F ratios versus O2 feedback track excellent under these conditions. If I set the cruise target to 14.2 the fueling tracks precise. Under light accel the same, very close fueling versus target, wtihin 5%. I can move timing up or down from 40 degrees to 30 degrees and there is no difference in the way the motor reacts, from a "surge" standpoint. What I can do is RICHEN the target A/F down to the mid to low 13's and the surge is greatly reduced. Now thoughts are you give the engine what it wants. If it likes low 13's for cruise, that's where you set it. My problem with that is over the years I have been able to cruise under similar conditions as lean as 14.9 with no problems what so ever. My other problem with that premise is running under this rich an A/F turns my plugs to coal. Therefore all I think I'm doing when I richen the mixture is to mask the underlying problem.

The other symptom when it surges/misfires/runs like crap is the EGT's swing wild. Typically in the past I would cruise in the 1380 - 1400 degree range. When the surge occurs EGT's will swing up close to 1500 then flicker to 1300 or 1250, back up to 1400, they are all over the place. This tells me something is funky with combustion.

Under moderate throttle and wide open the engine runs great, never better or more responsive.

Here is what I have done so far:

I have verified crank trigger and cam sync signal. Both are correct.
I have verified all plug wires and coil wire are in excellent shape
I have inspected for leaks at the headers, TB and anywhere else I can think of.
I have run in open as well as closed loop with no change
I have verified valve lash is set to spec.
All lifters and rockers are operation properly and the camshaft is in perfect shape.
Inspected wiring to verify that grounds are connected.
Verified all sensors are working correctly.
All the plugs were brand new on the refresh and appear in good working order.

NOW, here is a kicker. One of the reasons I refreshed the motor is I had an intake leak the sucked coolant. I know coolant can be hard on O2 sensors. Readouts on the O2 via the FAST though seem normal. I think the O2 is still accurate as plug readings show a drastic change from running the engine at 14.4 down to the 13.2 to 13.4 range. They went from a decent brown to coal tar black. I have also tried to tune without the O2 in closed loop with no change in engine operation, still "surges".

I am ssimply stumped on this. My gut centers around an ignition related problem. I have run a Blaster 2 coil and MSD Pro Billet distributor for the past 4 years with no issues. The cap and rotor are brand new and in perfect shape. Coil ohms out fine. Crank trigger is setup as it always has been and timing at idle is rock steady, although idle is not an issue. It idles fine at 14.5:1.

I am stumped to beat the band and would appreciate any thoughts.

 

[ijames]

If by coolant you mean antifreeze, then yes, it is hard on the O2 sensor. My understanding is that the silicates in the anti-corrosion additives block the pores of the sensor (just like lead) and make it lazier and lazier. Also, keep in mind that you absolutely cannot trust any O2 sensor data when the engine is misfiring. The oxygen in the cylinder that misfires goes into the exhaust along with the unburnt fuel, and some portion of the oxygen and fuel may react in the exhaust before the sensor, but not all, so the sensor sees some amount of excess oxygen and reports that the engine is lean. Running closed loop the ecm will then dump in extra, uneeded fuel and the cylinders that are firing will be running rich while the O2 may show rich, lean, or near-stoich from the misfires. If you are running the same tune, injectors, and fuel pressure that you used before the rebuild then I strongly suspect some kind of vacuum leak. Maybe do a leakdown test to check the rings and valve sealing, and then rig up an adaptor to your throttle body and do a low pressure leakdown test on the intake manifold (pull the pushrods first). Anyway, my $0.02 .

 

[norbs]

Look for random spikes in the o2 sensor data on light part throttle cruise. Or rapid swings of correction, this is how i found out my module was bad. If you have an lm-1 you can see this easily on a log, as the response time its real quick.

 

[RickWI]

Carl, agree with everything you said. I suspect though if running the engine in open loop and VE cells still need to be high to run, relative to the past base tune, then that basically takes the O2 out of the pic. It may be lagging, and may not be correct for reads but with no feedback if the engine requires lots of fuel still then something else is amiss.

I can do one better on the leak test as we have a smoker at the shop. This device will produce smoke (like a smoke machine at a club) and create a slight pressure within the crankcase. Within a few minutes any seepage points will begin to show up. We use this mainly to located oil leak points but it would work excellent to locate any potential vac leaks.

Based on oil consumption as well as the tailpipe I think the rings took a quick seat. What I am going to do right now though is pull each plug and visually inspect for varience. It's a major undertaking to pull the back 4 plugs when you have 2" equal length headers and relocated angle plug All Pro heads. A PITA that I'd not wish on even a Furd driver.

Norbs, I have run data logs with various resolution on the FAST and I'll be dang if everything looks just fine under the poor operating conditions. Doesn't matter if it's under closed loop or open loop.

 

[RickWI]

Well pulled all the plugs and every one looks exactly alike. Very rich.

 

[ijames]

I think some kind of intake leak is more likely than the O2 sensor. The smoker test should work great for you; the intake leakdown test is kind of nice for us turbo guys because it lets us find things like intercooler hoses that only swell and leak under boost (or that don't have clamps ). Any chance of a problem with the plumbing to the map sensor, or the sensor itself?

 

[RickWI]

I agree, I think the O2 is OK as well. It seems the map sensor is working properly and is reading as it always has but I can replace with a spare I have in the shop.

I need to inspect that intake as a potential problem like you said. Visual looks good but you never know.

It might also be worth changing the coil just to eliminate a variable.

 

[norbs]

lets see a datalog of the stumble?

 

[RickWI]

Norb, the laptop that I have the FAST system on does not have a modem on it. I don't even have a site on the internet to download it to. Being that I don't have access to broadband and am on VERY slow dial up I'm limited.

I just got back in from driving it after changing out the plugs. I datalogged the bad area and can paint a picture for you.

RPM is dead steady at 2560.
Target A/F is at 14.5.
Actual A/F is above and below that .2 a/f maybe three times a second. Basically occillates slowly up and down.
Map is steady.
ignition timing is steady
TPS is steady

Looking at the datalog you'd say it's running fine. Driving it you wonder where all the cylinders are.

Tonight it ran smooth with a 12.9 A/F ratio.

What a PITA this is turning out to be.

 

[norbs]

Rick the logs are very small you can upload them here as an attachment. As long as you have winzip, you create a new archieve name it, then "add fiiles to the archeive. Wish i could tell you an answer, Does this surge appear to be a lean surge or a direct stumble? miss

 

[RickWI]

Norb, it acts (feels) like it is between a lean surge and a miss. On my motor I have had past situation where a plug wire is burned causing a similar feel. Typically we associate a miss as a hard "feel" when it happens. With this motor that type of miss when I had the bad plug wires (7/8) was not symptimatic of that. It just ran crappy. This is a similar situation only more pronounced. It is almost like it looses a touch of power, then gets it back, then looses it, then gets it back. You can see it on the EGT's as they swing up and down although I can't say I really see a direct pattern between temps and the "surging" but in general they are very unstable. So from a exhaust standpoint something is happening. From an O2 sensor standpoint I don't see a thing.

Mechanically I can't think of what could be causing it either. I have inspected the valvetrain, checked and verified all lash, reset lash a tad looser all to no affect. Same problem.

There is no doubt though it "acts" like a lean surge as it really, really wants more fuel to make it better. Of course thats really hard on the plugs as well as the pocketbook.

 

[norbs]

Just for a quick test put a timing light on each plug wire at a time, watch and see if one cyl, is skipping a beat in the pulses. This is a clear indication of a misfire. The egt change is also a sign or misfire. Keep trying

 

[RickWI]

I'd have to do that while driving down the road as it idles great, or appears to. Nonetheless, doing it at idle could show something. It's on the list for tomorrow.