Overheat at Idle

[McQue]

sending the fuel trim article to your gmail address as a pdf document

 

[TheSloMoShow]

Thanks, it's a good read.

I'll have to disconnect some of my injectors while idling, and pressurize when pulled out and still connected to the rail to see if it confirms my suspicion of a leaking injector (s).

Injectors, O2 sensors, radiator. I feel it has to be one of these 3 components causing the problem. Likely first two, but I'll keep the radiator in mind.

I'll also check my IAT/MAP/IAC/TPS to make sure all are responding correctly and aren't dirty/malfunctioning. Worth a check.

 

[Ecomike]

For some reason I feel a leaking injector would show up on its respective spark plug.

Have you probed the manifold flange with carb cleaner?

I was thinking the same things, exactly. And old man is right, it looks like all the plugs are reading a wee bit lean, thus it is not likely to be leaking injectors. I would suspect 1-2 leaking injectors would show carbon on the ends. An oring(s) on the injectors could be bad and letting air in... making lean....

In the world of science, there are two types of errors (thinking O2 sensors), systemic error would be the O2 sensors output reading being a little richer than the exhaust really is, and the computer compensating based on biased data. An exhaust leak near the O2 sensors can cause that, or a bad O2 sensor wire/connection biasing the data to the computer, and a vacuum leak can make it run a little lean IF the computer does not compensate with more fuel at idle.

 

[Ecomike]

For what its worth, I chased my tail for years with my 87, getting incremental improvements doing/replacing a dozen things, but the last real fix was finding a bad donut in the exhaust manifold flange to exhaust pipe that was 3/4" from the oil pan that would blow hot exhaust on the oil pan and super/thin the oil, and run the engine about 25F hotter than normal. 3 Muffler shops missed it in idle tests. Found it by accident when swapping out the AW4.

Thanks, it's a good read.

I'll have to disconnect some of my injectors while idling, and pressurize when pulled out and still connected to the rail to see if it confirms my suspicion of a leaking injector (s).

Injectors, O2 sensors, radiator. I feel it has to be one of these 3 components causing the problem. Likely first two, but I'll keep the radiator in mind.

I'll also check my IAT/MAP/IAC/TPS to make sure all are responding correctly and aren't dirty/malfunctioning. Worth a check.

 

[TheSloMoShow]

What if it was just one or possibly two leaking injector per bank? Wouldn't the O2 sensor in each bank register rich, and adjust the injector pulse width run to reduce fuel, effectively running the 2 good cylinders lean (because they were already stoich) and running the leaking injector cylinder closer to stoich (albeit prob still a little rich)?

I'm going to pull them anyways just to check, and make sure the o-ring seals are 100% and not letting any unregulated air in. But In the case the injectors check out, then it would have to be the O2 sensors or exhaust leak right? Wouldn't know why else they are registering rich...

I did use some carb cleaner last weekend and probed the intake manifold but didn't get any indication that there is a vacuum leak.

Sent from my SM-G900V using Tapatalk

 

[Ecomike]

Can't argue any of those points, but I am not sure what the computer sampling rate is for your year, and how much overlap there is between cylinders in the O2 conc in the exhaust pulses for each cyl, and what the response time is of the O2 sensor to changes from one cyl to the next, or how fast the computer can adjust, which all could affect the answer to your model and mine.

I suspect that computer is using the last sample data point, maybe adjusted for a complex algorithmic average of prior data points, to set the next injectors open time. In other words the computer does not know which injector/cyl is responsible for the last data point (not precisely anyway), and does not sample and adjust injector by injector, it does it bank by bank on yours.... and mine only has one sensor on the Renix jeeps, 87-90, with a slower computer.

 

[Ecomike]

"What if it was just one or possibly two leaking injector per bank? Wouldn't the O2 sensor in each bank register rich, and adjust the injector pulse width run to reduce fuel, effectively running the 2 good cylinders lean (because they were already stoich) and running the leaking injector cylinder closer to stoich (albeit prob still a little rich)?"

In that case 1-2 injectors would normal and 1-2 would lean. All six of yours, based on the photos look lean

 

[TheSloMoShow]

Ok - good point.

Since I'll have the rail and intake out this week to look at the injectors, I'll check first that the manifold bolts are torqued correctly and that there isn't an obious crack or gasket issue. I think this should give me enough space to have access to all bolts.

Given one doesn't present itself, what's your opinion on best method for detecting hairline crack? I've heard spraying soapy substance on the suspect area, or, throwing some Seafoam down the TB and starting engine back up after 30 min will produce enough smoke to tell.

Sent from my SM-G900V using Tapatalk

 

[TheSloMoShow]

Hey Ecomike, I'm trying to run this through my head but am not coming up with anything.

How is it that an exhaust leak could fool the O2's into thinking the engine is running rich?

My thought would be that an exhaust leak (more notably at the exhaust manigold gasket and before upstream O2) would have lower pressure than the outside air, and the draw O2 in - leaning out the mixture, and tricking the O2 it thinking engine is running lean.

Or would this only be the case at idle, where scavenging is most present? I.e., when cruising the internal pressure is actually larger forcing the exhaust to exert O2 (and not combustion) to the outside atmosphere and trick the O2 it thinking the mixture is too rich?

Sent from my SM-G900V using Tapatalk

 

[Ecomike]

The ceramic part of the O2 sensor breaths and senses the difference (called partial pressure difference) of the O2 concentration on the inside (Exhaust, depleted, used up, less O2) and the outside O2 concentration.

If the outside O2 concentration is diluted by an exhaust leak at or near the O2 sensor it changes the differential partial pressure, and fools the O2 sensor into thinking the exhaust is not what it really is, makes it think it is lean IIRC? and the computer compensates by adding fuel (which throws out my earlier thesis) But if one fuel injector is leaking, it can confuse the computer..... and make it run otherwise normal stoich cyls a wee bit leaner. Also the scrubbing effect at idle can draw in fresh air from what I have been told making it look leaner than it is.

I need to sleep on the issue again as to which way it confuses the sensor and computer (lean or rich).

Keep in mind the computer does not measure unburned fuel, it assumes complete combustion, and uses the amount of O2 left (versus a normally constant outside O2 concentration) as a surrogate test for rich or lean. Rich means it ran out of O2 and still had fuel to burn.

Right now I am lost as to what the exact issue is or might be with yours. Been too long a day and week already.

Are you sure it is not a wiring/connector/ or O2 heater issue?

 

[TheSloMoShow]

Are you sure it is not a wiring/connector/ or O2 heater issue?

I was nearly positive myself after the last time I tested for those issues. I have some time tmrw. So that's when I'm pulling the injectors/checking exhaust system for leaks.

I'll revisit the O2's again as well.


Sent from my SM-G900V using Tapatalk

 

[TheSloMoShow]

I pulled the injector assy tonight. Was able to find as well with all of that out of the way that two of the exhaust manifold bolts weren't even finger tight (one between cylinder 3 and 4, exhaust manifold only bolt and other exhaust manifold only bolt, by cylinder 6). Tightened everything back up after a thorough visual inspection and then re-installed 6 new bosch injectors.

After install I reset KAM and then took her for a ride. LTFT's have gotten better, but still at -7%. Something at idle still isn't right.

So I believe I can rule out any issues with the injector assy, and any exhaust leak at the manifold/head/Gasket area.

I'll have to check for hairline cracks in the exhaust system this weekend, including where the two pre cats merge together. Didn't get a chance to test for exhaust leak there tonight.

O2's are still on the list for the weekend too, either re-test and verify or just flat out replace all 4.

Sent from my SM-G900V using Tapatalk

 

[Ecomike]

Only the front two O2 sensors are used by the computer to control the IAC air flow at idle and the fuel rate from the injectors. The two O2 sensors after the final Cat converters are just EPA verification devices that the engine and Cats are performing as designed to not emit excess HCs and NOx gasses by measuring for too little or too much O2, a surrogate test method for NOx and HCs and for a working catalyst in the Cats.

I pulled the injector assy tonight. Was able to find as well with all of that out of the way that two of the exhaust manifold bolts weren't even finger tight (one between cylinder 3 and 4, exhaust manifold only bolt and other exhaust manifold only bolt, by cylinder 6). Tightened everything back up after a thorough visual inspection and then re-installed 6 new bosch injectors.

After install I reset KAM and then took her for a ride. LTFT's have gotten better, but still at -7%. Something at idle still isn't right.

So I believe I can rule out any issues with the injector assy, and any exhaust leak at the manifold/head/Gasket area.

I'll have to check for hairline cracks in the exhaust system this weekend, including where the two pre cats merge together. Didn't get a chance to test for exhaust leak there tonight.

O2's are still on the list for the weekend too, either re-test and verify or just flat out replace all 4.

Sent from my SM-G900V using Tapatalk

 

[Ecomike]

So what is a normal LTFT for them?

Have you seen any change in the over heat issue at all. Hot exhaust blowing on the intake and engine could have been part of the issue too.

 

[TheSloMoShow]

...The two O2 sensors after the final Cat converters are just EPA verification devices that the engine and Cats are performing as designed to not emit excess HCs and NOx gasses by measuring for too little or too much O2...

I heard somewhere that the two O2 sensors post-cat do actually play a role in Fuel trims, by adjusting Fuel in order to get the pre-cat O2's readings to stabilize and the post-cat readings within normal limits as well. Rather than functioning as a go/no-go measurement and triggering CEL I assume, should the threshold of excess HCs and NOx be reached. Any validity to this?

I cleaned all sensors today to deal with the speed density system, and probed the engine with Carb cleaner. No intake leaks afaik.

The LTFT'S fluctuate between -7 and -11%, but when looking at combination of STFT and LTFT together the output is always around -7%.

I'd think this might be OK, considering the injectors I swapped in have a higher flow rating than stock. But the coolant temperature rises just as it always has, so I still think something is irregular.

I still need a sure fire way to cofirm there are no exhaust leaks, likely I'll use the SeaFoam trick this weekend.

If that doesn't result in any indication of exhaust leak (s), then I think I'll replace the two pre-cat O2 sensors.

Sent from my SM-G900V using Tapatalk

 

[McQue]

FYI - here is a cut and paste from my 01 FSM...

Downstream Sensors:​

Two downstream sensors
are used (1/2 and 2/2). The downstream sensors are
located in the exhaust downpipes just after the minicatalytic
convertors. The downstream is also used to
determine the correct air fuel ratio. As the oxygen
content changes at the downstream the PCM calculates
how much air fuel ratio change is required. The
PCM then looks at the upstream oxygen sensor voltage
and changes fuel delivery until the upstream
sensor voltage changes enough to correct the downstream
sensor voltage (oxygen content).
The downstream oxygen sensors also provide an​

input to determine mini-catalyst efficiency.

My point in posting this is that some people will tell you that the downstream sensors are only there to check if the cat is working... but I tend to believe what is written in the Factory Service Manual. PCM technology improves all the time. Believe it.

 

[Ecomike]

There are errors in that text, in that they refer to the first 2 inline sensors as downstream one minute and then call them upstream in the later description. The first 2 sensors, let's call them "upstream" between the large cats and the mini cats, they are the ones the computer uses for fuel trim adjustment. The down stream sensors, located after the big cats, down stream of both cats is used to measure Cat efficiency, as I said before. From that FSM detail you just posted, it seems that the newer OBD-II rigs do have the computer using the last two O2 sensors not just for EPA clean air testing, but to fine tune the fuel rate if the engine is too rich or two lean based on the final O2 sensor data to get the final exhaust after the large cats into EPA complaince, and perhaps to confirm that the first 2 sensors are working, using a more complex algorithm than older rigs. The O2 concentration does change from the inlet to the large cats to the exit of the large cats. So testing the final O2 sensors should show them running leaner than the first two.

Thanks for sharing that FSM data.

FYI - here is a cut and paste from my 01 FSM...

Downstream Sensors:​

Two downstream sensors
are used (1/2 and 2/2). The downstream sensors are
located in the exhaust downpipes just after the minicatalytic
convertors. The downstream is also used to
determine the correct air fuel ratio. As the oxygen
content changes at the downstream the PCM calculates
how much air fuel ratio change is required. The
PCM then looks at the upstream oxygen sensor voltage
and changes fuel delivery until the upstream
sensor voltage changes enough to correct the downstream
sensor voltage (oxygen content).
The downstream oxygen sensors also provide an​

input to determine mini-catalyst efficiency.

My point in posting this is that some people will tell you that the downstream sensors are only there to check if the cat is working... but I tend to believe what is written in the Factory Service Manual. PCM technology improves all the time. Believe it.

 

[TheSloMoShow]

Threw some SeaFoam in this evening. Engine needed it anyways I'm sure.

Result was this, an exhaust leak after the pre-cat's.

https://vimeo.com/179530777

Small pin-hole leak, but it's there. Unfortunately the dip-$h!t exhaust shop who welded the pipe back together didn't get it completely done.

I could go back and have them seal it up, but is it worth it?

I'm still struggling to see how a leak like this could fool the O2 sensor readings, especially when my O2 sensors voltage readings indicate rich condition but I would think an exhaust leak fools O2's into thinking the mixture is lean.

Sent from my SM-G900V using Tapatalk

 

[McQue]

There are errors in that text, in that they refer to the first 2 inline sensors as downstream one minute and then call them upstream in the later description. The first 2 sensors, let's call them "upstream" between the large cats and the mini cats, they are the ones the computer uses for fuel trim adjustment. The down stream sensors, located after the big cats, down stream of both cats is used to measure Cat efficiency, as I said before. From that FSM detail you just posted, it seems that the newer OBD-II rigs do have the computer using the last two O2 sensors not just for EPA clean air testing, but to fine tune the fuel rate if the engine is too rich or two lean based on the final O2 sensor data to get the final exhaust after the large cats into EPA complaince, and perhaps to confirm that the first 2 sensors are working, using a more complex algorithm than older rigs. The O2 concentration does change from the inlet to the large cats to the exit of the large cats. So testing the final O2 sensors should show them running leaner than the first two

Mike your are confused...there are no sensors after the big cat on a 2001. Read it again. it refers to the upstream sensors, but does not call the downstream sensors , upstream or vice versa. It explains how the pcm uses both up and down to adjust fuel ratio.

 

[Ecomike]

Mike your are confused...there are no sensors after the big cat on a 2001. Read it again. it refers to the upstream sensors, but does not call the downstream sensors , upstream or vice versa. It explains how the pcm uses both up and down to adjust fuel ratio.

Are you saying there are 2 sensors before the pre cats, between the head and the pre cats? And then 2 sensors after the pre cats but before the larger final cats? If so that is bizarre. Found a photo, you are right.

Never seen Detroit do this before. But then again the newest rig I work on is a 2001 Saturn. In this case I would look for leaks at the 2 donuts also and for bad precats in this case. I am still shocked the cooling problem was not the fan clutch, dirty radiator /condenser (inside or out), or T-stat or water pump, or radiator cap.