[cruiser enthuiser]

Hey Fellas. New to this forum. Hoping some of you guys with TBI experience can help me out.

I recently converted my early SBC 350 to TBI. I've worked through a few issues with rough idle, vacuum leaks, etc. It runs pretty well now. However, when I went to smog test (required in Nevada on my '68) my CO emissions at idle were very high and I failed. Interestingly, I am getting a 44 error code for "lean exhaust" . Apparently the computer thinks it's running lean and is enriching my fuel mix?? I've changed the o2 sensor with no change. I've also changed coolant temp sensor, checked for a stuck EGR, checked for vacuum leaks. Not sure what else it might be. Anyone got ideas?

Thanks in advance for your help.

Ed

 

[tnsmith10]

does the throttle body have an iac? are you running a mass air flow sensor? either 1 of those things could be the culprit. also the thought of a faulty knock sensor comes to mind as well.

 

[T-bucket23]

Hey Fellas. New to this forum. Hoping some of you guys with TBI experience can help me out.

I recently converted my early SBC 350 to TBI. I've worked through a few issues with rough idle, vacuum leaks, etc. It runs pretty well now. However, when I went to smog test (required in Nevada on my '68) my CO emissions at idle were very high and I failed. Interestingly, I am getting a 44 error code for "lean exhaust" . Apparently the computer thinks it's running lean and is enriching my fuel mix?? I've changed the o2 sensor with no change. I've also changed coolant temp sensor, checked for a stuck EGR, checked for vacuum leaks. Not sure what else it might be. Anyone got ideas?

Thanks in advance for your help.

Ed

I would be willing to bet your O2 sensor is not working or is not connected properly

 

[BogiesAnnex1]

Hey Fellas. New to this forum. Hoping some of you guys with TBI experience can help me out.

I recently converted my early SBC 350 to TBI. I've worked through a few issues with rough idle, vacuum leaks, etc. It runs pretty well now. However, when I went to smog test (required in Nevada on my '68) my CO emissions at idle were very high and I failed. Interestingly, I am getting a 44 error code for "lean exhaust" . Apparently the computer thinks it's running lean and is enriching my fuel mix?? I've changed the o2 sensor with no change. I've also changed coolant temp sensor, checked for a stuck EGR, checked for vacuum leaks. Not sure what else it might be. Anyone got ideas?

Thanks in advance for your help.

Ed

What exactly is the engine configuration, and what or who's TBI are you using?

OEM TBI is designed around a very specific engine configuration, 305 or 350 or 454 (each of the engines has a chip specifiic to it), with a very mild cam of 160-170 degrees duration an LSA of 112-116 degrees and less than .4 inch lift at the valve and compression around 8 to 1. Anything other than this will quickly get into custom chip territory. Same goes for headers instead on manifolds, this changes the back pressure which will affect the mixture by way of the computer and O2 sensor. Also the temp of the O2 sensor and engine coolant are both critical. Sensors come unheated (2 wire) and heated (3 wire); heated is better as it isn't sensitive to where it's placed from the head pipe. Two wire needs to be jammed as close to the manifold as you can get it or it won't stay hot enough to signal the computer to go into closed loop operation. The engine, also, needs to be at least 180 degrees to signal the computer to come out of cold start choke enrichment mode as well.

The system is dependent upon a catalytic converter and in some instances an AIR injection system as well. Most emissions as a percent of total exhaust output are at or just off idle, this is what they check for; not how many absolute tons of the stuff comes out the pipe. So a cat converter and possibly AIR do a lot to clean these emissions. However, AIR on an engine with a computer not originally designed for it will fake the O2 sensor into thinking the mixture is lean and it will try to correct with added fuel. The same thing will happen if there is an exhaust leak between the engine and the O2 sensor.

TBI is also very sensitive to fuel pressure, you need to know what it is, 10-12 pounds is plenty.

What did you do for a distributor? The OEM HEI is computer controlled. HEI is the minimum ignition that will work. Misfires on idle from old fashion ignitions will make the mixture hugely rich. While the computer if it's in closed loop would try to lean it thru the O2 sensor, its authority to correct the mixture is limited to about 10%, if that, so anything beyond that the computer can't fix given one misfiring cylinder makes the mixture at least 12.5% rich at the O2 sensor.

Another thing is idle speed and manifold heat, the TBI intake needs a lot of exhaust heat to vaporize the liquid fuel and the idle speed needs to be over 600 RPM to help mix the fuel and air. If these conditions aren't being met, it's likely one or more cylinders are see a stream of liquid fuel that can't be consumed given the amount of air present.

So you've got a lot of stuff to check out.

Bogie

 

[fast68]

tbi commonly have bad o2 sensors
try a heated one from a '90s 3800 engine etc.




IAC motors get carboned up bad on these as well

fuel pumps get weak in these tbi vehicles as well commonly

if it take any extra bit of cranking to get it to start up then i would check the fuel pump or replace it they are cheap for the tbi vehicles


make sure the injectors arent super old or in need of cleaning as well.

new filter ?

tbi is VERY picky about fuel pressure

 

[T-bucket23]

What exactly is the engine configuration, and what or who's TBI are you using?

OEM TBI is designed around a very specific engine configuration, 305 or 350 or 454 (each of the engines has a chip specifiic to it), with a very mild cam of 160-170 degrees duration an LSA of 112-116 degrees and less than .4 inch lift at the valve and compression around 8 to 1. Anything other than this will quickly get into custom chip territory. Same goes for headers instead on manifolds, this changes the back pressure which will affect the mixture by way of the computer and O2 sensor. Also the temp of the O2 sensor and engine coolant are both critical. Sensors come unheated (2 wire) and heated (3 wire); heated is better as it isn't sensitive to where it's placed from the head pipe. Two wire needs to be jammed as close to the manifold as you can get it or it won't stay hot enough to signal the computer to go into closed loop operation. The engine, also, needs to be at least 180 degrees to signal the computer to come out of cold start choke enrichment mode as well.

The system is dependent upon a catalytic converter and in some instances an AIR injection system as well. Most emissions as a percent of total exhaust output are at or just off idle, this is what they check for; not how many absolute tons of the stuff comes out the pipe. So a cat converter and possibly AIR do a lot to clean these emissions. However, AIR on an engine with a computer not originally designed for it will fake the O2 sensor into thinking the mixture is lean and it will try to correct with added fuel. The same thing will happen if there is an exhaust leak between the engine and the O2 sensor.

TBI is also very sensitive to fuel pressure, you need to know what it is, 10-12 pounds is plenty.

What did you do for a distributor? The OEM HEI is computer controlled. HEI is the minimum ignition that will work. Misfires on idle from old fashion ignitions will make the mixture hugely rich. While the computer if it's in closed loop would try to lean it thru the O2 sensor, its authority to correct the mixture is limited to about 10%, if that, so anything beyond that the computer can't fix given one misfiring cylinder makes the mixture at least 12.5% rich at the O2 sensor.

Another thing is idle speed and manifold heat, the TBI intake needs a lot of exhaust heat to vaporize the liquid fuel and the idle speed needs to be over 600 RPM to help mix the fuel and air. If these conditions aren't being met, it's likely one or more cylinders are see a stream of liquid fuel that can't be consumed given the amount of air present.

So you've got a lot of stuff to check out.

Bogie

High CO would not be due to unburnt fuel high HC would. I dont think it is a misfire issue. The first thing to check is what the computer sees coming from the O2 sensor . The conflict is that the O2 is reporting lean but the test fixture is reporting rich. This tells me the O2 sensor is not reporting to the computer correctly or the computer is interpreting the signal wrong. To much fuel pressure would not cause the O2 to report lean. It would be reporting rich as it would be trying to have the ECM lean out the mixture unsuccessfully. The O2 needs to be scoped to see what it is actually doing and the signal pattern at the computer needs to be looked at with a scan tool to see whet the computer is seeing. A 0 voltage at the ECM would tell the system it is lean and the fuel pulse would be lenghtened to try to compensate. Once it reaches its limit it will set the lean code. This could be due to a shorted wire from the O2. Probably not an open as most times if it is open it will float to .5v as a baseline.

 

[cruiser enthuiser]

Thanks for the responses everyone. I'll try to answer as many questions as I can.

The vehicle is a ' 68 Land Cruiser. The engine is a crate 350 put in by the previous owner. It's a Mr. Goodwrench job, bone stock, nothing special about it in anyway. I wish I could speak in greater detail to its particulars. I changed the intake/throttle body with one from a 5.7L TBI out of an '88 suburban and am running the OE distributor (with new cap/rotor). I'm using the well known 7747 ECM. During the conversion I changed most sensors from the 5.7 with new including the IAC. I upgraded to a 3 wire heated sensor and I am running rams horns. The oxygen sensor is about 12 inches downstream of the these on one side. I'm running dual exhaust, no crossover, no cats. There's no MAF or air rails. I've installed an aftermarket fuel pressure regulator on the TBI and was running it at 13 psi. I've since dialed it back to 12 psi.

I am using WINALDL to monitor performance and the computer is definetely getting a signal from the o2 sensor. I'm not sure what a correct reading (in volts I believe?) should be, tho. I don't have any data in front of me now, but the numbers bounce around a fair deal during operation which I thought was normal.

Could this problem be caused by an exhaust leak upstream of the o2 sensor? I've picked up a new set of collector gaskets which I'll be throwing on this weekend as the ones there look suspect. I don't think the manifolds are leaking, but maybe I should re-gasket, re-install these?

Thanks for your help guys,
Ed

 

[cruiser enthuiser]

does the throttle body have an iac? are you running a mass air flow sensor? either 1 of those things could be the culprit. also the thought of a faulty knock sensor comes to mind as well.

The knock sensor is I think the only sensor I re-used. Not sure how that would factor in to this problem tho? If somone explained to me how it could be the cuplrit I'd throw a new one on for sure.

 

[T-bucket23]

Thanks for the responses everyone. I'll try to answer as many questions as I can.

The vehicle is a ' 68 Land Cruiser. The engine is a crate 350 put in by the previous owner. It's a Mr. Goodwrench job, bone stock, nothing special about it in anyway. I wish I could speak in greater detail to its particulars. I changed the intake/throttle body with one from a 5.7L TBI out of an '88 suburban and am running the OE distributor (with new cap/rotor). I'm using the well known 7747 ECM. During the conversion I changed most sensors from the 5.7 with new including the IAC. I upgraded to a 3 wire heated sensor and I am running rams horns. The oxygen sensor is about 12 inches downstream of the these on one side. I'm running dual exhaust, no crossover, no cats. There's no MAF or air rails. I've installed an aftermarket fuel pressure regulator on the TBI and was running it at 13 psi. I've since dialed it back to 12 psi.

I am using WINALDL to monitor performance and the computer is definetely getting a signal from the o2 sensor. I'm not sure what a correct reading (in volts I believe?) should be, tho. I don't have any data in front of me now, but the numbers bounce around a fair deal during operation which I thought was normal.

Could this problem be caused by an exhaust leak upstream of the o2 sensor? I've picked up a new set of collector gaskets which I'll be throwing on this weekend as the ones there look suspect. I don't think the manifolds are leaking, but maybe I should re-gasket, re-install these?

Thanks for your help guys,
Ed

The O2 signal will constantly be fluctuating between .1 and .9 volts once warm.
An exhaust leak in front of the O2 can cause issues as it can introduce extra air into the system.

The other thing to look at is the MAP sensor signal. Make sure it is within spec.

 

[BogiesAnnex1]

High CO would not be due to unburnt fuel high HC would. I dont think it is a misfire issue. The first thing to check is what the computer sees coming from the O2 sensor . The conflict is that the O2 is reporting lean but the test fixture is reporting rich. This tells me the O2 sensor is not reporting to the computer correctly or the computer is interpreting the signal wrong. To much fuel pressure would not cause the O2 to report lean. It would be reporting rich as it would be trying to have the ECM lean out the mixture unsuccessfully. The O2 needs to be scoped to see what it is actually doing and the signal pattern at the computer needs to be looked at with a scan tool to see whet the computer is seeing. A 0 voltage at the ECM would tell the system it is lean and the fuel pulse would be lenghtened to try to compensate. Once it reaches its limit it will set the lean code. This could be due to a shorted wire from the O2. Probably not an open as most times if it is open it will float to .5v as a baseline.

High CO and high HC often go together, they are an indicator that the chemical reaction is not going to completion, it is running out of air before finishing the burn. This is an indicator of a rich mixture which has too much fuel for the available oxygen supply. It can also be an indicator of late timing either or both cam and ignition where the initial events of valve timing and or ignition are late relative to the piston cycle and the reaction does not go to completion due to insufficient compression causing a slow burn and/or the exhaust valve coming open before the fire burnt thru the reactants.

A simple misfire would read rich by the O2 sensor causing it to signal the computer to lean out the mixture. If the computer was successful in leaning out the mixture there would be 7 cylinders running lean and one spewing raw fuel. On a smog test this would tend to be high in HC and CO from the misfire cylinders and high in NOx and CO2 from the lean cylinders.

Excessive fuel pressure or a leaking injector would put excess fuel into the engine, the computer in closed loop would try to fix this the same as it would a misfire. In this case there wouldn't necessarily be 7 lean cylinders and one rich, but where I was going is that the computer may not have enough authority to totally correct the mixture which would end up again in the situation where there was not sufficient air (oxygen) to complete the reaction. If the richness is slight which my example of 10% authority against a 12.5 percent rich mix could end up with high CO but not a lot of HCs because there would be just enough oxygen to take the reaction mostly but not completely thru the first stage of the carbon burn which would yield CO but lacking more oxygen cannot then burn the CO to CO2. If the mixture was more rich then there would be a woeful lack of oxygen and some of the HCs couldn't burn at all and would go out the pipe with both high CO and high HC content. So he could be dealing with a very subtle problem between a little rich and a lot rich.

If he had a air leak into the exhaust ahead of the O2 sensor, the sensor would read the mixture as lean and throw fuel at a false problem. This in reality would richen the mixture which the O2 sensor would in combination with the air leak read as normal, or as normal as it can make it, and your back into the previous paragraphs problem statement of a rich mixture. The issue is how rich, a little rich will show a high CO but may not indicate excess HC. A lot rich will show a high CO and a high HC.

The problem he has in diagnosis is there's a computer in there trying to fix things, unlike a carb that just soldiers on without knowledge of nor an ability to correct problems. So he has to deal with the identified problem not only from a failure standpoint as a possible cause but also how the computer tries to correct problems which may in the curative process cause other problems.

The simple solution is probably to put a cat converter on both pipes just down stream of the O2 sensor.

Bogie

 

[carsavvycook]

What are the HC's, %O2, and NOX readings? Are they at, or below normal?

Just a high CO is caused by retarded ignition timing, with all of the other test numbers normal.

This could also be effected by the spark plug gap, and heat range, but not as much, as the retarded timing.

Your lean O2 sensor code is most likely from the exhaust leak prior to the O2 sensor.

 

[tnsmith10]

The knock sensor is I think the only sensor I re-used. Not sure how that would factor in to this problem tho? If somone explained to me how it could be the cuplrit I'd throw a new one on for sure.

on a knock sensor, it senses detonation which is usually created from either too much ignition timing or the lack of fuel.
if the sensor is faulty, it can give eroneous readings back to the ecm, which if it is sensing knock when there isnt, it can give off a signal back to the ecm saying the engine is lean.
if you have done any kind of engine modifications that involve anything to do with what old bogie mentioned, then i would follow his lead. if its stock, then you are going to have to start testing parts. dont just throw parts at the car, you could just be throwing money away.
another thought is from what was mentioned about the distibutor, if its a computer controlled distributor, it also is sending info back to the ecm and the ecm makes the proper adjustments accordingly if its functioning correctly, and when dropped in is supposed to be set @ tdc, the ecm makes all the adjustments based on readings from O2 sensors, the mass air flow sensor,temp sensors, the knock sensor, iac motor, map sensor, and probably a couple other things im not able to think of at this time. also, vacuum leaks make for lean problems as well. gl

 

[BogiesAnnex1]

The knock sensor is I think the only sensor I re-used. Not sure how that would factor in to this problem tho? If somone explained to me how it could be the cuplrit I'd throw a new one on for sure.

If the knock sensor activates the computer retards the spark, this might not provide enough burn time which would push a high CO and probably a high HC content in the exhaust.

Bogie

 

[cruiser enthuiser]

Wow, a lot of great information...thanks guys! I definetely have a better idea of which way to go in resolving this. Carsavvycook, I may post up the emissions test details later. Gonna be wrenching this weekend and hopefully resolve this soon.

Ed

 

[sbchevfreak]

On an OBDI system the PCM will seea missfire as a false lean, due to the added unused oxygen coming from the misfiring cylinder, and actualy richen the fuel mix, causing high CO an HC at the sniffer. If here is a misfire, it actually will show up extremely rich a the tailpipe.. An upstream exaughst leak can cause the same effect.

 

[carsavvycook]

We will have a better idea, as to what is going on, once he posts all of the test results.

I am putting my money on retarded ignition timing until then.