O2 sensor failing and indicating lean on heavy throttle
11-27-2016, 08:30 PM
#1
O2 sensor failing and indicating lean on heavy throttle
I have a 85 with an engine that I rebuilt. New long tube headers with the O2 sensor after the collector. On the right side same location I have a wide band O2 sensor. I am datalogging and can see the ECU output in real time on my laptop. The narrow band O2 sensor is a 4 wire Bosch heated type. No smog pump, and EGR is blocked off and turned off on the chip.
When I start the car it goes into closed loop and I can see the O2 sensor cycling, and crossing over like you would expect. Wide band confirms the AFR is reasonable, and BLM and INT are around 128. I can drive around and everything seems fine. As soon as I start going WOT, and heavy acceleration, the O2 voltage goes low, starts indicating lean, and stops cycling. The INT goes rich to compensate and starts driving the BLM up, and the wide band starts moving progressively rich, dropping below 10.
The O2 voltage stays low, and normal driving does not cause it to go back to crossing over after that. I have swapped the O2 sensors side to side in the past to make sure that the readings aren't being affected by an air leak on one side. I can turn the car off and after a short time the O2 sensor will go back to working on startup.
Is this just a faulty O2 sensor? It's brand new. I don't understand why it works until I use heavy throttle. Heat? Something in my exhaust temporarily fouling it? I don't have junk coming out of my exhaust. I am trying to find something other than the o2 sensor as the problem. I'm not experienced, and this doesn't make sense to me.
When I start the car it goes into closed loop and I can see the O2 sensor cycling, and crossing over like you would expect. Wide band confirms the AFR is reasonable, and BLM and INT are around 128. I can drive around and everything seems fine. As soon as I start going WOT, and heavy acceleration, the O2 voltage goes low, starts indicating lean, and stops cycling. The INT goes rich to compensate and starts driving the BLM up, and the wide band starts moving progressively rich, dropping below 10.
The O2 voltage stays low, and normal driving does not cause it to go back to crossing over after that. I have swapped the O2 sensors side to side in the past to make sure that the readings aren't being affected by an air leak on one side. I can turn the car off and after a short time the O2 sensor will go back to working on startup.
Is this just a faulty O2 sensor? It's brand new. I don't understand why it works until I use heavy throttle. Heat? Something in my exhaust temporarily fouling it? I don't have junk coming out of my exhaust. I am trying to find something other than the o2 sensor as the problem. I'm not experienced, and this doesn't make sense to me.
11-28-2016, 01:59 AM
#3
Pro
Yes, when the "PE" mode is enabled it returns to OL in correlation with the inputs.
Does your wideband show lean when the narrowband voltage drops? Does your wideband have a narrowband output you can use?
Does your wideband show lean when the narrowband voltage drops? Does your wideband have a narrowband output you can use?
Last edited by wikdwizard; 11-28-2016 at 02:01 AM.
11-28-2016, 06:12 AM
#4
Melting Slicks
sounds like a bad narrowband. Disable the narrowband and witness what the wideband says during WOT, and cruise. The fact it stops responding is also a clue that it might be bad.
An engine can be tuned such that in open loop, the a/f is reasonably close to desired. I always run my cars in open loop, never use a narrowband. I hold it between 14.8 and 15.2 during idle/cruise, and I dip into the 11.5's for WOT because I also see 15psi of boost and 9:1 compression with 93 octane.
An engine can be tuned such that in open loop, the a/f is reasonably close to desired. I always run my cars in open loop, never use a narrowband. I hold it between 14.8 and 15.2 during idle/cruise, and I dip into the 11.5's for WOT because I also see 15psi of boost and 9:1 compression with 93 octane.
Last edited by Kingtal0n; 11-28-2016 at 06:13 AM.
11-28-2016, 08:43 AM
#5
Yes the ECU should boost fuel (PE) on heavy throttle. I am mostly concerned that the narrow band seems to stop working after heavy throttle use, and doesn't recover, so that the BLM goes high and eventually it triggers the lean code.
The wide band will move lean momentarily on throttle application, then it goes to normal, and then on lift throttle goes rich momentarily then back normal. This is normal behavior. I am watching the wide band on a dash gauge, so I can verify that everything seems fine until the narrow band starts giving the bad readings.
I am thinking its a bad narrowband also, but it's easy to misdiagnose this stuff and spend money you don't need to.
I would prefer to keep the narrow band. My engine is not that modified, your turbo probably needs more attention to fueling than normal. Thanks.
The wide band will move lean momentarily on throttle application, then it goes to normal, and then on lift throttle goes rich momentarily then back normal. This is normal behavior. I am watching the wide band on a dash gauge, so I can verify that everything seems fine until the narrow band starts giving the bad readings.
I am thinking its a bad narrowband also, but it's easy to misdiagnose this stuff and spend money you don't need to.
I would prefer to keep the narrow band. My engine is not that modified, your turbo probably needs more attention to fueling than normal. Thanks.
11-29-2016, 12:08 PM
#8
Le Mans Master
Member Since: Oct 2002
Location: Las Vegas - Just stop perpetuating myths please.
Posts: 7,098
Received 373 Likes
on
356 Posts
I'm guessing a exhaust leak on that "lean" side. 1 of the header tubes is drawing in air as the RPM's increase. That's how good headers work vs cast manifolds they provide some amount of extraction of the exh gas from all the flow at the collector.
I'm also guessing that O2 is getting hotter as the RPM increases so not a unheated sensor problem as the opposite side seems to function OK.
From what I read is you have swapped O2's and that same side is still lean at WOT. I don't know what you mean by bad narrow band as swapping O2's shows good sensors and the instrument that reads it and controls it is the ECU. You have to remember the NB is controlling the mix and is much more sensitive to small voltage & stoichiometric change while you're WB is read only and has small changes in output for the same change in A/F. But in my mind the ECU should move to open loop with just heavy throttle let alone WOT and fueling should be from off the "tables" - not the O2's. What do your ECM tables look like do you know?
Why the O2 reading doesn't return to closed loop is another issue. What is the WB saying at the same time? The ECU needs certain criteria to return to closed loop. Do you meet them? You need to look these up and verify them for your car problem.
I'm also guessing that O2 is getting hotter as the RPM increases so not a unheated sensor problem as the opposite side seems to function OK.
From what I read is you have swapped O2's and that same side is still lean at WOT. I don't know what you mean by bad narrow band as swapping O2's shows good sensors and the instrument that reads it and controls it is the ECU. You have to remember the NB is controlling the mix and is much more sensitive to small voltage & stoichiometric change while you're WB is read only and has small changes in output for the same change in A/F. But in my mind the ECU should move to open loop with just heavy throttle let alone WOT and fueling should be from off the "tables" - not the O2's. What do your ECM tables look like do you know?
Why the O2 reading doesn't return to closed loop is another issue. What is the WB saying at the same time? The ECU needs certain criteria to return to closed loop. Do you meet them? You need to look these up and verify them for your car problem.
Last edited by cardo0; 11-29-2016 at 12:17 PM. Reason: Add no return to closed loop.
11-29-2016, 01:48 PM
#9
Pro
Did the OBD1 cars have two oxygen sensors? I was assuming he meant he swapped the wideband and the factory sensor side to side in an attempt to eliminate variables.
11-29-2016, 01:57 PM
#10
Melting Slicks
I'm guessing a exhaust leak on that "lean" side. 1 of the header tubes is drawing in air as the RPM's increase. That's how good headers work vs cast manifolds they provide some amount of extraction of the exh gas from all the flow at the collector.
I'm also guessing that O2 is getting hotter as the RPM increases so not a unheated sensor problem as the opposite side seems to function OK.
From what I read is you have swapped O2's and that same side is still lean at WOT. I don't know what you mean by bad narrow band as swapping O2's shows good sensors and the instrument that reads it and controls it is the ECU. You have to remember the NB is controlling the mix and is much more sensitive to small voltage & stoichiometric change while you're WB is read only and has small changes in output for the same change in A/F. But in my mind the ECU should move to open loop with just heavy throttle let alone WOT and fueling should be from off the "tables" - not the O2's. What do your ECM tables look like do you know?
Why the O2 reading doesn't return to closed loop is another issue. What is the WB saying at the same time? The ECU needs certain criteria to return to closed loop. Do you meet them? You need to look these up and verify them for your car problem.
I'm also guessing that O2 is getting hotter as the RPM increases so not a unheated sensor problem as the opposite side seems to function OK.
From what I read is you have swapped O2's and that same side is still lean at WOT. I don't know what you mean by bad narrow band as swapping O2's shows good sensors and the instrument that reads it and controls it is the ECU. You have to remember the NB is controlling the mix and is much more sensitive to small voltage & stoichiometric change while you're WB is read only and has small changes in output for the same change in A/F. But in my mind the ECU should move to open loop with just heavy throttle let alone WOT and fueling should be from off the "tables" - not the O2's. What do your ECM tables look like do you know?
Why the O2 reading doesn't return to closed loop is another issue. What is the WB saying at the same time? The ECU needs certain criteria to return to closed loop. Do you meet them? You need to look these up and verify them for your car problem.
11-29-2016, 07:35 PM
#11
I am using a 4 wire Bosch 15730 heated Oxygen Sensor, Universal Type Fitment. It was $35 shipped, no big deal if I have to replace it.
Yes, I have two sensor bungs just past the collectors, one on each side.
I had done this before, but just swapped the wide and narrow band sensors left and right, and drove for 12 minutes. Light throttle, low speed. Wide band was showing spot on AFR. Car went into closed loop normally, and I could see the narrow band cycling high and low, and my wide band kept showing good AFR. At 10 minutes the narrow band went to low voltage, stopped cycling, and the ECU started adding fuel. Wide band dropped to 10-12 AFR at that point.
So it seems that my narrow band is defective. Does anyone know if I am using the wrong type of narrow band? Like one expressly made to go way downstream after cats or something like that? I think it very odd that it would work briefly, then stop working. And then start working again after a cool down. I don't know anything about the details of O2 sensors.
My ECU is going back to closed loop. The problem is that when it does, the bad narrowband forces the AFR rich.
My AEM (30-4110) UEGO Air/Fuel Ratio Gauge works great though. I got that for $165 shipped online. I don't know how people can tune their cars without that and a tool like the Moates APU1 I bought.
Yes, I have two sensor bungs just past the collectors, one on each side.
I had done this before, but just swapped the wide and narrow band sensors left and right, and drove for 12 minutes. Light throttle, low speed. Wide band was showing spot on AFR. Car went into closed loop normally, and I could see the narrow band cycling high and low, and my wide band kept showing good AFR. At 10 minutes the narrow band went to low voltage, stopped cycling, and the ECU started adding fuel. Wide band dropped to 10-12 AFR at that point.
So it seems that my narrow band is defective. Does anyone know if I am using the wrong type of narrow band? Like one expressly made to go way downstream after cats or something like that? I think it very odd that it would work briefly, then stop working. And then start working again after a cool down. I don't know anything about the details of O2 sensors.
My ECU is going back to closed loop. The problem is that when it does, the bad narrowband forces the AFR rich.
My AEM (30-4110) UEGO Air/Fuel Ratio Gauge works great though. I got that for $165 shipped online. I don't know how people can tune their cars without that and a tool like the Moates APU1 I bought.
Last edited by BlueTwoToneCorvette; 11-29-2016 at 07:36 PM.
11-30-2016, 04:17 AM
#12
Melting Slicks
narrowbands are all the same. I walk into pep boys and find the cheapest one I can, often a 1-wire sensor for an old truck or something, for $6 or $8 or whatever, and it will work in place of any narrowband on any vehicle ever made, as they all function essentially the same: by flipping back and forth from 1.001~ volt to 0.001~ volt which alerts the ECU which way is "rich" and which way is "lean".
The other wires are for heating elements (faster warming up, staying warmed up) and redundant/additional grounds (improved sensor accuracy or faster response? or electron return line for sensor life improvement?I never did find out why they sometimes have the redundant grounds...)
The other wires are for heating elements (faster warming up, staying warmed up) and redundant/additional grounds (improved sensor accuracy or faster response? or electron return line for sensor life improvement?I never did find out why they sometimes have the redundant grounds...)
Last edited by Kingtal0n; 11-30-2016 at 04:18 AM.
11-30-2016, 09:14 AM
#14
Well, yes. That’s why I swapped the sensors. So now I know that the narrow band and wide band readings are the same both sides. That couldn’t occur if there is a pre-sensor air leak in the exhaust on one side. And if I had an air leak on both sides, the wide band would report lean. The wide band is telling me my AFR is near perfect until the ECU starts commanding a rich AFR in response to the lean narrowband signal.
11-30-2016, 02:52 PM
#15
Melting Slicks
I think you are assuming it is the sensor and ignoring it might be a different problem the sensor is reporting and can't fix. Not claiming to know what it is only something dumping fuel after some amount of time.
11-30-2016, 03:13 PM
#16
Yes you're right, I need to try to isolate the current narrowband o2 sensor to see if there is another problem. I am going to set the wide band to emulate narrow band, and wire it to the ECU narrowband o2 input. Thanks.
12-01-2016, 12:23 PM
#17
Melting Slicks
I would watch the sensors on a scan tool or laptop software and see if the change to a wrong but valid number an not set a code. If it has a cold start injector unplug it.
12-01-2016, 01:43 PM
#18
Le Mans Master
Member Since: Oct 2002
Location: Las Vegas - Just stop perpetuating myths please.
Posts: 7,098
Received 373 Likes
on
356 Posts
I am using a 4 wire Bosch 15730 heated Oxygen Sensor, Universal Type Fitment. It was $35 shipped, no big deal if I have to replace it.
Yes, I have two sensor bungs just past the collectors, one on each side.
I had done this before, but just swapped the wide and narrow band sensors left and right, and drove for 12 minutes. Light throttle, low speed. Wide band was showing spot on AFR. Car went into closed loop normally, and I could see the narrow band cycling high and low, and my wide band kept showing good AFR. At 10 minutes the narrow band went to low voltage, stopped cycling, and the ECU started adding fuel. Wide band dropped to 10-12 AFR at that point.
So it seems that my narrow band is defective. Does anyone know if I am using the wrong type of narrow band? Like one expressly made to go way downstream after cats or something like that? I think it very odd that it would work briefly, then stop working. And then start working again after a cool down. I don't know anything about the details of O2 sensors.
My ECU is going back to closed loop. The problem is that when it does, the bad narrowband forces the AFR rich.
My AEM (30-4110) UEGO Air/Fuel Ratio Gauge works great though. I got that for $165 shipped online. I don't know how people can tune their cars without that and a tool like the Moates APU1 I bought.
Yes, I have two sensor bungs just past the collectors, one on each side.
I had done this before, but just swapped the wide and narrow band sensors left and right, and drove for 12 minutes. Light throttle, low speed. Wide band was showing spot on AFR. Car went into closed loop normally, and I could see the narrow band cycling high and low, and my wide band kept showing good AFR. At 10 minutes the narrow band went to low voltage, stopped cycling, and the ECU started adding fuel. Wide band dropped to 10-12 AFR at that point.
So it seems that my narrow band is defective. Does anyone know if I am using the wrong type of narrow band? Like one expressly made to go way downstream after cats or something like that? I think it very odd that it would work briefly, then stop working. And then start working again after a cool down. I don't know anything about the details of O2 sensors.
My ECU is going back to closed loop. The problem is that when it does, the bad narrowband forces the AFR rich.
My AEM (30-4110) UEGO Air/Fuel Ratio Gauge works great though. I got that for $165 shipped online. I don't know how people can tune their cars without that and a tool like the Moates APU1 I bought.
Well, yes. That’s why I swapped the sensors. So now I know that the narrow band and wide band readings are the same both sides. That couldn’t occur if there is a pre-sensor air leak in the exhaust on one side. And if I had an air leak on both sides, the wide band would report lean. The wide band is telling me my AFR is near perfect until the ECU starts commanding a rich AFR in response to the lean narrowband signal.
Well I'm trying to figure your setup and testing. It reads like you have 2 bungs in each (side) collector with a NB in each side but only 1 WB on one side. So when you swap sensors you move the WB to the other side and the NB's of course are swapped. That leaves the WB paired with the same NB yes? What I'm not reading is whether the lean on WOT moves to the other side along with the same NB. I think you are trying to say that it does. Yes? Then it does look to be a NB. But I don't see that you verified the good reading NB with your WB. Or maybe you have but I don't see it. Tracking your good reading NB would help verify your diagnose the bad NB is truly misbehaving or that the WB is not accurate enough to verify this. I say this because WB don't have the resolution near stoichiometric that NB sensors do.
OK you do get back to closed loop. But that one side NB can't control correctly?
Something that can happen to O2 sensors is they become sealed up with sealer or other during installation. O2 sensors have to breath out through the wire terminal end. Often this end gets covered up with some type of goo.
Please let us know what you find.
12-05-2016, 09:40 PM
#19
My 85 had 1 narrowband. When I replaced the exhaust, I added another bung on the right side. That's what I mean when I say I am switching them side to side to verify the readings are the same to check for air leaks.
My engine mods include descreening the MAF, minor porting of the upper and lower plenum, and heads, valve job, 1.6 roller rockers, and long tube headers, with 2.5" dual exhaust.
I set my wideband to the narrowband emulation setting and connected its 0-1v output the o2 line to the ECU. I wasn't able to get it to work properly. I'll have to read more about that and try again.
When you disconnect the O2 signal line the ECU never goes into closed loop, so I played around with changing the MAF tables to trick the ECU into adding more fuel. You will get a O2 related error code, but the car will run fine and you can drive around and test MAF changes.
Just using the stock ECU MAF maps in open loop, my wideband was showing idle as lean 16-17 AFR, low RPM cruise was lean around 15-16 AFR, and WOT was rich, 11-14 AFR. I was easily able to change MAF table 1 and 2 and get the AFR from 14-15 at idle and cruise. I did quite a bit of cruising and WOT while watching the wideband, and the car ran fine. I was able to spin the tires in first and second gear by rolling on the gas, so I think the engine is working ok.
One of my oil pressure sensors had blown out in a geyser of oil previously, and I think the O2 connector was coated in oil. I was careful not to seal the ends of the 4 O2 wires by using a crimp splice, like they caution about "breathing" through the wire, but oil got on there, so that may be causing my problem. I'll have to look at that again.
Overall the fuel system seems to be working well enough except for the flakey O2 narrowband issue I have.
I'm really excited to get to drive my car, even if it's not perfect yet. I've been working on it for a year, rebuilding everything along the way.
My engine mods include descreening the MAF, minor porting of the upper and lower plenum, and heads, valve job, 1.6 roller rockers, and long tube headers, with 2.5" dual exhaust.
I set my wideband to the narrowband emulation setting and connected its 0-1v output the o2 line to the ECU. I wasn't able to get it to work properly. I'll have to read more about that and try again.
When you disconnect the O2 signal line the ECU never goes into closed loop, so I played around with changing the MAF tables to trick the ECU into adding more fuel. You will get a O2 related error code, but the car will run fine and you can drive around and test MAF changes.
Just using the stock ECU MAF maps in open loop, my wideband was showing idle as lean 16-17 AFR, low RPM cruise was lean around 15-16 AFR, and WOT was rich, 11-14 AFR. I was easily able to change MAF table 1 and 2 and get the AFR from 14-15 at idle and cruise. I did quite a bit of cruising and WOT while watching the wideband, and the car ran fine. I was able to spin the tires in first and second gear by rolling on the gas, so I think the engine is working ok.
One of my oil pressure sensors had blown out in a geyser of oil previously, and I think the O2 connector was coated in oil. I was careful not to seal the ends of the 4 O2 wires by using a crimp splice, like they caution about "breathing" through the wire, but oil got on there, so that may be causing my problem. I'll have to look at that again.
Overall the fuel system seems to be working well enough except for the flakey O2 narrowband issue I have.
I'm really excited to get to drive my car, even if it's not perfect yet. I've been working on it for a year, rebuilding everything along the way.