P0171 Code - Driving me crazy! Any ideas?
Discussion
Hoping someone on here can offer an insight to a problem I'm having.
I have a persistent P0171 code on my 2006 Lexus GS430 (Lean on Bank 1) and I cannot seem to get to grips with it. There are no other codes and the engine appears to run great.
However, using my OBD scantool, the fuel trim on Bank 1 is around 25-35%, rising when under load, so really lean, while Bank 2 is usually between 6-10% (fine). Everything I know suggests it's usually fuel, air or sensor issues, but it can't be fuel, as the car is LPG converted and the readings are identical on Petrol or LPG, which use different pumps, injectors, lines and tanks, it also can't be the MAF, or a leak that is too far upstream, as the other engine bank is running fine, so I've focused on checking Bank 1 specific air leaks and O2 Sensors.
What I've done so far:
- Installed new Cat - This was changed a while back, as the old one had died (I suspect this was the first sign of this problem, but the P0171 code only showed up a few weeks later, as I wasn't doing too many miles due to lock down).
- Fully smoke tested the Intake and Exhaust Manifold - Nothing.
- Replaced both Bank 1 O2 Sensors with new OEM ones - No change.
- Repaired three iffy wires on the O2 loom side connector for Bank 1 Sensor 1 - Made no difference to sensor readings or fuel trims.
- Checked timing (using Cam/Crank marks) - Spot on.
- Checked cam timing directly (removed the cam cover and eyeballing the internal marks on the cam) - Spot on.
- Checked compression - Front to back: 155/155/160/165 Psi - very healthy, 142 Psi is the minimum 'good' result.
I now have the LHS cam cover off and the spark plugs were all quite loose and there was oil in the spark plug tubes of two of them.
I am now at a loss - I don't think it's sensors and I can't find any air leaks. All I can do now is put it all back together, replace the spark plug seals and ensure the plugs are torqued correctly, but I can't see any reason why it's not just going to run lean again.
Can anyone see or think of anything I've missed? It's driving me nuts!!
I have a persistent P0171 code on my 2006 Lexus GS430 (Lean on Bank 1) and I cannot seem to get to grips with it. There are no other codes and the engine appears to run great.
However, using my OBD scantool, the fuel trim on Bank 1 is around 25-35%, rising when under load, so really lean, while Bank 2 is usually between 6-10% (fine). Everything I know suggests it's usually fuel, air or sensor issues, but it can't be fuel, as the car is LPG converted and the readings are identical on Petrol or LPG, which use different pumps, injectors, lines and tanks, it also can't be the MAF, or a leak that is too far upstream, as the other engine bank is running fine, so I've focused on checking Bank 1 specific air leaks and O2 Sensors.
What I've done so far:
- Installed new Cat - This was changed a while back, as the old one had died (I suspect this was the first sign of this problem, but the P0171 code only showed up a few weeks later, as I wasn't doing too many miles due to lock down).
- Fully smoke tested the Intake and Exhaust Manifold - Nothing.
- Replaced both Bank 1 O2 Sensors with new OEM ones - No change.
- Repaired three iffy wires on the O2 loom side connector for Bank 1 Sensor 1 - Made no difference to sensor readings or fuel trims.
- Checked timing (using Cam/Crank marks) - Spot on.
- Checked cam timing directly (removed the cam cover and eyeballing the internal marks on the cam) - Spot on.
- Checked compression - Front to back: 155/155/160/165 Psi - very healthy, 142 Psi is the minimum 'good' result.
I now have the LHS cam cover off and the spark plugs were all quite loose and there was oil in the spark plug tubes of two of them.
I am now at a loss - I don't think it's sensors and I can't find any air leaks. All I can do now is put it all back together, replace the spark plug seals and ensure the plugs are torqued correctly, but I can't see any reason why it's not just going to run lean again.
Can anyone see or think of anything I've missed? It's driving me nuts!!
Surely the obvious thing to have tested....is it lean ? And confirm yes or no.
No point wasting time or money throwing parts at it without establishing actual mixtures.
And if it's stating fuel trims are high on that side...that's clearly because it thinks it is lean. SO apply a simulated voltage to that o2 input and see firstly if you can read it, and what response the ecu gives in terms of fueling. Whilst also monitoring mixtures at the same time
No point wasting time or money throwing parts at it without establishing actual mixtures.
And if it's stating fuel trims are high on that side...that's clearly because it thinks it is lean. SO apply a simulated voltage to that o2 input and see firstly if you can read it, and what response the ecu gives in terms of fueling. Whilst also monitoring mixtures at the same time
stevieturbo said:
Surely the obvious thing to have tested....is it lean ? And confirm yes or no.
No point wasting time or money throwing parts at it without establishing actual mixtures.
And if it's stating fuel trims are high on that side...that's clearly because it thinks it is lean. SO apply a simulated voltage to that o2 input and see firstly if you can read it, and what response the ecu gives in terms of fueling. Whilst also monitoring mixtures at the same time
Thanks for the suggestions - I should have said, but the plugs did look like it was running lean, No point wasting time or money throwing parts at it without establishing actual mixtures.
And if it's stating fuel trims are high on that side...that's clearly because it thinks it is lean. SO apply a simulated voltage to that o2 input and see firstly if you can read it, and what response the ecu gives in terms of fueling. Whilst also monitoring mixtures at the same time
However, I'll see about checking as you said, although I'm doing this as a DIY'er so haven't done that before (this is the first time I've had to muck about with O2 sensors). A quick bit of Googling looks like it's simple enough, though, so I'll look at that, thanks.
(Forgot to add - the O2s that were fitted originally were aftermarket and I didn't fully trust them - their resistances on the bench weren't in the range what the workshop manual dictated.)
Edited by Mike_Mac on Saturday 20th March 22:10
Mike_Mac said:
Thanks for the suggestions - I should have said, but the plugs did look like it was running lean,
However, I'll see about checking as you said, although I'm doing this as a DIY'er so haven't done that before (this is the first time I've had to muck about with O2 sensors). A quick bit of Googling looks like it's simple enough, though, so I'll look at that, thanks.
(Forgot to add - the O2s that were fitted were aftermarket and I didn't fully trust them - their resistances on the bench weren't in the range what the workshop manual dictated.)
You've just answered your own question.However, I'll see about checking as you said, although I'm doing this as a DIY'er so haven't done that before (this is the first time I've had to muck about with O2 sensors). A quick bit of Googling looks like it's simple enough, though, so I'll look at that, thanks.
(Forgot to add - the O2s that were fitted were aftermarket and I didn't fully trust them - their resistances on the bench weren't in the range what the workshop manual dictated.)
Edited by Mike_Mac on Saturday 20th March 21:46
Edited by Mike_Mac on Saturday 20th March 21:50
Swallow the cost and fit genuine O2 sensors.
Really this is basic stuff.
If a sensor is telling you it is lean....the ecu is trying to add fuel but it's still reporting lean. YOu need to establish if it is lean. And if that sensor is working correctly.
It's not rocket science. Does the sensor behave as expected when you rev, on the overrun, does it ever indicate anything other than lean ?
If you swap bank to bank what happens ?
And again...measure the mixtures on each bank to determine if it is lean.
You're pissing in the wind without some common sense applied.
If a sensor is telling you it is lean....the ecu is trying to add fuel but it's still reporting lean. YOu need to establish if it is lean. And if that sensor is working correctly.
It's not rocket science. Does the sensor behave as expected when you rev, on the overrun, does it ever indicate anything other than lean ?
If you swap bank to bank what happens ?
And again...measure the mixtures on each bank to determine if it is lean.
You're pissing in the wind without some common sense applied.
stevieturbo said:
Really this is basic stuff.
If a sensor is telling you it is lean....the ecu is trying to add fuel but it's still reporting lean. You need to establish if it is lean. And if that sensor is working correctly.
It's not rocket science. Does the sensor behave as expected when you rev, on the overrun, does it ever indicate anything other than lean ?
If you swap bank to bank what happens ?
And again...measure the mixtures on each bank to determine if it is lean.
You're pissing in the wind without some common sense applied.
It's basic, not rocket science and common sense when you have the experience and know the answers. I've not had to dig into this kind of fault before, hence I'm asking questions, because I don't know the answer. If a sensor is telling you it is lean....the ecu is trying to add fuel but it's still reporting lean. You need to establish if it is lean. And if that sensor is working correctly.
It's not rocket science. Does the sensor behave as expected when you rev, on the overrun, does it ever indicate anything other than lean ?
If you swap bank to bank what happens ?
And again...measure the mixtures on each bank to determine if it is lean.
You're pissing in the wind without some common sense applied.
I have swapped sensors bank to bank with no change in Bank 1's trims, and the B1S1 Sensor (regardless which sensor I put in there) appears to be switching between lean/rich in closed loop as you would expect (with the knowledge I have, which as I said is basic).
The long term fuel trim for Bank 1 starts out high and then gets higher as the engine is put under load. Bank 2 stays in single digits and appears to be fine. Hence I don't think it's a Vac leak, have ruled out MAF, didn't see it being fuel and have therefore focused on the sensors and their feeds to the ECU. New sensors were bought because I didn't fully trust the original ones I found there, and the loom connector for the B1S1 sensor has had new wires spliced in because the old ones were damaged, although that may have been when removing the old sensor (they were very brittle though).
Right now, I suspect the wiring/signal still, but didn't know how to go about checking it - hence the OP, and I've now got a test light on order to see what happens when I put a lean or rich condition to the ECU (as you suggested originally).
How is a test light going to change whether the engine perceives something as rich or lean ?
You need to establish what wire is the signal beck into the ecu, and then apply a relevant voltage to this wire to see if you can see it on the scan tool. That will verify wiring etc.
Test on the good side to find relevant pins. From looking at the sensors available, it does seem to be a regular normal lambda sensor, nothing fancy so that should make it easier.
Or simply extend wiring and swap sensor sides so left sensor feeds into right wiring, and right sensor feeds into left wiring, and see if that changes anything.
Clearly if left bank/sensor has the fault, it should now appear on the right side on the scan tool and visa versa. This will also prove wiring etc.
You need to establish what wire is the signal beck into the ecu, and then apply a relevant voltage to this wire to see if you can see it on the scan tool. That will verify wiring etc.
Test on the good side to find relevant pins. From looking at the sensors available, it does seem to be a regular normal lambda sensor, nothing fancy so that should make it easier.
Or simply extend wiring and swap sensor sides so left sensor feeds into right wiring, and right sensor feeds into left wiring, and see if that changes anything.
Clearly if left bank/sensor has the fault, it should now appear on the right side on the scan tool and visa versa. This will also prove wiring etc.
Was the new CAT OEM? My wife had a Mk5 Golf that was worth a grand and needed a new CAT - VW wanted £1400 for one. So I fitted a non-VW CAT for <£500. The car would still throw the same CEL as mixture readings were outside VWs tolerances. However, it easily past MoT emissions and I didn't notice any increase in fuel consumption. The CAT worked fine but was outside of the OEM tolerances.
So if you haven't fitted a genuine CAT that maybe the issue and there will be no solution other than shelling out for a Lexus one. Ditto O2 sensor. Always by genuine OEM sensors, false economy to skimp on them. Certainly from my experience of Japanese cars.
So if you haven't fitted a genuine CAT that maybe the issue and there will be no solution other than shelling out for a Lexus one. Ditto O2 sensor. Always by genuine OEM sensors, false economy to skimp on them. Certainly from my experience of Japanese cars.
Did you buy the 'Genuine' O2 sensor from ebay?
I only ask because there are loads of fake parts around that have the wrong parameters.
When I tried to buy an air flow meter for a Vectra a few years ago, I was forced to use Dealer type outlet. Autovaux I think.
Everything on the Bay was sub £100.
Genuine part was £200.
The parts and packaging are unbelievably good.
They just don't work properly.
Caveat emptor!
I only ask because there are loads of fake parts around that have the wrong parameters.
When I tried to buy an air flow meter for a Vectra a few years ago, I was forced to use Dealer type outlet. Autovaux I think.
Everything on the Bay was sub £100.
Genuine part was £200.
The parts and packaging are unbelievably good.
They just don't work properly.
Caveat emptor!
stevieturbo said:
How is a test light going to change whether the engine perceives something as rich or lean ?
From this video here: https://youtu.be/7MDtYzHl6AYSuggests to simulate a voltage to the ECU you can use a test light.
Ref all your other suggestions - some have already been done (switching sensors), but I’ve not switched the feeds, so will add that to the list!
mickyh7 said:
Did you buy the 'Genuine' O2 sensor from ebay?
TEKNOPUG said:
Was the new CAT OEM?
Thanks guys and yes - it was all straight from main dealer. I’ve seen far too much utter garbage from eBay, Autodoc, whoever... this was why I swapped out the O2s - some of them seemed ok but others were well out of spec and I installed OE for peace of mind. Cheap is not worth it!!LarJammer said:
Sounds like your intake manifold may have a leak on that bank? Not sure what the setup is on that engine though.
I agree, and before throwing anymore parts at it grab a can of carb cleaner, let the engine idle and carefully, very carefully, spray cleaner around the intake manifold. If the revs rise you have a vacuum leak.NMNeil said:
LarJammer said:
Sounds like your intake manifold may have a leak on that bank? Not sure what the setup is on that engine though.
I agree, and before throwing anymore parts at it grab a can of carb cleaner, let the engine idle and carefully, very carefully, spray cleaner around the intake manifold. If the revs rise you have a vacuum leak.Thing is, with a vacuum leak the trim should have got better with revs, not worse, as far as I’m aware.
Edited by Mike_Mac on Sunday 21st March 18:11
TEKNOPUG said:
Is the issue noticeable when you actually drive the car or is it simply the ODB2 output values?
Usually it drove fine, but it did run unevenly occasionally, which usually happened under load (uphill or similar) but then cleared up when back to driving normally. It also didn’t matter what fuel it was on at the time.Generally this also set the P0171 code, although it would still set after enough time had passed.
Never had any codes for O2 Sensors, Cat, MAF etc, which was why I started looking for leaks, although subsequent reading up suggests the trims shouldn’t get worse on increasing revs if that was the case...
I'm still suspicious of a leak though, but the dodgy wires on the B1S1 O2 sensor still bother me, even though they've been fixed. I'll be very interested in what applying a signal to the signal wire will do!
Edited by Mike_Mac on Sunday 21st March 20:19
Mike_Mac said:
From this video here: https://youtu.be/7MDtYzHl6AY
Suggests to simulate a voltage to the ECU you can use a test light.
Ref all your other suggestions - some have already been done (switching sensors), but I’ve not switched the feeds, so will add that to the list!
SD is certainly a legend, but I would be cautious about using the test light to positive to try and apply a "high" voltage to the signal line to simulate rich especially if you do not know exactly what you're doing.Suggests to simulate a voltage to the ECU you can use a test light.
Ref all your other suggestions - some have already been done (switching sensors), but I’ve not switched the feeds, so will add that to the list!
IMO if you want to do that, get some sort of variable voltage supply and give it 1v first to test.
But first off in your case it would be safer to extend wiring and swap sensors side/side so the working sensor is now on the lean side, but still connected to the functional ecu input/wiring.
I certainly won't disagree with anyone saying to only buy OEM components....but IMO it's daft buying anything without proper testing.
stevieturbo said:
SD is certainly a legend, but I would be cautious about using the test light to positive to try and apply a "high" voltage to the signal line to simulate rich especially if you do not know exactly what you're doing.
IMO if you want to do that, get some sort of variable voltage supply and give it 1v first to test.
But first off in your case it would be safer to extend wiring and swap sensors side/side so the working sensor is now on the lean side, but still connected to the functional ecu input/wiring.
I certainly won't disagree with anyone saying to only buy OEM components....but IMO it's daft buying anything without proper testing.
agreed ref diagnosis via replacement - there were other reasons for doing so. IMO if you want to do that, get some sort of variable voltage supply and give it 1v first to test.
But first off in your case it would be safer to extend wiring and swap sensors side/side so the working sensor is now on the lean side, but still connected to the functional ecu input/wiring.
I certainly won't disagree with anyone saying to only buy OEM components....but IMO it's daft buying anything without proper testing.
Also noted comments ref var. voltage or swap first!
stevieturbo said:
SD is certainly a legend, but I would be cautious about using the test light to positive to try and apply a "high" voltage to the signal line to simulate rich especially if you do not know exactly what you're doing.
IMO if you want to do that, get some sort of variable voltage supply and give it 1v first to test.
But first off in your case it would be safer to extend wiring and swap sensors side/side so the working sensor is now on the lean side, but still connected to the functional ecu input/wiring.
I certainly won't disagree with anyone saying to only buy OEM components....but IMO it's daft buying anything without proper testing.
Definitely this - I'd probably apply the test voltage by knocking up a breadboard with a couple of pots to make a voltage divider, then very carefully test it and apply. Even then I'd be a bit nervous.IMO if you want to do that, get some sort of variable voltage supply and give it 1v first to test.
But first off in your case it would be safer to extend wiring and swap sensors side/side so the working sensor is now on the lean side, but still connected to the functional ecu input/wiring.
I certainly won't disagree with anyone saying to only buy OEM components....but IMO it's daft buying anything without proper testing.
Swapping the sensors side to side with extended wiring is a better test to start.
Gassing Station | Engines & Drivetrain | Top of Page | What's New | My Stuff