NewCastle1712

I attempted to add this topic to the forum yesterday, so pardon me if it appears twice.

I own a 2000 Honda CRV and have a problem with fuel trim that only occurs at idle. Rather large negative STFT numbers (as large as -18%) occur with the vehicle at idle and the LTFT also gradually creeps down to rather large negative values (as much as -6.2%) if the vehicle stays at idle for some time (say 6 minutes). Once the vehicle is in motion again the numbers approach normal values within a few minutes of driving (i.e. the LTFT approaches zero).

From the reading I have done on fuel trim, it appears that the engine is running rich at IDLE and ECU is trying to compensate for the condition by reducing the injector pulse width. Besides the appearance of the negative fuel trim, everything else about the engine appears normal. There is no stalling or stumbling at idle, the engine is idling at normal speeds and the engine operates normally during driving and is responsive to the throttle.

I have performed the following maintenance actions or tests (not necessarily in the order presented below) but have as of yet been unable to solve the problem:

(1) Ran the Honda idle relearn procedure
(2) Replaced virtually all of the ignition components (coil, ICM, rotor, distributor cap, wires, plugs)
(3) Replaced the fuel pressure regulator (vehicle doesn't have a fuel pressure Shaffer on rails, so took a chance and replaced regulator)
(4) Checked to make sure the fuel return to tank was not restricted (by temporarily directing fuel from return to a portable tank)
(5) Cleaned the IACV on the vehicle with Q-tips without removing the coolant lines. Stainless steel screen was quite dirty before cleaning.
(6) Made sure that the evaporation canister was not purging into intake manifold during idle (no vacuum pulled on input from canister to purge valve)
(7) Checked manifold pressure at idle (values were as expected and steady)
(8) Replaced the catalytic converter (this is what started the whole search for idle problem. The catalytic converted was almost completely melted inside. Bottom line is that I don't think there are now any exhaust restrictions that could affect fuel trim.
(9) Checked for vacuum leaks and found none (even though I think that in most cases a vacuum leak would cause a lean condition).

So, any one out there have any ideas on what may be occurring here?

NewCastle1712

Oops! Forgot to mention that I ran Sea Foam through the fuel system just in case the problem might have been caused by dirty injectors - no noticeable affect

Muckman

iTrader: (2)

Why are you assuming fuel trim is a problem? That's exactly how a closed loop system is supposed to work. Various engine and environmental conditions effect the air fuel ratio and your Ecu is using the oxygen sensor for feedback to adjust the correct amount of fuel.

NewCastle1712

Thanks for the response. I knew that fuel trim was supposed to compensate for wear and environmental factors but had read that numbers greater than about +-10% on STFT and +-5% on LTFT were considered as a potential problems. As indicated, the fuel trim numbers I have been seeing exceed this. Are the trim numbers I have been seeing actually considered in range for this engine ? Also, as mentioned, the trim numbers other than idle are well within range and the LTFT is very nearly zero when driving. Why am I seeing the larger trim numbers only at idle?

VegasInvasion

vac leak, poor injector flow, poor battery connection or alternator failure, air restriction, bad IAT, bad PCV valve, low compression, leaky valves, cracked head gasket.

NewCastle1712

Thanks for the input.

I had checked the IAT sensor by taking my IR thermometer and pointing the laser at the Inlet Air Temperature sensor on the engine and comparing that reading with the reading that I got from "live data" on the OBDII scan tool. The readings were within a few degrees of each other.

I compared the MAP sensor output from the OBDII "live date" tool with barometric pressure with the engine not running but ignition switch in the "on" position and saw that they were very close to each other (barometric pressure provided courtesy of home weather station)

I had previously done the "shake" test on the PCV valve and found no problems but went ahead and replaced the valve to make sure with no affect on the problem (also held a finger over the PCV tube back to the intake manifold to make sure that the tube was intact and found that there was A strong vacuum there)

I have checked the air input making sure there were no restrictions and I have replaced the air filter.

Concerning the injectors. I have ran the common injector cleaners through the system but have not performed other diagnostics. The engine performs great off idle and the mileage is better than it has ever been. Also as previously mentioned, the fuel trim off idle is "smack on" and very close to zero (did see the LTFT climb to 0.8% on a 30 mile trip yesterday).

I have not tested the voltage at idle but will do that today. I am wondering though if a low voltage would cause the injector pulse width to shorted and cause a leaner mixture?

I did check the manifold air pressure at idle using a vacuum gauge and found it within expected range and very steady. Leading me to believe that there is not an engine problem. Also as mentioned the vehicle operates great even at idle (not stalls, stutters, plenty of power, responsive to throttle, etc)

I had previously performed a vacuum leak test by carefully spraying a combustible fluid
(I think I used brake cleaner) on vacuum connections.


Since the problem (i.e. high negative fuel trim numbers) only occurs at idle I'm thinking that it must be associated with vacuum, throttle plate bypass or an improperly performed idle relearn procedure. Other?


By the way, there is more than just an academic concern about large negative trims at idle. What started this whole mess is that I burnt out a catalytic converter tracking down a problem with an intermittent vehicle stall during operation (a real thrill on the highway by the way). While I was attempting to find the problem (which turned out to be an intermittent coil) I overheated the catalytic converter during long runs at idle attempting to duplicate and isolate the engine stall problem (weak coil --> rich mixture --> overheated catalytic converter). While I am not now operating the vehicle at idle for long periods of time, you can understand why I am concerned about the implications of a rich mixture at idle implied by the negative trims I am now encountering.

Help!

NewCastle1712

I had previously checked the MAP at idle on the vehicle using a vacuum gauge and found that vacuum readings appeared in range and steady. For kicks I thought I would collect MAP readings at idle using live data from my OBDII scan tool. Here is what I found

MAP with engine off, ignition on 14.2 PSI (barometric pressure from home weather station was reading 13.9 PSI. That's about a 2% error)

MAP at idle, engine warm, all accessories off 5.7 to 5.8 PSI (steady)

Converting PSI to in-Hg, this results in a vacuum at idle between 17.1 and 17.3 in-Hg. From the vacuum diagnostic web pages I have reviewed, these reading while within "normal range" (17-22 in-Hg) are on the low side of normal. So is it possible that there is a very small vacuum leak in this engine causing the negative fuel trim problem I am seeing? If so what particular subsystem on the engine would cause a rich condition on slightly lower than normal vacuum? The engine by the way has about 175000 miles on it.

Jim Houston

NewCastle1712

From my other recent post on fuel trim I became curious on what the Honda Idle relearn procedure was all about. That is, exactly what parameters are used as inputs (for a speed-density fuel trim system) and what controls are affected?

Which of these are inputs? MAP sensor, RPMs, IAT sensor ECT sensor, O2 sensor, knock sensor, others?

Which of the following outputs are affected - Input Idle Control Valve motor, fuel injection pulse width, ignition timing, other?

Does the learn procedure build a new fuel trim map which is stored in memory?

Thanks,

Jim Houston

VegasInvasion

The honda ecu is pretty impaired when it comes to learning. Idle is a basic closed loop system, but IIRC it's just a short term trim. It reads rpm, tps, map and ect. It adjusts fuel, ignition, and idle duty cycle. There are some conditions that change the outputs slightly, like A/C and purge, but they're static values.

VegasInvasion

improper voltage will absolutely affect injector phase. particularly improper ground, but if there's an alternator or battery issue that may be the culprit as well.

I didn't see any results on a leakdown or compression test, which could easily be the problem.

I think you can rule out vacuum issues. You may have a dirty throttle body though, which could cause idle issues.

If you overheated the cat, there's also a good chance you damaged the o2 sensor(s), so if those haven't been replaced I would do that before anything else.

NewCastle1712

Thanks. If the exhaust manifold (i.e pre-CAT) O2 sensor is not used in the idle re-learn procedure, it might explain why resetting the ECU and running idle re-learn procedure is not making any difference in my situation (large negative STFT at idle). Perhaps the computer is assuming that a certain level of throttle plate air flow bypass is being supplied by the IACV based only on MAP, TPS, ECT and RPM sensors? Could it be that there is still some air restriction in the IACV resulting in a rich mixture after the computer has set the IACV to the desired idle level? I did clean the IACV on the vehicle without removing the coolant lines. Perhaps I need to go ahead and remove the IACV and give it a very thorough cleaning? Maybe also check the bypass air channels in throttle body to make sure they are also not partially clogged?

Jim Houston

NewCastle1712

Thanks. Got tied up today but will run voltage test first thing tomorrow AM. Think I will also talk to Auto-Zone about borrowing their compression tester.

When I purchased the new catalytic converter I bought a couple of Bosch oxygen sensors for insurance in case I needed them to pass Texas State inspection. I was able to pass inspection without replacing the oxygen sensors but perhaps now I need to consider replacing them.

The really confusing thing about this problem is that it ONLY occurs at idle. Fuel trim off idle is near perfect. In addition the engine seems to be performing great. No stumble or searching at idle, responsive throttle, plenty of power and good gas mileage! But I remain very nervous about the possibility of screwing up another catalytic converter due to an overly rich A/F mixture at idle.

VegasInvasion

could just be a small exhaust leak then. I merged your threads because it's for the same issue, and it takes me twice as long to respond to two threads. You're overthinking this problem, it's good to know how things work in detail, but a fuel trim issue is simply process of elimination. IACV issues are pretty black and white, you should rule this out as the problem.

NewCastle1712

Thanks. Thanks also for combining the threads.

I am assuming that the exhaust leak if it exists would be before the exhaust manifold oxygen sensor and that this leak would cause the oxygen sensor to "think" the mixture was lean causing the ECU to add more fuel? Would make sense also that a condition like this would only occur at lower RPMs as the ratio of leaked gas to total exhaust gas would be greatest.

Any hints on how to find a small exhaust leak?

NewCastle1712

I checked the battery voltage this morning and found the voltage at idle (no accessories) was at about 14.4 volts. I believe this is within normal range for this vehicle. Checked the fuel trim reading at idle at the same time. It is still showing a rather large negative STFT numbers (less than -14%)

NewCastle1712

Reviewing the OBDII live data today and noticed that the TP sensor was showing about a 10% deflection of the throttle plate at idle. I would have thought that the TP reading should be at zero at idle and that all of the air entering the engine be under control of the IACV. Could my problem be that the ECU is thinking that the throttle is partially depressed and is lengthening the injector pulse width accordingly? If this is the problem is the fix as simple as adjusting the tension on the throttle linkage?

Jim Houston

kyden

reading 10% is correct.

NewCastle1712

Thanks. Back to the drawing board!

Jim Houston

Johnny_9

compression test.

NewCastle1712

I thought it might be useful to capture some of the parameters from my OBDII scan tool and present them here in case someone saw anything strange (other than the large negative fuel trims) . Below are the last 3 frames of 131 frames for a test conducted this morning.

Fuel system 1 statusCL
Calculated Load Value(%)35.7
Engine Coolant Temperature(¡£F)194
Short Term Fuel Trim -Bank 1(%)-14.8
Long Term Fuel Trim - Bank 1(%)-14.1
Intake Manifold Absolute Pressure(psi)5.2
Engine RPM(rpm) 765
Ignition Timing Advanece for #1 Cylinder(¡£)16.0
Intake Air Temperature(¡£F)145
Absolute Throttle Position(%)9.8
Oxygen Sensor Output Voltage Bank 1-Sensor 1(V)0.155
Oxygen Sensor Output Voltage Bank 1-Sensor 2(V)0.210
----------------------
Fuel system 1 statusCL
Calculated Load Value(%)34.9
Engine Coolant Temperature(¡£F)194
Short Term Fuel Trim -Bank 1(%)-12.5
Long Term Fuel Trim - Bank 1(%)-14.1
Intake Manifold Absolute Pressure(psi)5.1
Engine RPM(rpm) 770
Ignition Timing Advanece for #1 Cylinder(¡£)15.5
Intake Air Temperature(¡£F)145
Absolute Throttle Position(%)9.8
Oxygen Sensor Output Voltage Bank 1-Sensor 1(V)0.895
Oxygen Sensor Output Voltage Bank 1-Sensor 2(V)0.875
----------------------
Fuel system 1 statusCL
Calculated Load Value(%)34.9
Engine Coolant Temperature(¡£F)195
Short Term Fuel Trim -Bank 1(%)-14.8
Long Term Fuel Trim - Bank 1(%)-14.1
Intake Manifold Absolute Pressure(psi)5.2
Engine RPM(rpm) 755
Ignition Timing Advanece for #1 Cylinder(¡£)14.5
Intake Air Temperature(¡£F)145
Absolute Throttle Position(%)9.8
Oxygen Sensor Output Voltage Bank 1-Sensor 1(V)0.195



Does anyone see anything strange that may be causing the large negative values for fuel trim?

NewCastle1712

NewCastle1712

Did a little nosing around on the internet but could not find anything definitive on what the calculated engine load at idle should be for the 2000 CRV engine. Nevertheless the approximate 35% load at idle seems awfully high. I assume that the fuel injection schedule is based on the calculated load. So, if the calculated load is greater than actual load wouldn't the ECU be adding more fuel than necessary to maintain the desired A/F ratio?

I assume that the calculated load (i.e. ratio of current air flow to maximum air flow) is based on both sensor data (RPM, IAT, MAP and maybe ECT) as well as known information on the engine (displacement, volumetric efficiency, etc).

As mentioned earlier I checked the MAP sensor reading against barometric pressure from my home weather station with the engine not running and found they matched within 4%. I also took a reading at the IAT sensor on the intake manifold with my IR thermometer and found that it matched the IAT reading from the OBDII scanner within a few degrees.

So, it boils down to two questions; (1) is a 35% load at idle abnormal for this engine and (2) can I count on the static readings I made on MAP and IAT to accurately reflect dynamic conditions with the engine running?

Jim Houston

Flr Power

Have you done a leak down test?

NewCastle1712

Not yet but plan one one soon.

NewCastle1712

I found Honda Service Manual on line and looked up specified values for Calculated Load Value (page 268 of 1395) the values were as follows:

28-41% at idle speed
13-20% 2500 rpm with no load

So, it appears that my 35% at idle is within the specified range