- This topic has 8 replies, 4 voices, and was last updated 7 years, 7 months ago by
.
-
Topic
-
The car is a ’96 Accord, 2.2L automatic non Vtec. Engine has about 190k on it. Compression is great, doesn’t burn oil. I started having intermittent coughing sputtering misfiring from the engine. I threw my scanner on it, no codes are showing up. What I am seeing is my upstream O2 is flat at 1.7 volts give or take a bit. LTFT sits at about 8, STFT is all over the place, jumping as high as 20 or so, and when ever I have the misfire, the STFT is very negative, around -30 to -50. This leads me to believe that my misfire is a lean misfire.
Obviously, the upstream O2 shouldn’t be flat at 1.7 volts. I will check the wiring to that to make sure nothing is torn up, loose or corroded, and then replace the O2 sensor if the wiring checks out. My question is, could that be enough to cause the jumping in my STFT and resulting misfire, or is it likely that there is something else at play?
-
Replies
-
IF your 02 sensor truly is stuck low at .17 volts. That would explain positive fuel trims and that would be a lean condition. Just to double verify you can add propane to verify it doesn’t change. Sometimes a snap throttle will do it. What is you down stream 02 sensor doing? It appears your back fire comes in when the car goes rich and is trying to pull back fuel with the negative numbers. Are you seeing black smoke. Smelling harsh fuel smell. Do you have MAF or MAP only? Does it do this at idle? Only after warmed up? Best of luck
Upstream O2 is at 1.3 constantly, I don’t see any fluctuation in it at all. Downstream O2 is not present as the car has not catalytic converter on it.
Fuel trim is generally running positive around 8 or so, but goes very negative when the misfire is present. Misfire can happen at idle, and also at driving RPM. Snapping the throttle will often clean up the misfire, sometimes only momentarily, sometimes longer term, and generally if I open the throttle up enough it will clean up and run pretty well, so seems to happen at part throttle, and idle mostly.
I did smell fuel around the engine after it was misfiring this morning. No visible leaks anywhere. I pulled the vacuum line off the pressure regulator and ran it for probably about 3 or 4 minutes to see if any fuel was getting past the diaphragm, didn’t see anything there. The fuel odor seemed to be mostly on the passengers side of the engine. Only things over there are the charcoal canister, and the pressure regulator. No visible leakage around either.
I haven’t noticed any black smoke, but the misfire mostly happened after dark last night, so I could very well have missed it, I will keep an eye out for that.
Oddly enough, I’m seeing 1.3 volts on my scan tool even with the O2 disconnected. Checking the wiring for the O2 connector (end that goes back to ECU) I see 12v on the heater, and only .4 on the O2 . . but scanner still says ECU is reporting 1.3. I’ll fully admit I am lost here, so if anyone can shed some light on this, I’m all ears.
Hmm, reinstall the catalytic converter . . . you really think that could be it? That could be the cause of my intermittent misfire? I’ll have to give that a try. Kind of surprises me that the car would run fine for 2 years without it, and then suddenly develop issues.
There seems to be no real pattern to when, and how long the misfire / bog lasts for. It acts, and feels very much like a lean misfire. When it is happening, if I get into the throttle, somewhere around the 2/3 to 3/4 area the engine snaps to life. I can watch the STFT jump up to 0, or slightly positive at the exact moment that the engine comes back to life.
Occasionally, I will see my O2 reading drop from 1.3 to .5 either at the onset of the misfire / bog, or at the end of it, and then slowly climb back up to 1.3 Most of the time it just sits at 1.3. I haven’t seen any crazy movement in intake pressure numbers to suggest something odd happening with the MAP sensor. The TPS also shows no signs of bad behavior. The ECM is reporting steady voltage of about 13 volts at all times, and I still have no check engine light. By now I would have expected one for the primary O2 being slow since it has sat at 1.3 volts for most of the day today, and possibly for voltage too high on the primary O2.
Is it possible that the computer is not going into closed loop, and that is why I am seeing the high, stable voltage? I would then assume that during those moments when the voltage drops to .5 or so, that the computer is in closed loop, but for some reason switches back to open loop. Of course, then my next question would by, why, and why isn’t it throwing a code if something is preventing it from going into closed loop.
Then again, maybe the whole issue is the O2 not doing what it should be, and keeping the computer from going into closed loop, but that still leaves me wondering why no CEL for the O2.
After doing a bit of thinking today, I came up with a theory: My theory was that the whole issue was indeed a crazy O2 sensor. My assumption was that the O2 was sticking at 1.23 volts, which would tell the computer that the engine is crazy rich. I then assumed that when the computer attempted to push the O2 sensor lean, it was going way lean attempting to get the sensor to flip. With the long term fuel trim holding at -18, and short term pushing close to -30, but since the sensor wasn’t responding properly, the computer was still seeing what looks like a rich running engine.
Every once in a while, the O2 sensor would jump to life, and readings would change to .3, .4, 1.0, or even 0. At 0, the opposite effect would take place, and short term fuel trim would go to 40 or more positive. At 1.0, the short term fuel trim would make radical swings from positive to negative trying to sway the O2. I then set up Torque on my phone, and set up real time data to show me LTFT, STFT, O2, as well as manifold vacuum, voltage and a few other things, and was able to see some of this happening while I was driving. I theorized that my misfire was indeed a lean miss, caused by the computer compensating for a bad O2 reading.
This would explain the engine running fine when I got into the throttle hard enough, as it switched from closed loop to open loop. I figured if there were a mechanical issue like leaking injectors, EVAP issues, or a leaking fuel pressure regulator, then I would continue to have problems even in open loop. To test this, I disconnected the O2 sensor, which forces the computer to run in open loop, and as I expected, it runs fine with the O2 unhooked.
I will check the wiring from the connector to the O2 sensor tomorrow to make sure there aren’t any problems there, and if that checks out, then I guess it is time for a new O2 sensor.
In the end, my troubleshooting made this more complicated that it needed to be, but I’ve never seen a car run this poorly due to an O2 sensor. Every time I’ve had one fail before, fuel economy got bad, and then the check engine light came on, but drivability never really suffered.
Oddly enough, I’m seeing 1.3 volts on my scan tool even with the O2 disconnected.
That sounds consistent with a problem in the O2 sensor circuit. It would be either a short or a problem with the PCM. Either way, the O2 sensor has nothing to do with reading 1.3 volts because it isn’t even connected.
On Civics from that era, the FSM gives this procedure:
With the O2 sensor disconnected and the key in the OFF position, remove the 4 pin primary O2 sensor connector and check for continuity between sensor terminal 1 (positive sending) and ground. If there is continuity, repair the short. If there is no continuity, replace the PCM with a known good unit and re-check.
You might check the service manual for your car to find the diagnosis procedure.
So, I finally got around to dealing with this.
First off, I’ll start by telling you all that the car has a header, and therefore has an extension on the O2 wiring to attach to the O2 sensor which is now under the car just before the flex pipe.
I was suspicious of that extension cable, so I pulled it out, and went over it. I did find nicks in the insulation on two of the wires, hard to be sure if they were significant enough to case any shorting, so I repaired the harness, and reinstalled it. When I was seeing the 1.3 volts with the O2 disconnected, that entire harness was unplugged, so it was not the cause.
With the harness reinstalled, I was still seeing 1.3 volts for the primary O2 in Torque. Next up, I removed the old O2, and replaced it with a new Denso O2 sensor. When I started the car up, I was still seeing 1.3 volts until the new sensor warmed up, then I saw voltage drop to nearly 0 briefly, and then it began cycling between about .3, and .8 volts just like it is supposed to. I also saw my fuel trims which had been at about -20 LT and – 8 ST climb to 0 LT and +4 ST.
So, apparently, when the ECU is getting no signal from the O2 sensor, for what ever reason what I see in Torque through my blue tooth OBD2 adapter is 1.3 volts. I don’t know why, but that is what was happening. New sensor in, everything works fine, engine runs amazingly, no crazy readings, life is good.
Thanks to all who offered up their advice!
- You must be logged in to reply to this topic.