Hi Folks –
Newbie here to the forum as well as being an owner of my first Honda as of two days ago. Live in MI but came to FL to buy and drive home a rust fee car for the college senior grand kid - something not possible in MI.

Anyway, my problem is virtual identical to rcabl, the original poster. The car drives beautifully for several miles and then the hesitation, bogging and rough running begin. In desperation, initially installed a new Auto Zone ignition coil simply because the situation was consistent with my experiences regarding the secondary ignition going astray. I was wrong because it changed nothing - thought for a while the problem was solved but in about 30 minutes, it came back. So I had replaced a good coil with another good coil - me bad! LOL!

After fiddling around with it for a day or so, I discovered the thing ran a ton better with the throttle position sensor disconnected - it goes into limp in mode then. There was none of the problems as described above, but when in limp-in mode, the trans shift points were at higher speeds and 4th gear was unavailable. Also had way less power because of the fixed sensor parameters when in limp-in. So, with the car running better with the TPS disconnected, I condemned the original TPS and replaced it with a new one. Guess what? Nothing changed - I had replaced a good TPS with a new TPS - me bad again! LOL!
　
Becoming somewhat super desperate late this afternoon, I sprang for a Harbor Freight scan tool that has real time sensor readings and other real time data observation. Exactly as described by the original poster rcabl, the engine was operating beautifully until the short term fuel trims began to go into negative numbers. Going down the expressway, the STFT was about +40 (yeah, super rich, huh) and the car still ran very well until the STFT got into the negative numbers. Anything below -10 to -12 (negative) results in the hesitation, bogging and all the other problems described above. The scan tool consistently had readings of negative 28 which seems to be the lean limit of the range of authority of the engine controller. It appears that +48 was the enrichment limit of the controller.

Connected the TPS connector once again and as expected the STFT went to zero - apparently there is no fuel trimming when in limp-in mode but rather the controller just sends a programmed amount of fuel from a base map of sorts. Even during limp-in mode, all sensor readings were normal. Coolant temp, intake temp, spark advance, TPS, baro and all the others were as expected and they changed as expected with speed and throttle changes. (think these sensor inputs to the controller were ignored during limp-in since the controller went into a default map of sorts)

So, the engine control module is seeing near perfect voltage and other events from all sensors but the engine controller is sending total garbage commands to the injectors. The fuel control is seemingly out of control since it is sending random full rich and then full lean injector commands when the engine controller gets warmed up. The engine also consumed excessive amounts of fuel – the thing is getting something less than 15 mpg on the highway which is consistent with the full rich STFT’s.


After seeing the scan data during several bouts of the problem, I was beginning to condemn the engine controller and after reading this post have become even more convinced the controller is going south as it heats up.

The question is – has anyone had any experience with rebuilt engine controllers from places like NAPA or any of the other after market parts places? Just not in the mood to give a ton of $$$ to Honda right now.

Sorry for the rant but I’m more than a little "upset" with this used car that’s so far from home! LOL!!

Jake

 

I moved your post to a new thread for better attention. Tagging onto an old post is tedious for new readers is questionable, since you don't know the source of your problem.

By the way w/ all that work, did you check for any trouble codes?

good luck

 

This is not a fuel starvation situation but rather it's a problem with the engine controller driving the short term fuel strategy full lean despite all the sensor inputs being normal for their respective operating conditions - at least they seem to be. The problem happens only after driving for about a half hour or so when for no apparent reason, the ECU decides to use its full range of authority and reduce the STFT to full lean. Immediately prior to these events, the O2 operation and response is totally normal, - there is no reason for the ECU to shorten the injector pulse width - but it does.

Disconnecting the TPS places the system in limp-in mode after which there is apparently no short term fuel trim at all. The STFT as displayed on the read-out box is always at zero during limp-in, it never moves toward richer or leaner. So, I guess what I was trying to say is that if faced with this bucking, surging and hesitation condition it might be prudent to unhook the TPS and get the system into limp-in mode. Limp in mode is way better than trying to drive the car with the problems listed above. When the STFT is bouncing all over the place, the engine goes from a coast condition (the engine is no longer running but just free wheeling) to a power condition and then back and forth again and when so doing, the driveline is taking a severe beating. The transistion from coast to power is very abrupt, very similar to turning the key on and off, and the driveline lets you know that it's not too happy when banging around. At least the limp-in mode will get you home, albeit with a corresponding loss of power and increased fuel consumption. The trans shift points during limp-in are elevated and there doesn't seem to be any fourth gear but the engine smooths out a lot - at least in my old beater - LOL!

There are no codes except when disconnecting the TPS. Curiously enough, the system stays in limp-in mode even if the TPS is connected once again. To get out of limp-in I had to use the scan tool or unhook the battery.

The original poster rcabl said that a replacement computer fixed his problem and I'm leaning that way. Conversations with a few independent techs yielded the same opinion even though they had not seen very many Honda ECU's fail. Went to a local Honda dealer and tried to talk with a tech there but understandly I couldn't get past the service writer.

Anyway, a used replacement engine controller will be here tomorrow. Tried NAPA and others for a reman but doesn't seem to be any available and a new one from Honda will cost over $700 including the require reprogramming fees. Hope it fixes the problem. We'll see and I'll update the thread as the situation develops.

Going to check for an after market alarm system too since one tech had seen problems with them before.

Again, sorry for the rant.

Jake

 

In an attempt to familiarize myself with the location of the ECU and to understand the service procedure in anticipation of installing the used ECU, I folded the carpet back from the center console area to expose the retaining screws and the electrical connectors. I left the carpet folded back from the center console area on both sides.

Off I went to get the used ECU just knowing that at any time I’d have to disconnect the TPS and put the system into limp-in mode, but guess what? The car ran like a top! I picked up the used ECU from the salvage yard and put it in trunk but never had to use it all the way back to Michigan from Florida – distance of about 1250 miles.

My current thought is when the carpet was folded back from the center console area it let in some additional air thereby cooling the ECU.

I have no idea if this car is fixed or not but it seems to be OK right now. Just a thought, but if others are experiencing this problem, it might be wise to fold the carpet back and cool the ECU a little to see if it helps. Perhaps older electronics are more susceptible to elevated temperatures than they were when new – I don’t know this for sure.

Anyhow, it takes all of about 45 seconds to fold the rug with no dollars involved so it might be worth a try. Not promising anything but seeming it has worked for me.

The car is now in the hands of the grand daughter and I'm sure if anything bad happens to the Honda I'll be the first to hear about it. LOL

Peace.
Jake

 

Here's a thought...

Maybe something is flaky with the wiring harness plugs? So by turning back the carpet, you flexed the wires a bit and the connection is better? That's not really a FIX, but it's something to check out if the problem comes back.

 

Hi, Jim – thanks for your response.

The situation might very well be caused by wiring or perhaps not. This is one of those times when I don’t understand all that I know. I’ve been casually thinking about this thing for several days now and I guess it might be possible for a stray voltage of some sort that has found it’s way into the O2 feed back circuit (wiring) thereby fooling the ECU into thinking the engine was running full rich.

But then again, I have to refer back to rcabl, the original poster (seemingly quite intelligent) with this problem, which was identical to mine in that the ECU was driving the STFT full lean during part throttle operation even though the O2 sensor input to the ECU was totally normal. After fiddling around for several weeks, rcabl finally solved his problem with an ECU replacement.

My engine would operate perfectly when the throttle was wide open at which time the feedback system was in open loop. All most all the time after the normal 20-minute warm-up period, the system would drive the injectors full lean when it was in closed loop (O2 feedback mode) until the throttle was opened a little (lot?) at which time it went into open-loop and ran normally. For whatever reason, during closed loop, the ECU was thinking the O2 sensors were full rich when they were not.

During the trip from FL to MI, I stopped north of Chattanooga and trimmed a couple of inches off the carpet where it was in close contact with the ECU. Both the drivers and passengers’ side got trimmed. The carpet got pushed back under the edges of the center console and the new opening is not obvious until you get down on your knees like when vacuuming the floor pan.

So, now the car is running with a "ventilated ECU" - LOL! But, it seems to work.

Another thought which crossed my mind is the exhaust system is in close proximity to the floor pan where the ECU resides and perhaps a heat shield has fallen off (removed) thereby allowing excess heat into that area.

As mentioned before, I really don’t know if this car is fixed or not. It’s cold here in MI so we might have to wait until next summer to find out for sure.

Anyway, just what I did – hope it might be of help to someone else with a similar problem.

Jake -