Surging and rough idle issue

[NSaneXr]

I'm having a bit of a problem narrowing down an issue with my car that started a couple months ago. It’s a 97 NSX-T with a stock engine with an AP-X exhaust. There is no check engine light and no codes in the system. The symptoms that I have are a more than normal lumpy idle, intermittent surging at low speed and low RPM and the intermittent faint smell of gas from the exhaust.

I recently rebuilt the alternator as I found it was shorting out and causing the car to shut off. Main relay is new as well. I replaced all 6 coil packs this year. Timing belt and valve adjustment were done 4 years ago and the fuel filter and spark plugs were replaced at that time.

Troubleshooting steps thus far:
- checked the fuel pressure and it seems to be within spec per the manual
- shorted out the fuel pump resistor and that didn’t change anything
- unplugged the coil packs one at a time to see if the idle changed (or didn’t) – no effect
- checked the coolant level and it’s good
- monitored the coolant temperature via OBD2 port and it doesn’t seem to fluctuate wildly

Checking the threads it seemed that a dirty throttle body could cause this issue. Before I cleaned the throttle body I wanted to measure the base idle (short the service terminal). I shorted out the service terminal per the manual and found that the idle fluctuates (I didn’t detect this when no it service mode). It doesn’t' remain smooth with the service terminal shorted. It starts to run normally then it drops to around 500 and seems like it is going to stall then it recovers. The cycle seems to be to fairly consistent in timing. Based on some other threads I have read it should be running at 600 steady if all is operating normally. It looks very similar to how the system reacts when you turn the A/C on.

I cleaned all the throttle body passages and it did not help the problem (base idle was still fluctuating).

I was thinking O2 sensors, but I thought shorting the service check connector would remove them from the equation. I also thought about an exhaust leak, but I didn’t find any immediate indications of any.

Am I missing anything obvious that I should be checking?

 

[Old Guy]

Does your OBDII scanner allow you to monitor the output of the primary O2 sensors? The secondary sensors are just for compliance and don't do any control action so don't worry about them. If you can access that it, it won't give you an AFR but it should give you an indication of whether you are running rich (smell of gas from the exhaust). When testing, make sure the engine is up to operating temperature because the ECU does not go into closed loop until its warm. If the front and back differ significantly, that might be an indication that one of the O2 sensors is out of whack.

The ECU gets a signal from the alternator to provide an idle up function when the alternator is charging. Since you had some alternator issues, perhaps check to make sure that function is working. The ECU also gets signals from the neutral position switch in the manual transmission, the clutch switch and the brake switch. These all perform functions associated with the idle control in the ECU. The service manual does not elaborate on what function these various inputs perform; but, it might be useful to eliminate them as a possible cause. The procedures for testing the inputs are in the fuel section of the service manual dealing with idle control. Non operation of these inputs does not generate a diagnostic trouble code so don't go bothering to check.

The engine coolant temperature sensor can also screw up idle control if it is seriously out. However, if you checked the temperature with OBDII tool and it is within reason compared to the dash gauge then that is probably not a problem.

Check your ignition timing to make sure that it is reasonably within spec at idle.

A vacuum leak in the intake system can cause a high idle problem on a MAP based EFI system. That doesn't sound like your problem so I would not go chasing vacuum leaks.

Because the DBW does not have an idle air control valve I don't know how important the cleaning of the passageways in the TB is to idle control. However, its important for other stuff so clean is a good thing. Also, in the service manual Honda does list 'throttle body clogged port' as a potential idle problem causer on DBW cars; but, doesn't provide any useful information like 'what port?'. I have never tried this on my car (yet); but, the DBW should be active when the ECU is powered up; but, engine not running on the off chance that you have to do a flood clear (wide open throttle). Pull the throttle boot off and with the ECU powered up, have somebody slowly actuate the gas pedal and watch the throttle plate to see if it opens and closes smoothly. I have never heard of it happening; but, at some point as these cars age somebodies throttle servo is going to develop problems which could be due to dirt or wear in the drive which results in jerky operation. If the throttle servo doesn't open and close the throttle plate smoothly, that's going to make idle control really difficult. In the service manual, Honda does provide some diagnostics for the DBW system; but, they all seemed to be focused on responding to a trouble code or the DBW being completely dead. The service manual does not seem to anticipate that after 20 years there might be some wear issues that might lead to some performance issues. Anyway, something to investigate as a last resort.

 

[Old Guy]

Update:

I suspect that the port Honda refers to in the service manual is the port that leads to the idle speed control screw. I suspect that this is a passage that bleeds air around the throttle plate to establish the base idle speed. I have never taken a close look into the DBW throttle body so I don't really know where this port is, other than the entrance should be upstream of the throttle plate. I don't know how effectively you could clean that port without removing the idle adjustment screw.

Kaz posted some pictures (post #17 in the following thread) of the idle adjustment screw on a non DBW car and the tip of the screw looked pretty bad.

http://www.nsxprime.com/forum/showt...trol-valve-discontinued?p=1974013#post1974013

Perhaps someone who has removed and cleaned a DBW throttle can provide details on cleaning. The service manual does not seem to cover this

 

[NSaneXr]

As always, thanks for the reply and insights. I did clean the idle speed control screw and port. They did not look as bad as the one that Kaz posted in that other thread. The tip was discoloured, but there was no build up. I wiped it down with some carb cleaner. I used an inspection camera to verify that that port was clear after wiping it down as well. After reinstalling it and setting the base idle as best I could, I am still getting the random drop in RPM with the service connectors shorted.

I am using the Torque app to monitor the ODB2 port information. It allows me to monitor all 4 of the O2 sensors. I did not see any any major differences between the voltages they were showing. I did replace all 4 in the last 4 years and one was suspected of being "lazy". I believe one test that could be performed is to unplug the front O2 sensors (it will throw a CEL) and see if it runs any differently.

You have a point about the throttle motor. I don't recall seeing any changes on the throttle open % when the issue happens. The only thing I thought was odd with the number was it seemed lower when the service connector was shorted (6% from what I remember). It's around 10% with the key in the ON position but the motor not running. I think it is more like 8 or 9% when running. The 6% might be normal in that situation though - I don't see any mention of it in the Service Manual.

I'll look at the alternator. It seems to work to charge the battery with and without accessories on and I don't notice any wild fluctuations, but it was something that was changed.

 

[Old Guy]

According to the SM, the wiper voltage on the TPS should be 0.5 volts closed and 4.5 volts open. The sensor bus voltage is 5 volts, so that probably corresponds to 10% for closed and 90% for wide open on your app. Your measurement of 10 % with the key on; but, engine not running would seem to be right on the money. I am not sure why the throttle position would drop when the engine is running, if anything it should be slightly higher, particularly during warm up when the ECU opens the throttle slightly to speed up warm up. Perhaps use your app to check the throttle position (key on, engine not running) to make sure the throttle % is increasing smoothly to max open. Repeat the open - close cycle a couple of times to make sure that the position always returns to the same number. If it doesn't return to the same number, that could be a sign of some stickiness somewhere which might be messing things up.

I don't think there would be any problem with the performance of the alternator. That alternator connection to the ECU does not affect alternator operation. It tells the ECU that the alternator is under load and that it needs to open the throttle to supply the load, otherwise the RPM takes a dip. If that alternator signal is missing it could contribute to your idle fluctuations as things turn on and off on the car. When you had your problem with the alternator, is there any chance that 12 V might have got applied to that alternator signal to the ECU? According to the SM, the signal out of the alternator looks like a 0-5 volt signal. I don't know what kind of protection the ECU analog inputs have on them. If it got hit with 12 V, it might have cooked that particular input. Only the alternator input test as described in the SM would confirm whether the alternator signal is working correctly.

 

[NSaneXr]

Today I was able to test the throttle body some more. The idle and WOT numbers are consistent. The only odd thing I noticed is that when the engine was fully warmed up and off, the % open fluctuated 0.4% without any input - at WOT the % open was consistent. I thought that must have been the problem for idle, however, with the engine running the RPM fluctuates (between 200-400 RPM) but the throttle body open % never changes. I tried to hold the RPM at 1200 and the throttle open % never changed but the RPM dipped to 1000 and went up to 1400 like it was trying to find 1200 RPM. I could see that the fuel trim and timing advance both changed; so I know the scanner was working. The upstream O2 sensors reported the similar values to each other and voltage was constant.

I did look at the alternator as well. Turning on the accessories (lights, a/c etc.) while at idle resulted in the throttle body opening (a few %) to up the idle rpm; both increases in value (% open and RPM) can be seen on the scanner. Turning the accessories off returned the open % to the previous value. It would seem that the ECU is getting signal in that case and is opening the throttle plate.

I haven't looked into your suggestion on the switches yet. I'll check into those next.

 

[Old Guy]

with the engine running the RPM fluctuates (between 200-400 RPM) but the throttle body open % never changes.

I did look at the alternator as well. Turning on the accessories (lights, a/c etc.) while at idle resulted in the throttle body opening (a few %) to up the idle rpm; both increases in value (% open and RPM) can be seen on the scanner. Turning the accessories off returned the open % to the previous value. It would seem that the ECU is getting signal in that case and is opening the throttle plate.

I haven't looked into your suggestion on the switches yet. I'll check into those next.

It seems like your engine is not going into closed loop idle control. When idling, if the engine speed drops you should see the throttle opening increase as the idle control / DBW opens the throttle to help the engine recover speed. At idle, I would expect to see a small relatively constant motion in the throttle opening. If I can find my Son's OBDII scanner and if it reports throttle position, I see if I can get a handle on what typical throttle motion is like when my car is idling.

Seems like your alternator control is just fine.

Update: Its a no go on the scan tool. As far as I can determine, it only reports DTCs and Freeze Frame data.