Help - RPM drops really low then goes back to idle

[soliloquist]

Hello,

I have a 1998 NSX-T with roughly 48k miles. It is completely stock/OEM except for the suspension. I recently had my local Acura dealership perform the 7 year / 105,000 mile maintenance. When I purchased the car it did not come with any prior service records so I have no idea when it was last done (or if it has ever been done).

Prior to performing the timing belt service the car was running without issue.

After the maintenance service I noticed that the car seemed to idle differently. Once the engine warms up to normal operating temperature and there is no load on the engine (setting at a stop light in neutral) the car seems to idle at the appropriate level (800 rpms). If you give it a little gas (rev the engine up to 1,500 or 2,000 rpms) and then let off the gas; instead of the engine returning to the normal level of 800 rpms, it drops very low (almost to a stall) of around 200 rpms. At which time the engine will return back to the normal 800 rpms idle level.

While driving (i.e. engine under load) everything seems to be fine.

I did find these posts in the forums that seem to be similar symptoms:
RPM drops to nearly 0 then goes back up to normal Idle (nsxprime.com)
Cannot find the idle adjust screw.... help please (nsxprime.com)

The second time I drove after the maintenance, the check engine light (CEL) came on. After that I have not driven it at all, and scheduled a return to the Acura dealership.

I did connect a code reader and got the following codes:
P0172 – System too rich (Bank 1)
P0170 – Fuel trim (Bank 1) – pending

While reading the codes I did start the engine and let it warm up to normal operating temperature and checked that when letting off the gas the engine speed was still falling way below the 800 rpm idle level before returning to the normal 800 rpm level. After this warm up and testing I shut off the engine and decided to check the codes again. This time the pending code was gone and the CEL was no longer illuminated. The P0172 code was still there.

Any ideas as to what is causing this?

 

[Honcho]

It sounds like you have a problem with your Bank 1 O2 sensor. Setting that aside, you may also have a dirty throttle body, where the idle idle air screw is not letting in enough air to maintain 800 rpm. I'm not as familiar with the DBW cars, but I believe dynamic idle speed is managed by the ECU controlling the throttle plate.

I would troubleshoot that O2 sensor first. Then, if that doesn't solve it, consider cleaning the throttle body (especially the idle air screw passage) and going through the idle speed setting procedure in the service manual. Acura will just spray some carb cleaner in there and tell you it's clean- it's not. The throttle body must be removed, the idle air screw must be removed, and that air passage must be cleaned out completely. It took me an hour alone just to blast all the crud out of mine.

 

[Old Guy]

Perhaps the easier things first.

The NSX service manual does not list a P 0170 code which is why it may have gone from pending to non existent. P 0172 is valid and is likely triggered by the long term fuel trim exceeding a threshold so it may effectively be the same thing as the P 170 fuel trim limit. Code readers that do not explicitly support the extended Honda error code set can get confused and display odd things on occassion.

P 0172 could be caused by a faulty primary (upstream) rear bank O2 sensor; but, there are other things that can trigger the error code. When O2 sensors fail they usually start to read lean (low output voltage) not rich. If you have access to an OBDII scan tool that can access the real time data use it to display the primary O2 sensor voltages while the car is running and up to operating temperature so that the fuel system is in closed loop. When it is in closed loop, the primary O2 sensor voltages should continuously oscillate from somewhere around 0.1 - 0.2 volts to 0.7 - 0.9 volts. If the sensor voltage is stuck above 0.6 volts then it is probably legitimately a goner. Scan tools that can graph the voltages (typically lap top style) can make it easy to spot this problem.

As noted, other things can cause the P 0172. Since you are not triggering a similar code on bank 2 that sort of rules out a fuel pressure problem. However, a really drooling fuel injector(s) on bank 1 can cause this problem. If you have access to a fuel pressure gauge set it up to measure pressure. Key on the ignition and let the pump prime to get the system up to pressure. If the system will not come up to pressure or it decays quickly when the pump stops you might have a leaky injector (a leaky pump check valve or leaky fuel pressure regulator can also do this). If the static pressure collapses quickly pulling the injectors out to check for a drooling injector might be a good idea.

The other thing that can trigger P 0172 is valve clearance and presumably a timing error on the bank 1 cams. Since the problem popped up after a TB change that is sort of a big red flag. However, checking the timing or the clearances is a fair amount of hassle so I would do O2 sensor voltage and the drooling injector checks first.

With respect to the idle problem. The first thing is to make sure that you have an accurate RPM measurement. An OBDII tool that displays real time values should do this. There is also a test loop in the engine bay to which you can attach an electronic tach for an accurate measurement. The capacitors in the dash tachometer tach circuit age and go out of spec leading to very significant errors in the displayed RPM. You definitely do not want to be using the dash tach to set idle RPM if you need to make an adjustment.

Honda has a whole bunch of Honda specific diagnostic trouble codes that are beyond the mandated OBD II diagnostic codes. Generic code readers may not be able to read these codes or will display them incorrectly. If you have an OBDII scanner that supports the extended set of Honda error codes, there are a number of error codes associated with the drive by wire and idle control system. Checking to make sure that none of these codes are stored in memory would be good because that might immediately point you in the right direction.

When you check the base idle speed, do it under no load conditions (no AC, no interior fan, no head lights ...). The alternator and the AC send an idle up signal to the idle control system. If the engine idle speed is correct with all that stuff shut off; but, switching on the AC or adding a lot of electrical load causes the idle to drop then one or more of the idle up signals may be missing. There are also switches on the transmission, brake and clutch which tell the ECU that you have come to a stop light and that it needs to go into idle control. If those switches do not operate correctly your will not have active idle control and the engine RPM may tank when you come to a stop sign and press the clutch in or shift to neutral. The service manual sets out the test and diagnostic procedures for all those idle control inputs (too long for an already long post).

The base idle speed adjustment screw on the drive by wire cars is on the top of throttle body on the intake side or the throttle body. The screw faces sideways toward the front of the car. Only consider fiddling with this screw as an absolute last resort. Mess with this screw and then find out that the problem is actually somewhere else and you are going to create a bunch of work for yourself getting everything correct. The idle adjustment screw sets a base level of by-pass air for the engine. What can get messed up is if something gets sucked into the port for idle air screw which reduces the by-pass air flow and causes the idle speed to tank. I can't recall where exactly the inlet port is for the idle adjustment screw; but, it should be fairly close to the screw on the up-stream side of the throttle plate. This is where you might want to try a can of throttle body cleaner with a straw jammed into that inlet port to see if you can clean it. Unfortunately, if something did get sucked in and is jammed you probably need to back flush the port and that requires removal of the throttle body. Try from the front first to see if that fixes it because it is easy to do.

Do you have an oiled element air filter? If they are over oiled the oil gets sucked out of the element and can plug up the idle air orifice and do other bad things. I personally prefer the rather expensive OEM air filter.

If you feel that you need to fiddle with the base idle speed, get a copy of the service manual and follow the idle speed setting procedure (page 11-129 on in the 1997 SM).

Good luck and hope its the easier stuff.

 

[soliloquist]

Thanks for all the helpful info.

I am getting the HO2S B1 S1 replaced. Will start there and see if that solves it.

A few additional details, my OBD II scanner is an Innova 5210. The live data for the engine RPM matches the tach in the display. It is not a readout issues, when the RPMs drop the engine noticeably bogs down (you can tell by the engine sound, another indicator is the interior lights dim for a brief moment when the RPMs are low).

The Acura techs are looking into it now and their scan tools brought back the same codes as mine. According to the service manual when checking the HO2S the Volts should be between 0.3 and 0.6. Mine was reading 0.74 volts. Going to replace that and see if it solves the idle issue too.

As for the air filter. I recently replaced the air filter. I am just methodically going through all the maintenance because I have no records of what was done before. I installed an OEM filter. The filter that I removed was a K&N filter that looked pretty new (to my untrained eye). When it was out, I did inspect the throttle body. I did not notice any indication of excess oil from the filter collecting on the throttle body. It looked perfectly clean (again to my untrained eye). I have no idea how long the K&N filter was in there, or if it was ever re-oiled.

It was certainly my thinking that the problem was related to the timing belt change. But maybe not, we shall see. I was thinking it was a small vacuum leak that was proportionally significant at low RPM and essentially insignificant at high PRM. Thats why it would manifest near idle and not under load. And then when the ECU detected the drop it was compensating to bring the idle back into spec. Looks like my theory was wrong.

Thanks again. Will post updates when I have them.

 

[Old Guy]

If your up stream O2 sensor voltage was reading a steady 0.74 volts that is a fairly strong indication that it has exceeded its best before date.

I didn't intend to create a distraction by referring to the oiled element filter. The primary point was that if something got into the up stream port of the idle air by-pass circuit it could be blocking air flow past the idle adjustment screw and causing low idle. I don't think your problem is a vacuum leak. The NSX engine uses a MAP based ECU and manifold air leaks elevate the idle RPM, not depress it on MAP based systems.

My comment about the valve clearance error was with respect to causing the P 0172 error code. Given that you likely have diagnosed the P 0172 error code, I don't think the timing belt change is material to the idle problem unless they accidentally pulled a wire loose in the idle control inputs during the TB change.

 

[soliloquist]

Update:

The HO2S was replaced and all the codes are gone for the moment (haven't had time to drive too much yet).

The idle problem remains unchanged.

I am now looking down this line of thinking for the possible solution:

Idle air control valve discontinued

Well, looks like Honda discontinued yet another part for our cars. I was able to find one on Amayama though for 183 bucks. After having my full motor build and having Shad analyze my tune he stills thinks this valve is the reason I go lean and almost stall sometimes on throttle lift. ‘My...

www.nsxprime.com

I know that thread is mostly about the 91-94 NA1 that had the electronic air control valve (EACV) and that my 98 does not. But it did seem like a similar issue and there was some reference to the 1997-2005 service manual.

In particular this quote by Larry Bastanza seems like it would be worth a try:
"Just wanted to mention I have had this issue with many NSX's. On a stock engine I have been able to resolve this easily by clearing the idle passage (remove idle setting screw and blow out the passage). ..., but just wanted to mention it, since I have seen it so often."

This seems simple enough to do but is there anything tricky about removing the idle adjusting screw? Is there anything I need to be aware of when putting it back? When I "blow out the passage" do I just use compressed air, or is there a better method?

Looking at the 1997-2005 service manual page 11-151 it does identify the "idle adjusting screw" but then in the Inspection section it has the caution "Do not adjust the throttle stop screw. It is preset at the factory." I could not find the throttle stop screw referenced anywhere. Is that the same as the idle adjusting screw? Or is it something different?

I have found the part number: 16016-PM6-005
Seems cheap, is it worth trying to clean the old one vs just replacing it with a new one?

By the way Kaz has some wonderful pics in that link that are very helpful!

He also has pics of a gummed-up throttle body that Honcho mentioned above. As I mentioned above as well, I did inspect mine and it did not look anything like that. I will disconnect the air box again and inspect it again. If the idle adjustment screw cleaning does not produce results I might try cleaning the throttle body.

Thanks for any help!

 

[Honcho]

Update:

The HO2S was replaced and all the codes are gone for the moment (haven't had time to drive too much yet).

The idle problem remains unchanged.

I am now looking down this line of thinking for the possible solution:

Idle air control valve discontinued

Well, looks like Honda discontinued yet another part for our cars. I was able to find one on Amayama though for 183 bucks. After having my full motor build and having Shad analyze my tune he stills thinks this valve is the reason I go lean and almost stall sometimes on throttle lift. ‘My...

www.nsxprime.com

I know that thread is mostly about the 91-94 NA1 that had the electronic air control valve (EACV) and that my 98 does not. But it did seem like a similar issue and there was some reference to the 1997-2005 service manual.

In particular this quote by Larry Bastanza seems like it would be worth a try:
"Just wanted to mention I have had this issue with many NSX's. On a stock engine I have been able to resolve this easily by clearing the idle passage (remove idle setting screw and blow out the passage). ..., but just wanted to mention it, since I have seen it so often."

This seems simple enough to do but is there anything tricky about removing the idle adjusting screw? Is there anything I need to be aware of when putting it back? When I "blow out the passage" do I just use compressed air, or is there a better method?

Looking at the 1997-2005 service manual page 11-151 it does identify the "idle adjusting screw" but then in the Inspection section it has the caution "Do not adjust the throttle stop screw. It is preset at the factory." I could not find the throttle stop screw referenced anywhere. Is that the same as the idle adjusting screw? Or is it something different?

I have found the part number: 16016-PM6-005
Seems cheap, is it worth trying to clean the old one vs just replacing it with a new one?

By the way Kaz has some wonderful pics in that link that are very helpful!

He also has pics of a gummed-up throttle body that Honcho mentioned above. As I mentioned above as well, I did inspect mine and it did not look anything like that. I will disconnect the air box again and inspect it again. If the idle adjustment screw cleaning does not produce results I might try cleaning the throttle body.

Thanks for any help!

This is actually what I meant by cleaning out your throttle body. While it is important to make sure the 4 vacuum lines are clear, the real key is the idle air passage. This passage is identical on the Cable and DBW cars, as it provides the base idle air for the ECU. On the 91-94 cars, the EACV and FITV control the idle after the base air. On the DBW cars, this is accomplished by the throttle plate motor.

Your symptom describes exactly what would happen if the idle air passage is obstructed. When the DBW throttle plate returns to the idle setting, the overall airflow is too low and the idle speed will drop low. The ECU will sense this and open the throttle plate (or EACV on the 91-94 cars) slightly to bring the idle back up. Remember that idle speed = Base Idle Air + Throttle Plate Air.

To clean the passage, you should remove the throttle body (not hard). Break the paint seal and unscrew the idle setting screw (this can be tough- just take your time). Using a carb/TB cleaner, spray into the passage and keep spraying until you get a clear stream out of the other end of the passage. The passage is only about 1/8" wide, so I found the straw helps. Watch your eyes- if the passage is blocked, the cleaner can shoot back out at you!

Once the passage is 100% clean, get your new idle screw, put some silicone grease on the O-ring, and screw it back into the passage. Then, perform the idle setting procedure in the service manual.

Here is a picture of the idle air passage on my 92, along with the old and new screws for reference:

 

[Bram]

A dirty IAT sensor cause all types of strange issues....need a good cleaning

Bram

 

[Old Guy]

The throttle stop screw is not the idle adjustment screw. I believe that it is the little allen head set screw retained by a lock nut and painted yellow on the front of the throttle body. It faces up and is close to the output shaft of the idle control motor. Definitely do not go messing with that adjustment. The procedure for setting the base idle speed is on page 11-129 of the service manual.

I suggest that you try to fire some throttle body cleaner through the up stream idle port as a Hale Mary and then try the engine to see if the idle problem is fixed. It will likely not fix the problem; but, takes 30 seconds to do so worth the try. If that does not fix the problem then you need to remove the throttle body and back flush the port (obviously need new gasket). I would unscrew the idle screw to help insure complete cleaning. Make a Sharpee mark to register the screw slot location and then count the turns out so that on reinstallation you can get it close before going through the idle setting procedure. The choice on idle screw replacement is yours.

Bram has a good point, if something gummed up the idle port your IAT sensor may also be gummed up. That can cause low low air temperature measurements which causes low air fuel ratios because the ECU thinks the engine is getting denser air that is actually present.