VERY rough idle at cold start?

[Matt'sNSX]

Hello all! I this finds everyone and their families all well and that everyone had a great Labor Day weekend! My name is Matt (Obviously!) and I recently purchased a very, VERY nice 96' NSX that I couldn't be happier with. The NA1/NA2 NSX is quite literally my dream car and it's all I've been thinking about since I got it. Quite literally a dream come true.

I do have a few concerns, one of which is a very rough idle at cold start. It feels like a dead miss, but no MIL and I'm assuming I'd be seeing a MIL if the engine were indeed misfiring. Curiously and thankfully, the rough idle goes away after 2-3 minutes and it runs quite well after that. While the car has less than 46,000 miles I did have the timing belt and water pump changed as the previous owner had not changed it and didn't know if any prior owners had either. It was his opinion that there was no need to change it due to the low mileage but I was paranoid due to the age so I had it changed. Turns out he was probably right as the belt looked perfect although the belt tensioner bearing was losing some grease.

I mention the timing belt/water pump swap because the vibration may have gotten worse after the work was done. I say "may" because there's a very realistic chance that's strictly my imagination/paranoia due to the timing belt change. I had only driven the car 3-4 times prior to the change so it's not like I was intimately familiar with the car's cold start behavior but there's no question that the cold start rough idle was there before the swap so maybe there really was no point in bringing it up. Better too much information than not enough I suppose.

One other possible clue is that the car (in my opinion) smells very rich at a true cold start (i.e., overnight or 12+ hours). I kind of suspect that I may be experiencing a rich mixture induced misfire from some sort of cold start strategy, but what do I know? Any thoughts and opinions will be sincerely ​appreciated!!

 

[docjohn]

where are you located?

 

[Matt'sNSX]

where are you located?

I live in Chester county Pa, about an hour west of Philadelphia...

 

[Old Guy]

Do you notice this more when you are in the car? In particular if you go outside and stand by the car the noise / roughness is not particularly noticeable and if you lift the engine cover and look at the engine it is not obviously rough or shaking? The reason I ask is that my 2000 NSX exhibits a similar phenomena. When it is quite cold and you are sitting in the drivers seat at idle it almost reminds me of a failing big end bearing. My car has done this since I purchased it - close to 9 1/2 years. The colder the car is, the more noticeable it is and it always goes away once warmed up. It is much less noticeable with cold starts in the summer and not noticeable on hot starts. Now that night time temperatures here are dropping close to freezing, with my unheated, detached garage the noise is quite noticeable on start up in the morning after the car has 'cold soaked'.

I attribute the problem to the front engine mount. In cold weather the rubber probably stiffens up and does a much better job of transferring noise and vibration from the engine to the back wall of the passenger compartment which makes a nice sounding board. The other contributing factor may be the engines 90 deg cylinder disposition. On a V6 the 90 deg angle results in an uneven firing order. I believe Honda uses offset crank pin pairs (not sure since I have never dismantled a C32B and don't want to) to improve the 'eveness'; but, it does not eliminate the problem (edit - my recollection being that the offsets introduce some mechanical unbalance).

Sometimes the front engine mount fails and that can result in a lot of noise in the car. Also, if you go out and observe that the engine is visibly shaking, then you do have an abnormality. Otherwise, that is just the way they are.

 

[goldNSX]

Did you DIY the TB/WP?

Rough idle can have several causes
- spark plug not firing: how do the rear coils look like? You could inspect the spark plug after a minute of rough idle and see if any are wet. While there check compression.
- a faulty sensor: engine coolant temp for the ECU
- an air leak somewhere: I just mention it here because I once had one but it was running lean in my case
- last but not least: TB off but I don't wan't to give you sleepless nights.
- valve adjustment
- injector spray pattern/seals
- bad fuel
Overall, a car not being driven can/will develop just kind of problems. Hope this helps so far. Good luck!

 

[Matt'sNSX]

Good information, all...

First of all, I wanted to say "Thanks" for all of the information, and I particularly like goldNSX's suggestions about pulling a plug or two while the engine's still cold. I apologize for the long response time here gents. For some reason I'm having a bear of a time with logging into NSX Prime and have to change my password nearly every time despite clicking the "remember me" icon.

As I mentioned previously, I had the timing belt and water pump replaced. In my paranoia fueled narrow focus I somehow, moronically, didn't think to change the plugs and adjust the valves while we were there (How stupid do I feel?!?). Anyway, would we experience an issue like this if we somehow missed a tooth or two? I like to think that there'd be a MIL if that were the case but who knows? Hopefully this is NOT the case and I'm just being paranoid once again.

As always, I appreciate any and all thoughts and input!

 

[Big McLargeHuge]

First of all, I wanted to say "Thanks" for all of the information, and I particularly like goldNSX's suggestions about pulling a plug or two while the engine's still cold. I apologize for the long response time here gents. For some reason I'm having a bear of a time with logging into NSX Prime and have to change my password nearly every time despite clicking the "remember me" icon.

As I mentioned previously, I had the timing belt and water pump replaced. In my paranoia fueled narrow focus I somehow, moronically, didn't think to change the plugs and adjust the valves while we were there (How stupid do I feel?!?). Anyway, would we experience an issue like this if we somehow missed a tooth or two? I like to think that there'd be a MIL if that were the case but who knows? Hopefully this is NOT the case and I'm just being paranoid once again.

As always, I appreciate any and all thoughts and input!

I've been off a tooth or two on my TB before and no noticeable misfires and never a MIL, just slightly more vibrations at idle really, so doubtful that's the main cause of your problem. If you're concerned the TB was installed incorrectly, you can pull off the valve covers as part of your diagnosis process to check the marks on the 4 cam pulleys vs. the crank TDC mark. That would also be a good time to do the valve adjustment/plug replacement since they should be done with a TB job.

 

[Old Guy]

First of all, I wanted to say "Thanks" for all of the information, and I particularly like goldNSX's suggestions about pulling a plug or two while the engine's still cold. I apologize for the long response time here gents. For some reason I'm having a bear of a time with logging into NSX Prime and have to change my password nearly every time despite clicking the "remember me" icon.

As I mentioned previously, I had the timing belt and water pump replaced. In my paranoia fueled narrow focus I somehow, moronically, didn't think to change the plugs and adjust the valves while we were there (How stupid do I feel?!?). Anyway, would we experience an issue like this if we somehow missed a tooth or two? I like to think that there'd be a MIL if that were the case but who knows? Hopefully this is NOT the case and I'm just being paranoid once again.

As always, I appreciate any and all thoughts and input!

There is obviously no code for a valve timing error. In order for a valve timing error to cause a MIL it has to trigger some other problem such as a fuel mixture error or a misfire. You have to be off by more than one tooth to trigger a problem which would trigger a MIL. The only way to confirm / rule out a timing belt error is McLargeHuge's suggestion to remove the valve covers and check the timing marks. Incorrect valve clearances can contribute to the problem. Checking the valve clearances isn't so bad. Adjusting the valve clearances, particularly the exhaust clearances is a bit of a root canal (actually the root canal is faster and not as hard on your back:smile.

At the risk of beating my own drum, if your engine runs smoothly once it is up to operating temperature chances are things are normal. Timing belt errors don't disappear when the engine heats up; but, excessive valve clearance can become less excessive making the problem less severe. However, excessive valve clearances will typically result in a noisy engine, particularly when it is cold.

Since you have a 1996 it should be OBDII. If you have access to an OBDII scanner you should be able to monitor your long and short term fuel trims. If the long term trims are small and your short term trims are small at idle you don't have any fuel mixture problems which, like checking your blood pressure, suggests; but does not absolutely confirm that things are OK. Removing and replacing the sparkplugs would allow you to check the color of the tips on the old ones which would provide some confirmation of the presence or absence of a fuel mix problem caused by a big valve timing error. While the coils are removed you can check for tracking and corrosion on the coils (common on the back cylinders).

If you remain uncomfortable, the only answer is exploratory surgery - valve cover removal and inspection of the timing marks. A valve clearance check at that time would be appropriate. If you are not doing this yourself you should find an NSX specialist who will take the time (and your money) to do it well. If the valve cover gaskets were not replaced when you did the timing belt you should probably plan to have new gaskets on hand for the job.