2005 Honda Accord 4cyl revving idle when warm

Question

My 2005 Honda Accord I4 is idling fine when cold but once warm it begins "revving" on its own between 1200-2000RPM at about 0.5Hz. The vehicle initially threw a DTC P0507 ("Idle Air Control System RPM Higher Than Expected") within a few minutes of starting but after disconnecting the battery (see troubleshooting below) I haven't been able to reproduce it.

Symptoms:

• At cold start idle is a little higher than normal, 2000RPM (usually 1500?)
• When warm, and vehicle is in PARK, idle revs between 1200-2000RPM at about 0.5Hz.
• Holding the accelerator when vehicle is in PARK ceases revving, but then it begins again when accelerator is released.
• Placing vehicle in DRIVE ceases revving, but idle is higher than normal, 1200RPM (usually 700?)
• Interesting twist: If I disconnect the Throttle Position Sensor (TPS) while its revving the frequency of revving increases to about 1Hz.

Background:

I had a different idle problem two months ago that I never got to root cause on, see writeup here: 2005 Honda Accord 4cyl low idle and stalling when warm.

Troubleshooting steps performed:

1. Suspected poor air-flow. I removed and inspected the air filter. Air filter was fine. Started car with air intake open (air filter removed) and no change in behavior.
2. Suspected bad Mass Airflow Sensor. Disconnected MAF electrical, restarted motor. No change in behavior.
3. Suspected engine controller needed reset. Disconnected battery, reconnected, held motor at 3000RPM for a minute. No change in behavior.
4. Suspected coolant system might have air in it. I removed the cap to the coolant reservoir and lifted the tube so I could observe coolant overflowing into the reservoir. The coolant reservoir was about 1/4 below MAX line when the engine was warm. No change in behavior.
5. Suspected battery might be low. Battery is dated 6/2011. Drove car around for 10-15 minutes between 35-45MPH, no change in behavior.
6. Suspected vacuum leak somewhere in air intake system. Sprayed throttle cleaner over a few connection points that I could identify as air hoses, no obvious change in behavior (but hard to tell when engine is already revving on its own). (Update: Learned later I did this incorrectly, you have to disconnect the TPS or else the computer may adjust automatically. See How to Find a Vacuum Leak.)
7. Suspected dirty throttle body or dirty Idle Air Control Valve (IAC). I removed the throttle body and IAC and thoroughly cleaned with throttle body cleaner. The interior of the IAC caked in grime; took me a half hour, a rag and a handful of q-tips to get clean. Reassembled, no change in behavior.
8. Suspected bad alternator / electrical. Checked voltage across battery terminals, 14.5V.
9. Per @Paulster2 below, suspected faulty IAC. Disconnected IAC while engine was revving. Revving continued, so it appears unlikely its the IAC.
10. Per @Zaid below, suspected vacuum leak again. Disconnected MAF electrical, started engine and waited for revving. Plugged air intake with heavy rag. Revving ceased but still idled high. Removed rag, revving resumed. Covered air intake with heavy-duty plastic bag. Revving ceased and idle sounded more normal.
11. While engine revving, removed electrical connection to Manifold Absolute Pressure (MAP) sensor. Engine sputtered momentarily, then resumed. Revving gone. Idle in PARK a little high (1100) but idle in DRIVE normal (700).

Any advice on what to try next would be greatly appreciated. Thanks!

Resolved:

It was the Idle Air Control (IAC) valve. Thanks everyone for your help!

Convinced it was a vacuum leak, I took it to a mechanic. He tested for leaks and found no evidence of a leak. He performed troubleshooting and his reference material lead him to the conclusion it was the IAC.

On vacuum leaks: The repeated oscillation of the idle changing was inconsistent with a vacuum leak, according to him. He said if it was a vacuum leak then the idle would be more consistently high or low and I probably would have gotten a different DTC.

On cleaning the IAC: He thought it was foolish of me to attempt cleaning the IAC. His assertion was that carb cleaner could find its way into the actuator and permanently damage it.

Answer 1

My suggestion is that your IAC valve is sticking. This would mean it is jutting in/out and not providing the amount of air for the computer to get the engine to settle. Since you've already cleaned it and the passageways, about your only other recourse is replacement.

Answer 2

You need to check for leaks in the air intake

A leaky air intake can explain hunting idle and I don't think spraying carb cleaner is always conclusive as some cracks can be really hard to see/detect.

There are several ways to check this:

• block the air intake completely

  The easiest way to do this is disconnect the airbox and cover the piping before the MAF. If the engine doesn't die, it's drawing air in from an alternative source

• oil dipstick test

  With the engine running at idle, remove the dipstick. If there is no change in engine RPM or idle, the engine is drawing air from an unmetered location

• monitor fuel trims

  Specifically the long-term fuel-trims (LTFT's). If they are positive and high, chances are your car is running with unmetered air.

  Note that unmetered air could also be caused by an underreading MAF. You'll also have to confirm that insufficient fuel supply is not an issue.

• smoke test

  Seal the intake plumbing and introduce smoke to see if it billows out from any holes or cracks. If smoke can be seen, that is where the unmetered air is being pulled from.