View source for Rough Idle

Found this good article in the explorerforums on troubleshooting a rough idle. Including it here in case it is removed:

I just wanted to share my experience in correcting a rough idle issue because it was particularly frustrating and I got quite an education in the process. Lots of information here. If you have the same problem I had, hopefully there will be something useful to you. This is on an '05 Aviator with the 4.6 DOHC engine but I imagine most Ford engines will be similar.

I noticed that after a transmission rebuild my engine seemed to idle pretty rough. It ran at higher RPM and load fine, and seemed to get normal mileage. I don't know if it ran poorly since I bought it or if I just noticed after the transmission was quieted down. Seems like sometimes you fix one problem only to expose the next weakest link.

First off, if you are dealing with an older high mileage engine it's very likely that you have multiple issues to deal with. So if it is a high mileage engine do yourself a favor and do the basic maintenance items first. New plugs, and PCV just to rule those out. Aviator has COP so no wires to worry about.

To do this properly, you are going to have to invest in scanner software. Without it you are just shooting in the dark on a lot of issues. I bought a USB OSB2 adapter on Amazon for $8, and have ScanXL software. I think it costs $79. That's not cheap, but it's cheaper than the cost of replacing one component unnecessarily. And it's a lot of fun to play with!

Regarding plugs, for the most part spark is spark. Some will say that you need to use OEM plugs or there can be odd problems. I think that's about 99% not true. But in the end, if spark is spark, just use the OEM and don't confuse the issue.

PCV valve test. I found a great test for PCV valves.
# Remove valve.
# Throw it in the trash
# If it bounces out and re-installs itself, it's good, otherwise buy a new one.

The rattle test on a PCV is only a fail test, not a pass. If it doesn't rattle its almost certainly bad. But if it does rattle, that doesn't necessarily mean it's good. To really test if it is effecting idle smoothness, just pinch off the hose and see if the idle changes significantly. If there is no change in idle, then chances are it's not a factor. Something I read on the web is that a leak in the oil filler cap or dipstick can cause a rough idle. This is not true if the PCV valve is working correctly. It has a controlled metered flow of air so even if the crankcase is wide open it's not going to have any noticeable effect. I tested this by removing my filler cap and dipstick and it had absolutely no effect. Don't forget to check the breather tube running from the air inlet to the valve cover.

On the Aviator, the EGR valve is a “module”. It has the DPFE sensor and MAP sensor integrated into it. If the EGR is old and dirty, buy a new one and don't even frustrate yourself trying to troubleshoot it unless you really know what you are doing. And only use OEM. I paid about $70 on Amazon for an OEM EGR module.

Fuel pressure. Highly unlikely a fuel delivery problem will create a rough idle, but don't rule it out. Use your scanner software or pressure gage and verify that the pressure is in range. '05 Aviator does not have a fuel pressure port so you need to use scanner data.

Vacuum leaks. Most rough idle issues are caused by vacuum leaks. So go through the engine carefully and use whatever means necessary to find the leaks. Smoke testing is not going to find many leaks because it applies a positive pressure to the manifold. Vacuum may only leak in one direction. And vacuum can be as high as 14psi whereas smoke testing only goes to 2-3psi. I had zero luck using propane to find vacuum leaks. I had some success spraying carb cleaner, but you need to use a scanner and monitor the fuel trims and O2 sensors to see it. The computer will compensate for the carb cleaner better than you can listen for it. The water spray test can be effective on some engines if you can direct the water where you need to. On the Aviator, the lower intake is so recessed down into the block that its almost impossible to check for leaks.

On the Aviator it's pretty simple to check for vacuum leaks in hoses. There's only 4 hoses connecting to the intake, so you can systematically pinch them off and verify for changes in idle or manifold pressure.

But I did find another very good way to verify if there is a vacuum leak. Start and warm up the engine. Turn off all the accessories. After the engine is completely warm and idling at it's lowest RPM (About 600RPM) unplug the IAC valve. If the engine RPM stays the same or drops, you can be fairly sure you don't have any significant vacuum leak. This is much better done with the scanner software. The IAC count should be between 21 and 40 at base idle. On Ford's, 20% on the IAC PID is actually fully closed. I think the reason for the low end being 20% is that even with the valve closed, there is a small controlled air flow that is roughly 20% of full open. Try and wait for the PCM to adjust the IAC to 21% or as low as possible before unplugging it and watching for the RPM drop.

The Ford IAC is a fairly robust device. It uses a variable pulse duty cycle to modulate how far the solenoid opens the valve. It seems like it is pretty much good or bad. Probably rare that they fail half way. Again, you can use scanner software to verify it's functioning. If you see the IAC counts changing in direct relation to the RPM changing, it's probably working correctly. If you unplug the valve, it will fully close from engine vacuum. You can verify it is shut by pinching off the bypass hose going to the manifold. There is a small controlled flow even closed so you should see a very small drop in RPM when you pinch off the bypass hose. If you see a big drop or engine stalls, chances are the valve is very dirty and needs replacing. If you do need to replace the valve, only use OEM not aftermarket. I used a Hitachi, but from what I understand, the Hitachi IS the OEM. It was half the price of OEM from Ford.

Intake manifold leaks. I went through hell on this one. Fortunately the Aviator is really simple to work on. I suspected I had an intake leak because of the way the engine ran. I used a Felpro gasket and installed the way they recommended, without any sealer, just clean and dry. The engine seemed to run much better for a day then started getting worse. I eventually suspected the gasket was leaking again so resealed it with a thin layer of RTV. That lasted a couple hours and went bad again. Finally, I went and got the OEM Ford gasket. Turns out the OEM gasket and Felpro are identical, just a different package. (The Victor Reinz is also the exact same gasket so just buy the cheapest one) I reinstalled with Edelbrock Gasgacinch with as much care as possible. I was absolutely certain it was not leaking when done. To this day, I'm still not sure I ever had an intake leak.

While I had the intake off, I thoroughly cleaned it. I also used brushes and cleaned out the intake runners. They weren't all that bad but the deposits were like molasses so I doubt any fuel treatment would have touched it. Very helpful to have a camera you can get in there when doing this. Mine were pretty gummed up but even so I don't think it had any real effect on the way it ran. If you have a camera you can take a peak at the valve condition and see if they are burned. The Aviator and Mustang heads have a problem with the rear valves getting burned from poor coolant flow. Fortunately mine were fine.

Throttle body. Unless its super dirty, not likely it's causing an issue. But its easy to clean so just do it. Mine had leaking shaft bores so I disassembled and replaced the orings. It wasn't so difficult. I just used plain old carb and choke cleaner and not special throttle body cleaner.

Fuel injectors. It's a PITA to remove the injector rail on this engine with the intake attached. I'd suggest removing the entire lower intake with fuel rail attached. It's way safer and you can take your time and get everything clean. Mine had a ton of dirt packed around the injector orings. There's no way I would get this done cleanly in the engine. I tested my injectors with a syringe full of carb cleaner and a 9v battery. There's lots of good videos on Youtube of how to do this. If they squirt cleanly, they are good. If they don't, they are bad. Not much grey area with these injectors. I also removed the brass filters and cleaned them just as a precaution. Might as well do the resistance test on them and verify they are all the same even though the resistance test is pretty unreliable.

MAF sensor. Don't forget to clean the sensor. It only takes 5 minutes. (Going to need a security torx bit to remove it). You don't need expensive MAF cleaner for this. Just use plain carb clearner. You can use your scanner software to verify it is functioning linearly. If the flow increases fairly steady with RPM's, then it's pretty safe to assume it's still good.

Intake air leaks. Make sure everything is tight on the intake air tube and filter housing post MAF sensor. Inspect carefully for any cracks or potential leaks. I like to use silicone grease where the snorkel attaches to the throttle body. It makes it slide on easy and helps to create a good seal. Even a small leak between the MAF sensor and throttle body is unmetered air and will cause almost as much trouble as a vacuum leak.

Fuel Trims. This is where you will need scanner software. The “spec” for fuel trims is +/-20% combined long and short term. No codes will be thrown until you get to that threshold. At 20% your engine will be running like shit, maybe not running at all. I think they set the spec so high so they don't have to deal with any more warranty claims than necessary. Experts will say that if your trims are anywhere near 10%, you likely have something wrong. A well running engine will generally be within +/-3%. When I started my long terms were up around 9-10% at idle. When finished everything is under 2% most of the time. It's too complex to get into here, but you can use these fuel trim numbers to help troubleshoot your problem. For example, if the trims are high at idle and drop to normal at higher RPM you probably have a vacuum leak or some other source of un-metered air. An IAC valve that's not behaving correctly can simulate a vacuum leak. Bad valves and rings can also simulate this issue so verify your engine is sound before spending too much time tuning it.

O2 Sensors. Again this is where you MUST have scanner software. The upstream O2 sensors are most important. That is how the PCM decides if the air/fuel mixture is correct. Even if you don't have an outright failed sensor, they do gradually go out of spec. If you have over 100k on them and are experiencing drivabilty problems, it's probably best to just replace them because they'll probably fail soon anyway. I don't know about the Explorer, but it was a 10 minute job on the Aviator and only cost $70 for the pair from Amazon. The upstream sensors need to switch back and forth between .9v and .1v. If you don't see this rapid switching on the scanner software, or if they are switching at different speeds, then chances are they are bad. If they are reading steady ov or 1v, don't assume they are bad. Go through the correct troubleshoot procedure to rule out wiring issues. The downstream sensors won't effect how the engine runs. They only monitor that the cats are doing their job. But it's an important function nonetheless. If you have a malfunctioning cat it's best to know soon so something else isn't damaged.

COP's. The Aviator is particularly prone to problems with the coil on plugs. So because of this, I think they get blamed far more than they should. I read that 75% of the COP's returned to Ford under warranty were actually good. For the most part they either work or don't work. Intermittent failures are fairly rare. You can test by purchasing a new COP and substituting it into the engine cylinder by cylinder until it runs correctly. Or, you can do a cylinder drop test and see if there is a noticeable change. On my engine, I did the drop test and couldn't detect any change in engine smoothness when any of them were pulled. Probably that was because I had some other issue masking it. Best way to check spark is with the adjustable spark tester. If they all can jump a 30kv gap, you can be pretty confident that is not the issue. If it is only intermittent then you will need to substitute COPs one by one to find the problem. Should perform a resistance check on each of them and verify they are all the same resistance.

NOW THE MOST IMPORTANT THING! This was the missing link that finally made my engine run the way it should. It was night and day and I wouldn't have believed it if I didn't see for myself. The Aviator engine (Maybe all Ford's) is extremely picky about the reset and learning procedure. You must follow the correct procedure or you will likely have lingering issues. Another forum member (A Ford tech I think) posted a video on this a while back. I think this procedure should be done on a cold or cool engine if possible.

# Disconnect battery
# Jumper the positive terminal to negative cable for 10-15 minutes to fully reset the computer.
# Reconnect battery
# Turn off all accessories and headlights
# Turn ignition on
# Wait at least 10 seconds
# Start engine. DO NOT TOUCH ACCLERATOR UNTIL STEP 11!
# Let engine run and completely warm up at least 10-15 minutes so that it can map all the idle parameters. The R's may go up and down a little during this time while it's dialing everything in.
# Turn on headlights and let run 1 minute
# Turn on AC and idle 1 minute
# Take car for a test drive but be rather gentle with it. Come to full and prolonged stops and let the idle stabilize at each stop. (Don't hinder traffic doing this though).

If you do any live troubleshooting of the engine after the reset procedure such as unplugging the IAC valve, vacuum hoses, swapping plugs, etc, the PCM will learn some bad habits again and you will likely have to redo the reset procedure. Probably it will eventually relearn correctly but I think it takes a long time. I found this out the hard way. I finally figured out that this is probably the reason my engine ran poorly after the transmission rebuilt. The battery was disconnected for a couple days and the computer was reset. No doubt the tranny shop just fired it back up and drove it out cold without following the reset/learning procedure. The PCM is factory calibrated for a NEW engine. So when new, it does not need to learn anything when reset, as everything has already been optimized. But an older engine has different performance characteristics such as lower vacuum, lower cylinder pressure, more exhaust restriction, etc. Seems that this PCM does a rather poor job of learning to make an older engine run smoothly.

EDIT:
Wanted to add a little more to this in case somebody actually reads it. I thought my idle problem was cured, but after a few days it would get worse again requiring another reset. Turns out the throttle body adjustment had been tampered with, or maybe came loose on it's own. I adjusted the base idle speed to 400-500RPM. Basically this is with the throttle plate completely closed and then back off the throttle screw a half turn so the throttle plate doesn't contact the throttle body bore. This made an amazing difference in how the engine performs. Almost completely cured the rough and erratic idle, and I picked up about 25% city fuel economy.

Here is a great video on how to properly adjust this. I would not suggest adjusting this screw unless you are sure you know what you are doing. I imagine it is factory set to a specific airflow and normally should never be touched. Regarding the TPS voltage, they are usually just under 1v from the factory, so if it's more or less that's a good indication it's been tampered with. Mine was hovering right at .99v, so I'm not sure it was changed, or maybe just with wear the airflow changed. My final setting was about .95v. The computer does not care what the voltage is, as it just reads the fully closed voltage and assumes that is zero.

https://www.youtube.com/watch?v=ayjhL9jpFH8