1996 3.8 intake manifold runner control stuck open

P1537- intake manifold runner control stuck open. Can this in any way hurt my van-(make it run to lean to rich) can I fix It by running a can of carb medic thru the intake.

The default position of the intake manifold runners is OPEN. They are open when the car is not running. When you start the car, vacuum is applied to the actuators, and they close.
Above a certain RPM they open.
There are 2 of them. One for the front bank of cylinders, and one for the rear bank of cylinders.

With the car OFF, you can try to move each one....and see if it moves.

The actuators can go bad. On my '96, the actuators get their vaccum source through red lines that connect to the top of each actuator.

Here is a link to some pictures of my '96....that may help you to locate the actuators...as well as a few other things. You can click on "view full size" to blow the pictures up to see more detail.

http://community.webshots.com/album/201931518cScpNK

I have the same code from autozone's scanner. I know where the runner bank is located but am not sure what the actuators are or what they do or if I can fix the problem. Any help is quite appreciated. My engine is running too rich right now as well, although it may be an oxygen sensor, although the code didn't come up. After cleaning my IAC valve the smell has been cut in half. Could this runner stuck open be the problem with my high CO levels? Everything else is within normal. Thanks in advnace,

Jennifer

There are 2 actuators.....1 for the front bank....and the other for the rear bank.
The actuators are made up of 2 parts. One is the vaccum "diaphram"....which, when vaccum is applied...CLOSES the runner. If the runner is showing that it is stuck open....you will want to verify that the vaccum is getting to the actuator.....(which it will be at IDLE).

The other part is a electrical resistance which will indicate if the runner is OPEN....or CLOSED. If the actuator pulls the runner part way closed.....then you might be able to clear the problem up by cleaning / lubricating the linkage. You can see that there are stops.....that a flange on the end of the runner shaft (right where it goes into the lower intake manifold) will move from stop to stop. If it hangs up.....you will need to make sure that it can move freely from stop to stop. Crud can also build up on the plates inside the lower intake manifold....preventing full motion of the runner. It is not uncommon for the actuator to go bad. And they are a little pricey.....which is why I am suggesting that you make sure that the mechanism can move freely from stop to stop.

With the engine OFF......the runners will be OPEN......and should be against the fully open stop.

With the engine at idle.....the runners will be CLOSED.....and should be fully against the fully closed stop.

If BOTH of the above conditions are not true....then you will need to correct the problem.
If they are not both against the fully closed stop at idle....you might CAREFULLY (As you are next to the drive belt) try to gently help it along.

Refer to the pictures in the link that I posted above......and...hopefully my ramblings make some sense.

I understand what you are saying, thanks and well I will ask someone a little more brave than I to put his hand in there and try to un stick it, lol.

The van idles *really* lousy, too, doesn't it?

IIRC, the intake manifold runner controls are fairly expensive ($200-range), but they are fairly straightforward to troubleshoot. When you finally find someone who's willing to stick their hand in there :smile:, they will quickly be able to tell if the vacuum servo's diaphragm is bad if they simply draw a vacuum on each of the two vacuum ports using a piece of small-diameter vacuum line, and see if each unit toggles the intake manifold runner. If both manifold runner controls toggle, then you likely have a vacuum leak somewhere upstream. If one of the units doesn't toggle (which is more likely), then you probably have a bad Intake Manifold Runner Control (or possibly a "stuck" runner control shaft, but that's doubtful). Regardless, it'll be pretty obvious.

Heck: Another option...have someone watch the IMRCs while you start the engine. If one servos closed but the other one just sits there, you've probably just IDd where you want to concentrate your attention.

One thing about those controls: They don't even open until the engine is turning over ?2,900-3000? RPM (IIRC...may be as high as 4,000 RPM?). If you rarely (if ever) zoom past 3500 RPM, you could as a temporary fix, wire the darned things closed...your idle will be much better, it'll drive "just fine", but as a tradeoff, won't have as much "zip" at the top end.

I had this same problem. The way to tell what the problem is to pull off the vacuum hoses. Have someone start the car while you put your fingers over the ends of the hoses. if you start feel the vacuum you know its not a vacuum or electrical problem.

In my case I felt the vacuum so I figured the actuators were bad. Now you take a screwdriver and press down on the linkage if they move easily you know it’s the actuators that are bad. Only one of mine was bad but both fail to operate I guess because they run together in vacuum system.

Now to tell witch on if not both are bad with the vacuum hose off push down the linkage to fully closed put your finger over the vacuum inlet and let off the screwdriver if it stays closed you know it works properly if it opens again you know there a leak in the actuator diahphram.

One thing about those controls: They don't even open until the engine is turning over ?2,900-3000? RPM (IIRC...may be as high as 4,000 RPM?). If you rarely (if ever) zoom past 3500 RPM, you could as a temporary fix, wire the darned things closed...your idle will be much better, it'll drive "just fine", but as a tradeoff, won't have as much "zip" at the top end.

OK. I am having this P1537/P1538 issue on my '98 Windstar w/ 81K miles. Following Bittmann's advice, I wire the two IMRC valves closed. Now the CEL is gone and I have driven the car a few hundred miles and it seems fine. I have NEVER exceeded 3000 RPM on this V6 3.8L engine so I guess it's fine.

I hope there is no side effects. Any comments?

IIRC, the intake manifold runner controls are fairly expensive ($200-range),


These units are listed on E-Bay for $60 for the RH & $55 for the LH. Says they fit '96 thru '98 models. What happened in '99? Just curious as mine don't seem to be having a problem, but would like to be in the know if the situation comes up.
http://cgi.ebay.com/ebaymotors/RS-INTAKE-MANIFOLD-CONTROL-VALVE-FORD-TRUCKS-3-8-4-2L_W0QQcmdZViewItemQQcategoryZ33557QQihZ005QQitemZ 150088876232QQrdZ1QQsspagenameZWDVW

These units are listed on E-Bay for $60 for the RH & $55 for the LH. Says they fit '96 thru '98 models. What happened in '99? Just curious as mine don't seem to be having a problem, but would like to be in the know if the situation comes up.
http://cgi.ebay.com/ebaymotors/RS-INTAKE-MANIFOLD-CONTROL-VALVE-FORD-TRUCKS-3-8-4-2L_W0QQcmdZViewItemQQcategoryZ33557QQihZ005QQitemZ 150088876232QQrdZ1QQsspagenameZWDVW

in '99 Ford wisely went to electrical actuators for the IMRC's.

As to the ones on Ebay: judging by the picture they appear identical to the Dorman valves that rockauto was selling. At least two of us have bought these from rockauto only to find out they did not work with our vans (mine is a '96). They worked properly from a vacuum standpoint, but did not send proper position information to the PCM resulting in continued "IMRC stuck open" codes. I returned mine to rockauto and have been living without. I wired closed the IMRC, and I have the old valve plugged in with the actuator wire-tied in the closed position to keep the light off.

I have a 2000 Windstar with the electronic actuators and am wondering if the open with engine off, closed with engine on scenario is still true. Mine don't move when the engine starts but seem to want to.

Thanx

I have a 2000 Windstar with the electronic actuators and am wondering if the open with engine off, closed with engine on scenario is still true. Mine don't move when the engine starts but seem to want to.

Thanx

My experience (limited) with the '99 and '00's is that the actuator stays "fixed" under normal ops, i.e. closed regardless of engine running or not. With your engine running, if you grab the throttle and push it toward full open quickly and let go - rev it, you should see the IMRC's move briefly toward open, and then closed again. In the life of your van, those two actuators should likely not move much unless you routinely floor it.

Just as a heads up on Ford Dealerships. I have a 96 Windstar. I had both 1537 and 1538 codes. I found the best price for them at NAPA, $104.00 each. I found out that Ford sold them for $162.00. The codes cleared and I still had two other codes that lead me to believe I had a vacuum leak. I took it to a Ford dealership and they found the problem. Two days after I got the car back my 1537 and 1538 codes came back. I took it back to Ford and told them that both IMRC's were only 3 weeks old. They did some extensive testing and told me that both of the IMRC's were bad. I told them to go ahead and replace them and that I wanted the parts back. They charged me $966.00 for the repair!!!!! When I went to pay the bill, I noticed that they charged me $ 268.00 each for the IMRC's. I mentioned that that was a rip off since I could get the parts from another dealership at a better price. They said they could remove the parts and I could go there. When I got home I called the manager of the dealership and told him what I thought of the rip off. He had me contact the parts manager. I told the parts manager and he said he would honor the other dealerships price. When I went in to get reimbursed for the difference, he mentioned that the price I found was probably an internet price. when i got home I called the other dealership that said the part was $162.00 on line. I talked to a parts person and he said that the part was $268.00. So the online price was a better deal. If you can't replace the parts yourself, I would order the parts online and take them to someone that can. P.S. I did get my money back from NAPA too.

This is great info. Thanks. I wonder if anyone can help me with a slightly different senario? My 3.8 96' has the CEL on and the code came up 1538. The mechanic (non Ford) told me it meant IMRC Bank 1 was bad. Ford however says they don't go by bank 1 or 2 but that I needed to know if it was the IMRC nearest the radiator or the one nearest the firewall (firewall side part number F65Z9S514DA) or the radiator side ( number F65Z9S514CC). Does anyone know which one that would be for Bank 1? I looked at wiswind's photos but didn't find any identification for the individual IMRCs. Photos where great however!!

Also I followed the advice above and discovered that with the engine OFF or ON at IDLE both IMRCs remain open and do not close. I have tested and checked the actuator arms and butterflies. They are all lubed and work easily. Taking the vaccum lines off (they are 2 white tubes that come from a Y vaccum line) I don't feel hardly any vacume and next to the front IMRC is the red vaccum line going to the fuel pressure regulator and it has great vaccum. The question.... are the IMRC vaccum lines vaccum source, controlled by the electrical IMRC plugs or by an independent other source? Because I am wondering if I have a vaccum leak (I can't see where the vaccum lines go as they disappear behind the rear injector/maniford) which could be why neither IMRC units close at idle. Or is the vaccum controlled by the IMRC electrical inputs and thus the IMRC's are both bad? I did do the test where you attach a vaccum line to each IMRC and suck and they both close easily.

I almost went and replaced the one for $268.00 but now not sure if that will fix my CEL code. Any suggestions or advice as to the solution and more troubleshooting is greatly appreciated. In the intermin I may wired them closed as suggested but ultimately I want to fix them correctly.

Thanks for your help.
Mack
mackjazz@hotmail.com

... someone has already posted on this forum (long ago) which part number went where. I don't remember much more. Perhaps the poster is still with us. ??

My guess would be that the one closest to the radiator reaches back and actuates Bank #1.

Hi
Welcome to the club. First, 1538 is IMRC stuck open on bank 2 (your front bank with cylinders 4, 5, 6. Ford also calls them left and right. I think the right is the rear and left front as the engine actually "faces" the right fender- hopefully others will correct me if that is backwards. If you are going to do much DIY work, I think a hand vacuum pump (mitivac) available at sears, autozone and similar places is a great addition. In this case, you could use it to measure the vacuum on a given line, produce vacuum "pressure" to test your IMRC's and help track the problem. If you are getting no vacuum to either IMRC at idle, you MAY have a problem with the IMRC vacuum solenoid. The vacuum supply goes via the solenoid which is contolled by the PCM to dole out vacuum to the IMRC's. It is located on the back of the upper intake plenum and is nearly impossible to see without the cowl off. I did have to change mine a while back and it is not a terribly expensive part, but is a dealer part and not widely stocked. If you have vacuum and power to the solenoid, and yet no vacuum to the IMRC's while at ide, that is likely your problem - assuming the IMRC's are OK (again, a vacuum test is the best way - BTW you could swap your fuel pressure regulator vacuum to test them as well -as I recall the van idles fine without the regulator hooked up. The IMRC's should receive vacuum at all times below about 3000 RPM, then the PCM signals the solenoid to close (no vacuum) and the IMRC's spring to the open position.

The path for vacuum runs from the rear plenum to the vacuum tank on the right wheel well behind the radiator expansion/fill tank, then back to the solenoid and on to the IMRC's.

Let us know what you do/find.

Best of luck - these things are notorious for failure. Many of us have manually locked them closed and live with the recurring light as they are expensive to replace, and the OEM ones are the only ones that work.



Let us know what you find.

Yup! Busboy's right ... and alert! 1538 is for Bank #2, not #1.

So I now guess it's the one toward the firewall that reaches to the front unit. Are you sure of the number, or the bank?

So I now guess it's the one toward the firewall that reaches to the front unit. Are you sure of the number, or the bank?

Hi 12ounce,

I'm not sure what you are getting at. Could you help me/us out?

Thanks

Thanks for the Welcome and responses. The mechanic wrote and told me IMRC bank 1. Later I asked about the code and thought I remembered 1538. Last night I wired them both closed.

Today I am going over to Autozone to have the OBD redone and this time I will note for myself which code it is. It was mentioned you live with them closed and the light on. I thought I read where if you wire them closed the CEL would go off and stay off. I ask because this whole issue came up when I tried to pass CA smog and although the van passed the emissions I can't be given a certificate due to the CEL being on. I can't re-register it or sell it until I get that darn CEL light off. Is it not true that it goes off (after sufficient cycles) with the IMRC's wired closed? As I said before the FPR gets great vaccum but neither of the IMRC's do. Attaching a vaccum gage shows little vaccum yet with a hose manually I can easily close either one of them with my mouth. So at idle they do not close.

All vaccum hoses are attached in the back (near firewall) and others appear to get sufficient vaccum which is why I asked if the vaccum failure could be the result of the electrical resistance connection to the IMRC's? Each IMRC has a 3 point electrical plug that is connected to the IMRC. Does anyone know what they do, are, or how they work in conjunction to the IMRC's themselves? Could these electrical parts of the IMRC's be the cause of little or no vaccum? I don't mind replacing the IMRC's if that will fix my CEL problem and they will both close at idle. But if I replace them and they don't close because there is still little vaccum to them then I would still need to address the vaccum issue. Any input on that? Does one IMRC affect the other, either vaccum wise or electrically? Where is the IMRC Solenoid located exactly? Is the cowl part of the car body near the wipers? How did you get to the IMRC Solenoid to replace it? Does it have its own fault code if I does not work?

Thanks for the great help and input.

Mack

Hi Mack,

First - you are essentially correct. If you can wire the IMRC closed and leave it connected electrically, it should sense a closed valve and clear your light. Be aware that by forcibly closing the vacuum valve itself that you may damage/destroy it internally. I say so because I have done so. My front valve worked with a vacuum pump, but was not getting vacuum. I tried to wire tie it closed which is somewhat difficult actually due to the limited space. I ultimately succeeded, but also discovered that I damaged the valve, and it no longer responded to vacuum as supplied by my hand pump. So bear that in mind.
As to the IMRC electrical: it is passive as to the operation of the valve - the valve only responds to vacuum on/off. The electrical provides a signal to the PCM as to the position of the valve. I believe you have ground, power and then the sensor line to the PCM which I believe gets a variable voltage based on valve position.
If you have the time I think you can run this problem down. As both of your valves respond to vacuum you either have a vacuum leak/broken line or the IMRC vacuum solenoid has failed. The IMRC vacuum solenoid is attached to the back of the upper intake on the rear horizontal "tube". I'll not kid you, it is hard to get at, but you should be able to feel it easily enough. As I recall, working from right to left (as you face the engine compartment) you should see/feel the vacuum hose attachment point(s), then the IMRC valve, and then the EGR vacuum solenoid. I mentioned earlier that Ford doesn't much keep the solenoid on the shelf. My local dealer, took a couple of days to get it. I think the part number is F1HZ9H465A. It is more than I remember. List is/was about $53. I bought it at fordpartsonline.com for $39. If you choose to troubleshoot it at all, you should have battery voltage at the solenoid with the key in run. Then, you would check vacuum in - as I said earlier coming from the vacuum tank on the right wheel well, then through the valve to the IMRC's. If you have good vacuum at the solenoid, and can apply vacuum to the hose out to the IMRC's such that they close I would vote the solenoid. If you can apply vacuum to the hose out at that point and the IMRC's do not close, then you have a leak in the line downstream.
You asked if one IMRC can affect the other: yes. If one is leaking it can cause both to stay open as their vacuum sources are connected.


Hope that all helps.

Mack, one other thing. I do have a photo of the upper manifold removed and it shows the valve, not in great detail but it is there. My sign on does not allow me to post attachments - not sure why. Anyhow, if that would be any help shoot me your email address either here or in a private message.

Busboy, its now no longer moot ... as you have made a good description of Mackjazz's possible solution.

But isn't it true on the pneumatic actuators: that the rear (RH) actuator is for bank #2; and conversely, the front (LH) actuator is for bank #1? I've never looked at the IMRC controls on anything earlier than a '99, but always understood that the linkages kinda "cross". ??

Busboy, its now no longer moot ... as you have made a good description of Mackjazz's possible solution.

But isn't it true on the pneumatic actuators: that the rear (RH) actuator is for bank #2; and conversely, the front (LH) actuator is for bank #1? I've never looked at the IMRC controls on anything earlier than a '99, but always understood that the linkages kinda "cross". ??

Hi

thanks for the reply. On the pneumatic actuators they control their respective bank i.e. front for front. The valves have about a 21/2" arm that moves the "rail" if you will controlling the bank's internal valves. I have looked at the later model IMRC's and do not clearly recall whether they "cross". I think that other than the actuators being electrical rather than pneumatic the construction is identical.

Thanks

Well I got the codes. Both P1537 & P1538 which is both runners (valve stuck open).

I wired them close late last night and am driving it to get it to cycle enough to see if the CEL will go off. I was careful to wire them both about 4/5th close. It was mentioned the butterflies can be broken or was it the valve diaphram inside the IMRC? Busboy 4 is correct in that the IMRC bodies are sort of back to back and their respective actuator arms travel away from the IMRC body, meaning one arm heads to the firewall and the other toward the radiator. There is no crossing on 96's. This photo shows them fairly well. http://rides.webshots.com/photo/1201995628011220610ESHmJI

I really am beginning to think everything would be okay if there was vacuum to the runners. There is nil currently. I will go look for the photo that was mentioned. I can feel around the back of the black air tube and even got a small mirror back there and can see the vacuum hose connections of which there are 2, one with a single line and the other with 3 or 4 hoses coming together at the point of attachment. Where in relation to these vacuum hose attachments would the IMRC vacuum solenoid be or is that what the vacuum tubes attach to? I have to check and make sure the vacuum hoses do not have cracks in them and as you said which my prelimnary shows them okay and check to make sure vacuum is getting to the valve solenoid. Again I can easily close the IMRC actuators with a hose and my mouth so I hate to purchase IMRC's and discover that is not really the issue. I assume when you say you have a photo of the valve you are referring to the IMRC vacuum solenoid, correct? I would like to see the pics and I did put my email address in my original post so here it is again. mackjazz@hotmail.com

I think the online parts place calls this solenoid part VLV ASY-THERM AIR CO. Does that sound correct? I couldn't find it by using IMRC but using your part number this is what came up?

Thanks for all the great help, wisdom and knowledge all are willing to share.

Mack

What you have will probably work for you. Another approach might be to attach the control vacuum hose directly to some prong on the vacuum manifold mounted on the firewall. There is usually a connecting point on the manifold that is not in use. This would "skip over" all the non-working vacuum controls.

Would this work? Might actuate at the wrong time ... not sure.

What you have will probably work for you. Another approach might be to attach the control vacuum hose directly to some prong on the vacuum manifold mounted on the firewall. There is usually a connecting point on the manifold that is not in use. This would "skip over" all the non-working vacuum controls.

Would this work? Might actuate at the wrong time ... not sure.

I have thought about that too. These things only need to be open at high rpm and many have wired them shut rather than fix them. So, I would think adding a T somewhere to a good vacuum source and then running lines to the respective IMRC's might just do it. As to actuating at the wrong time, the valves actuate, if you will, closed on start, and in almost all cases stay closed until you shutdown and the vacuum bleeds.

Good thought you had.

Hi for the persons who have wired their runner controls closed. I understand that your check engine light turned off. So if you hook up a scan tool it will show that their are no codes that everything is okay? i ask because i bought the dorman replacement and it still doesn't work. I don't run my van past 2k as far as the rpms go. I need this light off because of state inspection. Please advise as these parts are too expensive for me on a 14 year old vehicle. Thanks in advance for any info you can give me.

Bad news......I don't have personal experience.......super lucky to have mine last so long.
I remember a post or so on this forum that someone had the SAME problem with the Dorman replacement......seems that they are not ELECTRICALLY compatible and will cause the CEL and code.
It is a shame because the genuine ones are so expensive......but it seems that this is the way to go.

Hi, I would still like to hear from the person who wired them closed and said that there check engine light went off. I want to know if that would create a no codes stored situation when they connect the scan tool. Once again thank you for the info.

If you still have your original actuators.....you could try putting them back on......and wiring them closed.
Remember, the springs hold them open.....the vaccum pulls them closed.

The threads that I have seen where folks wired them closed were with the 1999 and newer....which use a different actuator.

Also, take note of the fact that the CEL will stay lit, with code(s) stored for a while AFTER you have corrected the problem.
It takes a certain period of time without a failure for the code to clear.

The default position of the intake manifold runners is OPEN. They are open when the car is not running. When you start the car, vacuum is applied to the actuators, and they close.
Above a certain RPM they open.
There are 2 of them. One for the front bank of cylinders, and one for the rear bank of cylinders.

With the car OFF, you can try to move each one....and see if it moves.

The actuators can go bad. On my '96, the actuators get their vaccum source through red lines that connect to the top of each actuator.

Here is a link to some pictures of my '96....that may help you to locate the actuators...as well as a few other things. You can click on "view full size" to blow the pictures up to see more detail.

http://community.webshots.com/album/201931518cScpNK
Once again this forum and it's wonderful members, like you, have helped me immensely! Knowing where to look and what to look for was invaluable. Found a missing clip that holds the arm into the lever that controls the valve. Should be fine now. THANKS!!!! Tom C.

After a long research

part # for irmc solenoid ford 3.8l

is e7tz-ph465a-a valve (as shown on the box)

Great thread, guys.

1998 Windstar, BTW.

So if I'm following the logic here.......I checked the vacuum and have a small amount of vacuum on the rear most IMRC valve (the one closest to the firewall), but can't seem to detect any on the front valve (the one closest to the front of the car). That would seem to indicate a vacuum line leak, or possibly a blockage, on that front valve line, correct? Since BOTH have to have vacuum for BOTH to work, I can probably just replace that one short piece of line and it should clear up, right? TIA for the help.

Jim

Took a break and decided to look at this again. The one thing I didn't try the first time around was to remove both vacuum lines at the same time. When I did that, I have good vacuum on both (by covering both lines with my fingers). Seems I have "pretty good" vacuum when I disconnect the back line, but almost none when I disconnect front line. I'm guessing that means the back IMRC has failed and the front one is probably on it's last leg? I was really hoping this would turn out to be a vacuum line problem. But it looks like I'll be spending money on new IMRC(s). Dang.

Jim

One poster, long ago .... may have been on another forum .... tried repairing the broken diaphragm with some success. I have no knowledge of what was required ... but it sounds like something I would try.

One poster, long ago .... may have been on another forum .... tried repairing the broken diaphragm with some success. I have no knowledge of what was required ... but it sounds like something I would try.

Ya know what, I was thinking about that driving home this afternoon. I may take a shot at that. After all, it's already broken. All I can do is break it more, right? Thanks for the tip.

Jim

I would also take a look at replacing with a different pneumatic actuator ... such as the one on D's '93 4 cyl Nissan truck ... "bullet proof". Pretty sure it could be Gerry-rigged into place.

This is a very educational thread. My 1996 3.8 has both codes for IMRC stuck open. I have them both wired closed for the time being (might be a vacuum problem upstream). Actually seems to run better off the line. However, now I have P0171 and P0174. Lean on both banks. The one by the firewall would not stay open when pried open and figure over the vacuum hole. Replaced bad with Rock Auto but still codes. Electric connection seems to be correct since code when away with wired shut solution. So I guess work on finding vacuum leak and MAF? Any help is appreciated. thanks

I used the same one from rock auto if you got the dorman one it will not work. Someone on this forum mentioned that the only ones that work are from the dealer. I am surprised that you did not get the buyout offer from ford. I got mine a few months after i had donated my winnie. Anyhow hope this info helps

This is a very educational thread. My 1996 3.8 has both codes for IMRC stuck open. I have them both wired closed for the time being (might be a vacuum problem upstream). Actually seems to run better off the line. However, now I have P0171 and P0174. Lean on both banks. The one by the firewall would not stay open when pried open and figure over the vacuum hole. Replaced bad with Rock Auto but still codes. Electric connection seems to be correct since code when away with wired shut solution. So I guess work on finding vacuum leak and MAF? Any help is appreciated. thanks

That usually excess means air is leaking into the intake somehow. With the engine running, you can carefully spray carb cleaner around the intake. When you find the leak, the engine will stutter or run rough for a moment. Check the intake hoses, air filter housing, etc.

If you've never done it (or it's been a really long time), you might consider replacing the upper intake gaskets and the PCV valve and grommett. Those are pretty cheap parts that can also cause an intake leak that's difficult to detect.