Horrid Idle with b16

[ProblemSponge]

I had a b16a1 swapped into my 90' hatch several months ago. It came with a chipped ecu. Since then i added a dc header. But since a little while before i put on the header, my motor idles like its drowning or something. It completely flips out. the car idles at 2200 when cold and when it drops down to about 2000 it starts gasping like its out of breath.Sometimes it really goes off, and sounds like a tired fat guy trying to have sex. It'll gasp every 2 seconds or so. After its warm, its good for about 15 miles, then it goes back up to 1500 or so and starts the chicken dance again. Someone please help. I just want some normalcy, please help.:confused:

[Weston]

I'm having the same problem with my B18B... I tried setting the idle several times and it works for a little while, but the next day it'll start idling at 2000 rpm when cold, and 1500 rpm when warm, but sometimes it'll drop down to the correct speed (750 rpm). It went all crazy on me tonight... it kept going back and forth between 1000 and 1500 rpm all fast. It did this a few different times too.

My gas mileage has been really shitty (20mpg) lately too. I've been beating on it more than usual, but the mileage still seems low.

I'm guessing this problem is IACV related, so either it's bad or something is causing it to send more air in. I have a bunch of spare sensors from my old engine, but I need to know which one is causing the problem...

Anyone have any ideas?

[gothemperf]

Use a multimeter to check the voltage of the various sensors and make sure they are compliant with the specs listed in the service manual.

[marcrx5]

Yeah, I've been having a similar problem in my crx for a while. I checked the tps voltage (.5v closed, 4.5V wot), make sure there is no air in the coolant, check for vacuum leaks, and if none of that works, we're all screwed

[stu]

My car usually idles fine, but just today it was acting up again. I personally think it's the cold weather that's to blame because last time it idled shitty it was freaking cold out. Does that help anyone trouble shoot?

[Weston-work]

Yeah, my surging at idle seems to be related to the cold. I have the throttle body coolant bypass mod, so maybe that's letting the air get too cold, which makes it more dense so the engine's getting more than it expects when it opens the IACV. I know that mod isn't recommended for the cold, but I've yet to see any explaination as to why. Hopefully that's the cause, because it's an easy fix.

[Weston]

Well, it just threw a "Check Engine" light on the way home for lunch... code 14, "Idle Control System Malfunction". So I guess that narrows it down to the IAC valve, which is what I had been suspecting. Fortunately for me, I have a spare one...

[rmcdaniels]

Isn't the IAC plug identical to another nearby plug? Maybe the MAP or IAT or TPS or MAF plug? At any rate two of the connectors in that area have identical plugs. I remember a few months ago that my car idled like crap after replacing the intake manifold and I traced it back to switching the two plugs. The car would start, but it idled poorly because with the two plugs switched the mixture was way off. I remember my mechanic switched out the IAC from the old manifold because he thought it was bad, but it made no difference. As soon as I switched the plugs back, everything went back to normal.

[stu]

I should check my CEL that I've been getting. I thought it was the O2 sensor, but I replaced it today and it is still coming on, and coincidentally, it only come on when I let it idle for a minute or more. Can someone walk me through how to check it?

[Weston-work]

On the 3rd gen Integras, there's a service connector on the passenger's side kick panel (up near the lower right corner of the glove box). It's a connector with two wires and it may be plugged into a rubber block (which may be green). Once you find it, just use a paper clip or piece of wire to connect the two wires together, then turn the key to the run position (but don't bother starting the engine). The CEL should then flash out the code's two digits, pause, and repeat... It'll start with the first digit in long pulses, then it'll do the second digit in short pulses. So, if it flashes long-long-short-short-short, that's 23.

A list of codes and more info is here: http://www.londonautoclub.org/html/tech_info/Honda_CEL_Codes.htm

[prkrnt]

ummm...has anyone thought that this could be *similar* to the problem in the other idle thread (http://www.highaltitudeimports.com/vbulletin/showthread.php?s=&threadid=13734). The problems sound different, but the root cause could be the same...anyone?

[95GSRB18C]

thanks for posting that. :)

[Mr. NoSkills]

yeah, check out the other thread, i added a little info that might help.

weston, did you change out the iacv at anytime? maybe the gasket is fucked if you did. i'd put back the tb hoses though, i've read stories about people having fucked up idle after doing that.
that shits way picky.

[rmcdaniels]

Is the service connector by the passenger side kick panel the same one you short if you are setting the timing?

[Weston]

I swapped out the IAC valve, and replaced the fuel filter and PCV valve, but no luck. So then I swapped the throttle body (and TPS and MAP sensor with it), but still no luck. And now it's idling high all the time, and surging much more often. It doesn't seem to be temperature related. It also threw the CEL again this morning, another code 14. So I'm wondering why it thinks something's wrong with the IAC valve, because both the old one and the new one are obviously working (you can hear it in the intake, and it is what makes the RPM able to surge).

If I unplug the IAC valve, the idle will drop and stay steady, but the engine will still sound like it's surging a little. So, I'm thinking it's a bad O2 sensor... maybe it's throwing out bad information, which makes the ECU screw with the fuel mixture and open the IAC valve??? When the ECU doesn't care about the O2 sensors (ie WOT), it runs really well. It seems to run fine when I'm not at WOT too, it just has a big idling problem.

I think I'm going to talk to Acura tomorrow... emissions components for 96 and 97 models are covered under a 14 year warranty, so it may be free to get this fixed...

[marcrx5]

That is almost exactly my problem. You can get it to not surge as often by adjusting the screw on the throttlebody. But as far as getting it fixed, let me know. I'm sick of it. I've checked and swaped out all the parts you listed, plus a differen't ecu.

[Weston-work]

I had my throttle body coolant line bypassed for performance, but I just hooked that back up in an attempt to fix this problem. Unfortunately, it's still idling way too high, so that wasn't it.

I called Acura today to see if O2 sensors are covered under the extended emissions warranty, and they are not (which doesn't entirely make sense, so maybe I'll ask again). They want $95 just to test them or figure out what the problem is. I can't see spending about $100 plus $200 for each bad O2 sensor (I have two) or whatever else is wrong with it. I'm looking into faking the O2 sensor signal right now... I'm thinking the right resistor could do the trick.

I also called my connection at Acura and described what it's doing and what I've tried, so now I'm waiting for a response... I think he must be out of town or something.

[95GSRB18C]

my car is having the same problem as all of you. My idle jumps also, but my idle is very rough. When i also have the IAC Valve plugged in, it wants to rev REALLY high, but when its disconnected, it just seems to bounce up and down. if i cant figure this out myself soon, i might just have to get it towed somewhere and have it looked at. If anyone finds out a problem to theirs or maybe has an idea, please let me know.

Thanks
Bryan

[Weston-work]

Looking around more, my best guess is the primary O2 sensor or maybe some coolant temp sensor. I think I'm going to fake the O2 sensor (or try the one out of my Prelude, if it'll fit) and see what happens...

[stu]

Tobi told me that the O2 sensor doesn't really affect idle, and when I got a new one, nothing changed.

[Weston]

Well, I just tried unplugging the primary O2 sensor, some sensor that's on the #1 intake runner, and three sensors on the side of the block (I'm thinking at least some were coolant sensors). I unplugged them one at a time and started the engine to test it each time, but none of them kept it from surging. ARGH!

I took a look at the Haynes manual's troubleshooting section for this kind of idle problem, and I have narrowed it down to three things: 1) vacuum leak (I doubt it; there's only 2 vac lines on my IM and they both look fine), 2) bad fuel pump (the car runs great when it's not idling, so I'm thinking not), or 3) worn camshaft lobes (only 65k on this motor, and it runs great, so I think not). But being that this is the Haynes manual and not the Helms manual, I'm sure it's troubleshooting section is not complete.

I was hitting pretty good speeds at Second Creek yesterday (better than I've done before), and it feels like it's running really well pretty much all of the time, so I know the engine is running great. It just has that damn wacky idle and I'm completely clueless... :(

I'm still trying to see what free info I can get from Acura...

[Weston-work]

Well, no help from Acura. I did however look through the Helms manual for my CEL code 14 (IAC valve), and by their flow chart, it is determined to be a simple glitch, so it doesn't really mean anything.

I installed an AFC today and I noticed that as I increase fuel at idle, it reduces the surging, but it still idles high. So I'm thinking that maybe there's a leak somewhere and it's getting more air than the MAP knows about. The vacuum lines look good, but maybe there's a bad gasket somewhere...

I searched on google a bit and it seems that many people who had this problem with the high, surging idle, and the cause seems to be a leak around the intake manifold. Now the problem is: how the heck do I find out if there is actually a leak? and where is it? Any ideas???

[Weston-work]

I found the problem (and the solution)! Hopefully this is what is happening to everyone else...

It turns out that my throttle body doesn't close all the way (even with the throttle cable disconnected; I think it has something to do with the spring and torque on that nut)... what caused it to become a problem is the fact that the ECU thought that the throttle body was all the way closed when it really wasn't. The TPS sensor output is around 0.5v when closed, but there's some slight variation in every sensor, so the ECU adapts to it. It remembers the minimum voltage it has seen and considers that to be completely closed. Well, mine was never completely closed, so the ECU never saw the true low voltage, and it always regarded the throttle being slightly open as fully closed.

I found this when I attempted to set my idle again... I got it up to operating temp (so the fast idle when cold valve wasn't open) and unplugged the IAC valve. Spec for a B18B with the IAC valve disconnected is like 460 rpm, so I kept turning the idle screw down to make the rpms drop, but I only got it down to around 900-1000 rpm when the screw was all the way closed. Normally the engine should die with the idle screw all the way closed, so obviously it was getting air from somewhere. So then I opened the screw back up and pushed on the throttle cable in the closed direction and bingo, the rpms dropped. I set it to spec, shut off the engine, plugged in the IAC valve, reset the ECU (pull the 7.5 Amp "back up" fuse for 10 seconds on Integras), and started it back up.

Now it's idling within spec (750 rpm for B18B's) and no surging. When I hit the gas, the throttle wont return to being all the way closed, so it'll still idle high, but now the ECU knows that it's not fully closed and controls the idle system accordingly.

But if I were to reset the ECU and not push the throttle all the way closed, I would get this problem again. I'm also expecting it to eventually come back, because I think the ECU continues to adapt to the TPS, so it'll need to see it fully closed from time to time or it'll forget. So, the real solution is to adjust my throttle so that it closes all the way. I had swapped throttle bodies, but they both sort of stick before they're all the way closed, so I'm thinking that's caused by the spring and the torque on the nut or something. I didn't bother to torque it to spec, as I didn't think it mattered, but now I see that it probably does.

[stu]

But if the throttle body closes all the way, won't the engine die?

[marcrx5]

Thats kinda where I'm at now. It will idle close to spec ( I have a different cam so I think this is part of my problem), but one I rev it, it stays high for about 20 sec, then drops. There has got to be a problem still, but I can't figure it out.

[Weston-work]

Only if the idle screw is closed all the way. Maybe the throttle is supposed to be slightly open, but if that's the case then the ECU expects it, and it's not much (I had no problem getting it to idle really low, I bet closing the idle screw would have killed it).

In my case, the throttle didn't close as much as it was supposed to, so I just pushed on the cable a little and it went to where the ECU expected it to be, and then it learned the TPS voltage for that position. So now when it doesn't close as far as it's supposed to, the ECU knows that and doesn't go nuts trying to get the idle right.

[marcrx5]

I got a SAFC, and when I let off the throttle, tps voltage goes to .5V. Even if it is idling high, voltage is still .5V. At least that is what it is doing for me, this is why I know there has to be another problem also. Try this with yours and let me know if the results are the same. I'm going to try swaping my eacv (for newer hondas it is a IACV) tonight and see if it fixes anything. I'm not throwing a code, but just in case.

[Weston-work]

My AFC said around 0.5v too, but after I pushed the throttle closed, the ECU and AFC learned that the voltage is really slightly lower when it's fully closed. I'm sure it's a very small difference in voltage, but it matters. My AFC was showing 0.2% or 0.3% throttle after I revved it and the throttle went back to where it was, but it hasn't seen the TPS level that low since then and it seems to have forgotten about it now. Maybe shutting off the engine (and therefore the AFC too) made it forget. But the car still idles fine, so the ECU still remembers and that's what matters. It idles just a little high sometimes because of the throttle being slightly open, but no higher than around 1000 rpm and no more surging.

[95GSRB18C]

i still havent figured mine out yet.... :(

[marcrx5]

Neither have I. I tried what you tried, but nothing. I swapped out my eacv, nothing. I have it so that it isn't horrible to deal with, but it still isn't right.

[AcuraBro]

Alright fellas, I have some more info that may help. I just fixed my idle the other day, and the cause was the FIT valve underneath the throttle body. Typically, if your idle is messed up, but you're not getting any CEL's, it's the Fast Idle Temp valve.

There's 3 10mm bolts holding it on... oh, I recommend you drain the coolant first. Once you get it off, there's a side with 2 8mm/philips screws, take one out with an 8mm and slide the cover open. You'll see a plastic white valve with a slot through it and a solid center "axle". Take a flathead to the slott and try to tighten it. If it's tightening, keep going and pop a beer open for you and me because this is the cause for your problem. Once you get it all tight, put everything back (including the coolant).

The thing about this is it's only a temporary fix, the only way to completely fix the problem is get a new FIT valve.