Intermittent high idle while coasting

[Avery4]

Hello everyone, several months ago I swapped my 2005 Honda Civic 1.7 VTEC from an automatic transmission to a 5 speed manual and I have been having an idle problem ever since. When I am coasting, the RPM will usually not go down to an actual idle. Instead, the RPM usually hangs around 1000-1500 for no apparent reason although I have seen it go down to an actual idle a few times. There isn't much consistency to the high idle either. Sometimes turning on loads such as the AC will increase the idle, but sometimes the RPM will stay constant. The only consistent behavior is that the idle is always a perfect 700-800 RPM when I am stationary. Also, if I unplug the speed sensor so the ECU doesn't know the car is moving, the idle is always perfect.

I have the proper 5 speed ECU, proper IACV, no vacuum leaks, TPS reads idle. I also tried disconnecting and plugging off brake booster, PCV, and evap solenoid, but no luck yet. I swapped the IACV with an extra one I had just to try it, which as I expected didn't solve the issue. I did initially think I fixed it though since immediately after that the idle was a perfect 700-800 RPM the whole way to college the next day regardless of loads or speed, but it started acting up again on the way home. I have also tried with the alternator unplugged since I read it can cause a high idle if it is incorrectly reporting a high load to the ECU, but it made no difference.

I am starting to wonder if I have a defective ECU since the idle is so inconsistent while coasting. Seems somewhat unlikely since the car drives perfectly otherwise, but at this point I can't think of another explanation since the engine is obviously mechanically capable of idling properly if it wanted to.

If anyone has any other ideas please let me know. Thank you very much in advance!

 

[WylieCoyote]

Every manual car I've ever owned kept the idle elevated if the car was in motion and the clutch disengaged, or transmission in neutral. Nothing unusual as far as I'm concerned.

I don't know when Honda moved to throttle-by-wire, but if you have a cable-operated throttle, you should have an Idle Air Control Valve that's responsible for increasing and decreasing idle speed as the A/C cycles. They can become inconsistent when they're dirty or worn out. This also controls high idle while in motion. I'd advise replace versus clean.

 

[ron17571]

My first thought was the TPS or a vacuum leak. Maybe the ECU. I guess changing these parts may be the answer, but I like finding the problem first. Good luck.

 

[Avery4]

Every manual car I've ever owned kept the idle elevated if the car was in motion and the clutch disengaged, or transmission in neutral. Nothing unusual as far as I'm concerned.

I don't know when Honda moved to throttle-by-wire, but if you have a cable-operated throttle, you should have an Idle Air Control Valve that's responsible for increasing and decreasing idle speed as the A/C cycles. They can become inconsistent when they're dirty or worn out. This also controls high idle while in motion. I'd advise replace versus clean.

Good to know, thanks. I have read that too, but if that was the case why wouldn't the idle be consistently high while rolling? If it would consistently idle slightly high at about the same RPM every time I coast I could see that, but I have a hard time believing this is normal behavior since the idle is so inconsistent. For example, on the way to college today, it was idling at around 1400-1500 RPM while coasting for the first half of the trip (12 or so miles), but then for no apparent reason the idle dropped to around 1K-1100 RPM for no apparent reason despite everything else being the same as far as I could tell.

Also, I asked a friend with the same generation Civic what his does and he said his idle is always under 1000 RPM while coasting. It is certainly possible that there is a slight difference in ECUs though.

I have a cable throttle. I did clean the IACV and it rotates freely. I have a hard time believing that the IACV is what is causing the problem since the idle is always consistently perfect when stationary, but who knows.

 

[Avery4]

My first thought was the TPS or a vacuum leak. Maybe the ECU. I guess changing these parts may be the answer, but I like finding the problem first. Good luck.

Thanks, I would certainly like to find the problem too. Swapping parts for extras I already own is one thing, but the ECU would be an expensive gamble. Even if I could find one at a junkyard cheap, I would still need to have it programmed by the Honda dealer for $130 before I could use it. I can't think of another way I could definitively prove that the ECU is or isn't the problem though.

 

[WylieCoyote]

You can't ascertain an IAC Valve's performance based solely on the fact that it rotates freely. The IAC, or a sporadic vacuum leak are pretty much the only things that will give you an inconsistent idle speed when all other circumstances are identical. I wouldn't go down the ECU road until you've ruled out much less expensive fixes.

The IAC went bad on my '99 Explorer, and under identical circumstances, sometimes it would idle fine at a stoplight, and sometimes it would die.

Replaced the IAC, never did it again.

 

[Avery4]

You can't ascertain an IAC Valve's performance based solely on the fact that it rotates freely. The IAC, or a sporadic vacuum leak are pretty much the only things that will give you an inconsistent idle speed when all other circumstances are identical. I wouldn't go down the ECU road until you've ruled out much less expensive fixes.

The IAC went bad on my '99 Explorer, and under identical circumstances, sometimes it would idle fine at a stoplight, and sometimes it would die.

Replaced the IAC, never did it again.

Thanks, I agree. I suppose its possible that both IACV's I tried are electrically faulty, but that seems unlikely.

At this point I am wondering what ECU inputs are used to determine an appropriate idle speed and which inputs could be changing without my knowledge. I know coolant temp, alternator load, TPS, AC clutch, and power steering pressure are idle inputs but I have ruled those out. I'm wondering what other inputs could affect idle speed.

 

[mk378]

Did you change to a manual ECU? If the ECU still expects an automatic it probably isn't going to run right.

Rolling the car in neutral is incorrect driving anyway. Stay in gear except to stop.

 

[Avery4]

Did you change to a manual ECU? If the ECU still expects an automatic it probably isn't going to run right.

Rolling the car in neutral is incorrect driving anyway. Stay in gear except to stop.

Yes, I got a 5 speed ECU shortly after changing the transmission. Good point.

What's incorrect about coasting in neutral? I understand decel fuel cutoff, but I also understand that an unpowered engine creates significant drag and that I can coast much further in neutral, saving fuel when used under the right conditions. When I started coasting when safe and appropriate, my MPG increased from high 30's to low-mid 40's.

 

[vavavroom]

I don't know the first thing about Honda vehicles but the cars with MT I've had would drop to normal idle RPM's in neutral or with the clutch disengaged, whether the vehicle was moving or not, usually to well under 900 RPM's. If there were any hunting, I'd suspect the IACV or a vacuum leak.

 

[oldhp]

Both of my Challenger R/T's ( 2016-2019) were 6 speed manual trans. Both would idle at 900RPM when in neutral and car was moving at any MPH.

 

[svtrit]

My Jeep idles high while coasting in neutral. I’ve just come to accept that. I can’t recall if my sons’ GTI did or not, and the last car I owned that was stick was an 86 Dodge, which would be too old to compare for sure.

 

[Avery4]

I don't know the first thing about Honda vehicles but the cars with MT I've had would drop to normal idle RPM's in neutral or with the clutch disengaged, whether the vehicle was moving or not, usually to well under 900 RPM's. If there were any hunting, I'd suspect the IACV or a vacuum leak.

Thanks, that's what I would expect. I see no reason for the speed the car is traveling at to have an effect on the idle speed when in neutral, clearly the ECU is doing something strange for some unknown reason.

There isn't any hunting, just usually a steady high RPM of a seemingly random number between 1000 RPM to up to around 1800 RPM, although it typically doesn't go much over 1500 RPM.

I wonder if intake air temp is an idle input. I noticed that the engine seems to somewhat consistently idle a bit higher while rolling until a few minutes after the coolant is fully warmed up when the intake air temp starts to warm up by a few degrees. I will see if I can influence the coasting idle RPM by heating/cooling the IAT sensor while the wheels are spinning on jack stands and report back.

 

[vw7674]

Honda had a switch, to sense speeds over 10mph, back in the day (1980-ish)
theory being extreme throttle snaps produce excess emissions. They were fitted to 5 speed manual cars. Over 10 mph the engine would not drop to idle right away.
Pretty sure modern injected Hondas also keep the engine speed from falling in anticipation of a shift if you're over a certain speed.

 

[Avery4]

Honda had a switch, to sense speeds over 10mph, back in the day (1980-ish)
theory being extreme throttle snaps produce excess emissions. They were fitted to 5 speed manual cars. Over 10 mph the engine would not drop to idle right away.
Pretty sure modern injected Hondas also keep the engine speed from falling in anticipation of a shift if you're over a certain speed.

Good to know, thanks for sharing! The RPM does seem to drop more slowly while rolling, presumably like you said in anticipation of the next shift to help rev match. That isn't the problem though, the problem is when I am coasting and the RPM won't drop to idle.

I understand that on some vehicles this may be normal behavior, but I have a hard time believing what I'm seeing is normal on my car because it is so inconsistent. If the ECU was programmed to hold a higher idle while moving, I would expect it to do the same thing consistently and mine doesn't.

 

[Avery4]

Update- I tried to test my theory of the IAT affecting the idle RPM by heating the IAT sensor with a heat gun while the engine was idling and the wheels were spinning on jack stands but I wasn't able to noticeably change the idle. Didn't have time to try cooling the sensor to see if that would increase the idle though.

Today the idle seemed to be fine. Every time I would push the clutch or shift to neutral while rolling, the RPM would consistently drop to around 1K ish over a period of 2-3 seconds once the coolant fully warmed up. The idle RPM held relatively constant when turning loads on and off too. Still ~200 RPM higher than when sitting still, but since it behaved consistently and was never excessively high I could believe that is normal behavior for this car. Too bad it won't just do that all the time.

 

[Avery4]

For the first time in weeks, the RPM actually dropped down to around 800 while coasting once the engine warmed up but only for a few miles. The idle then returned to around 1K-1100 RPM while rolling. I'm starting to think that's normal for this car since that's usually where it likes to idle when I'm coasting. Why it sometimes stays higher or drops all the way down to idle I have no idea.

But for whatever it's worth the engine is not happy idling at ~1K RPM when I'm rolling, it feels quite rough and annoying and rattles everything. The normal ~800 RPM idle when I'm stationary is smooth though. I do have polyurethane engine mounts, but the only significant roughness I ever feel is at idle around 1K RPM. Other than that it's smooth, I can barely feel the engine running at normal idle.