Rep Power: 0

Hi all,
For those with a short attention span, just read the bold text.

Honda Civic 1.6 Sport (2004) Hatchback
EP2, Mk VII
SHHEP23904U201890
D16V1 4103654

I am competent enough to solve and repair pretty much anything on any car.

This problem has me at my wits end.
Those of you reading the title will already conclude 'Inlet air leak' or 'Sticky/faulty Idle Air Valve'.

Some months ago, I had a cam belt shred (mia cupla - 185000 miles on it). Bent *all* the valves, so e-bay for parts list and out with the spanners...
I have done similar work on my Wife's Civic (same model, 2005) when she cooked the engine and warped the head due to a faulty radiator cap.

I took the battery off and charged it so the computer was without power for a month or two whilst I was doing the work and waiting for deliveries.
I had a bit of a problem getting the crank bolt off, otherwise a normal tear-down. The only damage to the pistons were small curved indentations in the crowns about 0.5mm deep, bores were clean, camshaft, bearings (very slight scoring) lobes and followers all clean and looking like new.
Replaced all valves, ground in, valve seals, head gasket, inlet, gasket, exhaust gasket, timing belt, camshaft seal. Cleaned out the carbon from the EGR channels and shined all the mating surfaces.
Rebuild went without incident and finally the engine started first time and sounded great.

This is the beginning of the problem.
Starting idle was 1100 rpm, which is fine for a cold engine, usually the RPM creeps up as the engine warms and then a stab on the throttle will see it drop down to 800-900 rpm - normal, but this time it raised to 2000 rpm and stayed there.
Well, that's frustrating...
Went for several drives and it runs like a champion, just a pain in the *** when arriving at traffic lights.
There is no Check Engine light and using Torque to list errors shows a clean slate.

Did all the usual checks to ensure I had connected everything, all pipes fitted and that I had not missed something - all in order. Checked the water level several times to try and top it up further, nothing but it is circulating fine and the throttle body gets hot when the engine is at 86Deg C.
Maybe cam belt is off by a tooth - so took the cam-cover off and set TDC, checked the cam pully - no - it is correct.
The throttle body had been exposed to atmosphere under the bonnet (hood?) and the carbon had become a bit sticky. So unmounted that and thoroughly cleaned it with carb cleaner, along with the Idle Air Valve - all looks nice and operating freely. Same problem.
Ordered a second hand throttle body with all sensors etc. and swapped it - same problem.
Tried blocking off the Crank case breather pipe and the Brake Vacuum pipe no difference.
Took the whole inlet manifold back off for examination - no faults cracks or anything that might leak, rubber seals were in good condition at the throttle body and the manifold joints.
Refitted with a very thin smear of instant gasket on the gasket just in case. Made extra certain that I had got ALL the nuts and bolts in and tightened properly.
At this point I am pretty certain that there is no inlet leak.
Maybe it is the O2 sensor - that was exposed to atmo whist the exhaust manifold was removed, so replaced that - no difference.
Perhaps the EGR is stuck open?... Removed the EGR valve (initially closed) and verified its operation with power accross the pins - working fine, and ... no difference.

Google turns up many references to this sort of behaviour on older and newer models and their solutions (mostly idle valve or throttle position sensor) did not work for me.
There is a possibility on the 2007+ Mk8 that the throttle position sensor 'learned values' can be reset and then re-learned, but I have seen very little that relates to the 2001-2005 model.
There are suggestions that driving it will eventually sort itself out, after 1/2 a tank of fuel it is still the same.

I have disconnected the battery numerous times, left it for hours. Reconnected and allowed to idle, done the same thing holding the revs at 3000rpm (as suggested by another Honda reset procedure) - it does not make any difference how I hold my face, results are the same.

The only things that seem to affect it are that:

• Putting the Air Con on will drop the revs to 1000 rpm whilst the pump is engaged (which is odd as the engine should increase speed to allow for the load)
• Pulling the Manifold Absolute Pressure sensor connector reduces the revs to a steady 950 rpm (which I take as evidence that there is NO inlet leak, since it would not change.)
  When I rev it, the manifold pressure goes DOWN (to about 5psi) when the idle is high and goes UP (to about 7psi) when the Aircon or MAP sensor disconnection is used to drop the revs.

I ordered and replaced the MAP sensor (after-market) and Torque showed that is was measuring the same pressure as the original, so no difference.

I took the car to Honda and asked them specifically to solve the idle problem, they did a diagnostic and I quote:
Their report included a number of other unrelated MOT related issues that I have since been remedying and charged me £120 for not fixing my car.
As I understand it, timing issues would show up on a diagnostic and are entirely computer managed anyway (no adjustment possible).

As you might conclude I have fairly thoroughly explored this problem and followed a good deal of discussion on the subject elsewhere on the internet and not solved the problem yet.

Any suggestions are welcome - I find that discussion even with respect to things that are unlikely to work can often be fuel for ideas.

 
Rep Power: 223

Usually a high idle would be caused by a vacuum leak.

 
Rep Power: 0

Thanks for the response.

Agreed - the most sensible conclusion.

However, if it were a vacuum leak, why would disconnecting the MAP sensor bring the revs down to normal... if it were a vacuum leak, I would suppose the idle valve would not be able to compensate for it.

I do have a worry that when I cleaned out the injector ports on the injector/EGR plate, I may have been too aggressive with a wire 'ball' on a drill.
It felt ok going back in and I assumed that the injector rail would 'press' the injectors into the holes against their end-stops.
I am going to try taking the injector rail off and supplementing the rubber grommets with a smear of instant gasket.
I could also very slightly bend the rail so the curve presses the centre two injectors more effectively, and then perhaps shave a little off the (bakelite?) spacers to give it another 1mm of traction.
If either of these works I'll get new grommets and a replacement plate from a scrappy.

I have done the butane probe test and could not find anything other than the carb mouth to get a change in engine idle.

 
Rep Power: 0

Took the injectors/fuel rail out and took measures to ensure they are tight and sealed.
Absolutely no difference.

I have ordered a bi-directional diagnostics device:
https://smile.amazon.co.uk/gp/product/B08B5RMFBG

Hoping that a throttle learning values reset might work.

 
Rep Power: 0

I know it’s been a while but I’m having the same issue, did you manage to figure it out?

 
Rep Power: 0

Sorry mate, I ended up scrapping the car.
Given the amount of things I reviewed, checked and replaced I had to conclude that maybe I had twisted the crankshaft getting the pully off (wedged the flywheel starter ring and used a very heavy impact socket on a high power air impact)
I can tell you all the things it wasn't?