zekgb

So as the title says, for the first couple of days after I completed the intake job (which included all new fuel lines, plug wires, coils, injectors, MAF, TPS, CPS and throttle body orings) I noticed that the car would idle and run great on the initial drive no matter how long the drive was (including a 60 mile commute) but would then idle somewhat lower with a slight hunting in subsequent runs until completely cooled down. The idle was definitely worse, but there were no other drivability issues other than slightly firmer shifts.

zekgb

(Continued)
Today I received PorKen's chips and to baseline I did a couple of 1/4 mile runs with the old chip set after coming up to temperature. Car ran well with a couple of typical S3 deadspots. The chips took me a couple of hours to get in and the car felt amazing. I did a couple more 1/4 mile runs and the car was .4s faster than earlier. I took it home to adjust the bowden cable as the 2-3 shift was happening at 6500 and on the next run the dead spots were there again. When I slowed down to make my U-turn the car was now idling much rougher than previously and this time would stay rough at below 2k Rpm on part throttle but would smooth noticably above that range.

When I got it back to the garage I noticed a pronounced whistling coming from the intake. I sprayed some ether at various places looking for a vacuum leak and the idle slowed pretty drastically at the base of the runner for 6 (I was expecting the idle to surge, but still obiously a leak, right?) So re-torqued all of the runner bolts and indeed somewhow the bolt for 6 and the shared bolt between 6 and 7 both took an extra half turn or so.

Re-started and the whistling was gone but was still idling roughly and responding poorly at part throttle. I went back with the ether and saw similar behavior for the hose to the ICV from the throttle body (idle drops drastically when hit with ether)but unfortunately I didn't do a great job locating the hose clamp for the ICV and center T so now I need to remove at least one of the intake runners to get to it. I'll obviously re-tighten everything WIAIT and re-orient all of the clamps so I can get to them without having to lift the fuel rail and runners.

Questions: Does this behavior make sense that a vacuum leak is a likely cause? Would an ICV leak specifically affect drivability in this way (most of the threads I've read about ICV leaks caused high idle)? Am I ok to remove just the one runner to gain access and reuse the paper gaket that's now been in use for a couple of weeks? Anything else I should be looking at other than vacuum?

Mrmerlin

from the info provided I would say you have a vacuum leak the system is full of potential leaking areas, and any seals that are reused have a higher potential for loose fitment.
That said if you have identified a leak then it should be investigated.
Did you use any DC 111 on the manifold injector O rings?
removal of the intake runner shouldnt hurt anything but the fuel rail comes out first.
posting a picture of the injectors and fuel rail free from the engine would be a good idea

Imo000

Sounds like a vacuum leak to me and/or turbulent air or leak at the MAF sensor. Keep poking around with the starter fluid and see if you find more leaks. The running OK when cold and rough when hot is due to the richer mixture when cold so the leaking air is not as much of an issue but when the engine warm up and the mixture is leaned out by the ECU, the extra leaking air created a lean missfire.

Dadddio

Smoke test is really the best way to go, if you have
access to a smoke machine.
That'll pinpoint the leak, and probably find a few more you didn't know you had.
I had a similar problem.
Ended up being my intake gasket.

zekgb

Injectors are from Roger so o-rings are fresh. I did use DC 111 on them. Plan now is to remove the rail on the passenger side which will give me access to the clamp for the center T and from there all the other clamps. I'll post pictures when I have it apart.

zekgb

Well so much for reusing the intake gaskets, they are already fairly oil soaked after less than 500 miles and tore as I lifted the front right runner (intake for 1 is much wetter than for 2.) It's an S3 so there's lots of oil in the intake (which I've always had) but I'm surprised the gaskets are so wet. I didn't wire wheel the runner mating surfaces after powder coating, could that be the cause? Should I be using the metal gaskets instead?

Imo000

Yes it could be the powered coated mating surface that lets the air in.

zekgb

The mating surfaces are primered but not powder coated and I am feeling stupid for not catching this. What's my best bet to knock it down back to the metal?

Lizard928

Belt sander, or random orbital sander.

Mrmerlin

if you dont have a sanding machine ,
tape some 220 grit paper to a smooth surface and work your part in a figure 8 to surface the part. finish with 400 grit

Landseer

Possible to nick an injector O-ring during installation.

Jfrahm

If a vacuum leak cannot be found it's worth looking at the coils, they can fail when warm/hot and work when cold. A spark checker comes in handy for this, the spark has to jump 10mm or so to pass the test.

zekgb

So I finally got everything reassembled today after hitting the mating surfaces of the runners with a random orbital sander to remove the primer (also added Greg's Oil Filler Baffle WIWIT), tightening all of the vacuum hose clamps and re-checking the injector o-rings (no issues) and while it's a little better from the previous behavior as far as idle and part throttle response, it's still far from good even when cold. Checked vacuum at EZK and only seeing 12 inches at idle with a steady needle, so obviously a pretty big vacuum leak. Nothing located using starter fluid, so I guess next step is a smoke test?

zekgb

Found a disconnected vacuum line for the diverter valve that I thought I had eliminated, now getting a steady 17.5" at the EZK while idling. Drivability much better (actually driveable now) but idle is loping/low and part throttle 1500-2k Rpm is improved but still noticeably not smooth.

Next up:

o Smoke Test
o Replace O2 Sensor (current is 20k miles/4 years old)
o CO/idle adjustment (If I can manage to fabricate a blink'r)
o Check timing (adjusted by Bill Ball 6k miles ago using 32ver)
o Examine plugs after a few hundred miles
o Compression/leakdown?

Anything else I should be thinking about?