KevinD

iTrader: (56)

so i have a customers car at the shop (nikkadanny from the forum). the car came in for a new timing belt which was perform by me, and nothing out of the ordinary during the install. the issue arose when he went to drive home. under full boost the car hessitated bad, and lit up the knock CEL. he immediately called me and i had him bring it back. hooking up the laptop i confirmed the CEL, 25 counts of knock when giving the car gas (any gas).

this car is equipped with the tephra mod for map switching. on the base map it has the pump gas tune. a safe tune, 18psi, conservative fuel and basically a stock ignition map. the alt map is a high boost, meth tune. nothing aggressive, roughly 28psi. the meth kit is an aquamist kit which is what controls the map switching.

1. misaligned timing belt
at first i suspected i miss-aligned a balance belt or the timing belt. we stripped the belt covers off, and verified that the timing belt is spot on. we also confirmed that the front balance shaft is in phase properly, as well as the rear balance shaft being in time.

2. boost leaks
during diagnosis of the problem, i realized the car was running extremely rich from what i had tuned it. this happens at idle (11-13afr), and at WOT (maxing the zeitronix wideband rich). i figured a pipe may have popped off so did a boost leak test. everything held just fine, no leaks at all.

3. map switching rom
we created a stock rom with his immobilizer to see if it was an issue with the map switching rom. we actually wanted to keep it safe, so we turned the methanol off, and made the stock rom a basic low boost pump gas tune. at first it seemed to be better. the car still ran very rich, but it would boost to 18psi tapering off bad to wastegate. no crazy hesitiation, and no knock counts above 2. heres where it gets weird though. with the car not moving, i free revved the engine and could generate 30+ counts of knock even with a mild blip of the throttle with the car in neutral. experimenting further with the ROM id, i put the map switching rom back in, and tried in on the base map (low boost pump gas). it repeated what i saw with the stock rom. no knock at WOT in 2nd gear, massive knock counts free revving, and car completely falls apart with the meth on. can't get above 10psi (to rich, and knock light flashes with 25+ counts of knock during spool up). car idles extremely rich at idle with both stock rom, map switching rom, and map switching rom on the alt map. car also runs extremely rich driving it, with meth on wideband maxes out rich and stumbles. it now wont even hit 18psi with the meth off as whatever is the issue seems to be getting worse.

4. compression test
fearing something was not right inside the motor, i did a compressor test. numbers came back alright (160-160-155-165). i have not done a leakdown test because i was satisfied with the compression numbers.

5. MAF going bad
because i feel like the huge knock counts are due to the car running extremely rich, i tried swapping MAF's and air filters. exact same results with both.

6. fuel pressure regulator
again trying to figure out why it was running so rich, i went on the assumption that if the fuel pump were failing it would run lean, not rich. but if the regulator is plugged it would run rich. i replaced the regulator with another OEM regulator, and it still had the same problem.

7. CEL
running on the assumption since this problem makes the car literally undrivable beyond the parking lot, i figured if it were a sensor, it would throw a code. no codes have been seen, or are pending.

looking back into the history of the car, danny had reported huge knock counts before when it was tuned by another local tuner. he said it was really bad, and would slow the car so drastically that cars a hundred horsepower less would take him. he then brought the car to me, i tuned it, and the knock counts went away. at the time it had no issues with running rich. its possible the two are related, but because a simple retune fixed it, and didn't have any problems for a long while, i can only assume now it was just a bad tune causing the problems at that time.

looking more recently, before i did the timing belts, danny installed a new intercooler, and a ported exhaust manifold. when doing the exhaust manifold he had a bunch of bolts break off, but as it sits he has everything back together correctly. also, we topped the car off with a new jug of methanol i just picked up from another local shop.

additionally, the car does have no muffler, cat or resonators. it is straight pipes (making it hard to hear rattles or vibrations). it is a full bolt on evo 9 on stock turbo. stock injectors, walbro 255, intake, intercooler, ported manifold, o2 dump pipe, TBE, intercooler pipes (buschur), forge diverter valve, brand new spark plugs (white tips though, so not indicitive of running extremely rich which would be a black plug).


i've got a list of things to check still:
1. fuel return line - any kinks or restrictions will cause high fuel pressure and a rich condition
2. check the wideband with another wideband to make sure its not the sensor itself
3. check the fuel injectors - one might be stuck open slightly
4. swap out the knock sensor make sure its not causing excessive knock
5. check the exhaust and around the block for anything rattling against it.
6. map sensor - try swapping it out.

so if anyone has come across an issue with the car running very rich and super high knock counts, what did you try? any other ideas? i'm hoping i don't have to throw parts at this thing until it fixes but that might be the route to go.

mitsuv

Check the crank angle sensor

Ev0ikon

iTrader: (18)

Are you saying that the Crank sensor is flipped backward?

Sievers

iTrader: (12)

This might just be talk but I will put in my $.02 to see if it is any help.

I am no expert so if I am wrong about any of this let me know

I would maybe check the static and/or dynamic fuel pressure while driving and if everything looks alright and if so, check injectors as you stated.

Also, to my knowledge a motor will not knock just free reving unless mechanical timing is way off. If it does show knock without a load it would definitely not be as much as you said this car is showing.

I would venture to say that the knock sensor is crapin out or it is hearing a mechanical noise it thinks is knock.

I am sure you know this but you are probably going to need to make it stop running pig rich before you can diagnose anything.

Maybe try reading the plugs for knock rather than relying on the sensor???

Hope this was of some help

Ryan

Sievers

iTrader: (12)

By the way Kevin, my motor is about ready to go back in so I will be giving you a call for a tune shortly...

heavens turbo

usually, white plugs and high knock indicate a lean condition. i haven't heard of a car running rich with high knock unless you have bad gas. is fuel pressure ok? tps reading ok? ecm bad? is it an audible knock or just through a scan tool? intake mani gasket or other unmetered air leaks? throttle plate operating ok? ummmmmmm........ man, this is a wierd problem. good luck, keep us posted.

KevinD

iTrader: (56)

TPS reads correct as logged. i haven't checked fuel pressure because he has a stock rail. i'll need to use my billet rail to get my pressure sensor on it to test it. might be doing that though...

the knock is not audible, it is just logged through evoscan (and picked up by the knock CEL). theres definately not any unmetered air though, boost leak tested both the intake manifold by itself, and the intake track from the turbo to the valves. no leaks at all. thats why the rich wideband readings are throwing me off.

also, the plugs had no indication of pitting either, so i'm not positive there actually is real knock. i'll shove a bore scope in the cylinder to inspect the piston tops to make sure i'm not seeing any pitting or damage.

bsmart82

Can you post some logs?

barneyb

I hope this doesn't sound goofy but how loud is the exhaust? Could that be setting the knock sensor off?

KevinD

iTrader: (56)

exhaust is quite loud, i considered adding a muffler to it to see if it changes anything.

here is a log attached. notice the 35 count of knock with only 40% throttle. i have a hundred logs from this so i don't remember the specifics of it, other then there is some free revving and looks like gentle acceleration (in second or first gear)

mitsuv

Did you unbolt ckank pulley and crank sensor blade?

Can't you install a wideband to check AFR

Jhutterer

^ +1

KevinD

iTrader: (56)

no i did not pull the crank sensor blade off. and if it were the crank sensor, the car wont run and it will throw a check engine light.

there is a wideband on the car, i do check afr constantly. i don't log it through evoscan though, i log it with zeitronix because the rate is much higher. and as mentioned i'm going to add a second wideband to ensure the first wideband is reading correctly (although i know it is running rich because the fuel trims are all very negative)

nikkadanny

iTrader: (31)

hey kevin ill bring my exhaust with me when i go up there today to meet up zach. so you can check it out, i was thinking maybe its the cams after the timing belt. ill prolly get some fresh 93 as well.

mitsuv

I have no reading of afr with the datalog you uploaded.
I know that a crank sensor problem would flash the cel, i was thinking about a wrong setting of the sensor blade or the sensor itself, causing higher timing advance.