General Bad ECU?

[boiseDSM]

Hey guys, this is my first post. I've had an annoying problem for a while that I can't seem to solve on my own. I'll try to be as detailed as I can.

My car is a 1990 TSi AWD. Stock fuel system except for a 255lph fuel pump. It has always ran great except for the occasional boost leak.

The car run runs fine when cruising at a constant speed and high vac. However, pressing into the throttle at lower engine speeds, I get massive knock. So much so that it pins the data logger at 40+ counts and the EGT's go up. If I downshift to get the RPMs up to say 4k and go WOT, it feels great.

Here is what I've done so far to try and fix it...

Replaced knock sensor
Replaced cam angle sensor
Tested the ISC (good)
Checked O2 sensor (cycles normally)
Pressure tested the intake (no leaks at all)
Checked grounding straps for ECU and engine
Checked idle position sensor (good)
EGR is blocked off

Did I miss anything?

The car runs fine when warming up and when WOT at higher RPMs. But driving at part throttle and lower RPMs, the car runs like someone else has the throttle.
The randomness of the problem leads me to believe that it is electronic in nature. Sometimes a jab at the throttle will produce acceleration, and sometimes it just makes the EGT's go sky high.

The ECU has never been serviced. Does anyone have any info on how to diagnose or test the ECU.

Thanks in advance for the feedback!

 

[evil_eagle]

How tight did you tighten the knock sensor? If you over tourqe the sensor then it will get alittle wild on you. Pull it back out and re tighten it, most of the time I just snug it, no more the say 10ft lbs.

Other thing to check is the timing marks, make damn sure they are all lined up, most of the time it ends up being the cam gears are off. That will cause knocking issues as well.

But sounds to me it's a faulty knock sensor, or it' just too damn tight.

Another thing to watch for is noisy lifters, they can also cause false knock at part throttle, what you have my friend is phatom knock, happens to the best of us.. You can clean the lifters fairly easy, just pop them out and soak them in carb cleaner for a few hours. Then pop them back in.

The reason for the high EGT's is caused by the timing being pulled when your getting the 40 counts of knock, timing retard= high egt's= loud intake=crapy performance.

I would have to rule out the ECU, if it was the ECU such as leaky caps you would know it for sure, it would miss fire on you and you would smell a roten sea food smell from the leaky caps.

Has your car allways done this, or just recently? If just recently what changes did you make? Most of the problems are caused by the last mod you have done, back track to the last time it was running good. The work from there.

Also check base timing at idle, should be set to 5degrees BTDC, don't forget the ground out the timing strap when adjusting timing on the car.

I live in nampa if you need any help, been working on dsm's for almost 11 years now.

-Jake

 

[boiseDSM]

Thanks for the response. Yeah it sounds like you know everything there is to know about these cars. 11 years is a long time. Very nice car by the way! I'm sure everyone is asking for a ride.

The knock sensor is new. The old one that came out of it was VERY tight. It took me a while to get the old one out. The new one is tightened as you would suggest (not too tight). The cam timing is right on. With the two hash marks on the cams at 3 o'clock (exhaust) and 9 o'clock (intake), the crank is pointing rght at the marker.
I don't think it is the lifters. When the oil is cold and viscous at startup, I sometimes hear lifter tick for a while but the car still drives and runs smoothly. They are 1G lifters so I will take them out and soak them in some solvent. Base timing was just set at 5 degree BTDC (set w/ plug grounded).

The problems history is sort of difficult to back track. It showed up sometime probably 6 moths ago and has been getting worse lately. One thing I've done to my car was to replace my home depot special ABS plastic intake manifold (yeah, I know) with a nice mandel bent pipe I welded together. There are no leaks in it however so I don't think that would be the cause. All clamps are tight.
Another thing I've done about that same time was to install an EGT block off plate. I wonder what the consequences of a leaky plate would be? I'll have to check it.
I'm not certain if there is any connection but my A/C compressor clutch won't engage even though the fan turns on. And my cruise control won't turn on. The small orange LED on the column turns on but the green "cruise" light will not come on. I've checked both connections and everything looks good, as does the fuses. I know the A/C problem is not connected but the cruise control went almost at the same time. The problem seems to go away if I reset the ECU. It runs great but start to act up after a few hours of driving.
I've been reading the thread "knock under and load" and it has helped for ideas. I will have to data log it again and post it. I know that under WOT, my o2 is at .81 which from what I've read is pretty lean. My fuel trim is at 118% low and 104% mid and high.

I'll get back to you once I've got some logs in. Thanks again for helping a novice DSM'r.

 

[Turbo442]

Couple of things I would look at. Is your data logger showing any unusal sensor readings? Watch every sensor reading for a few minutes. Any check engine lights? Whats your fuel pressure at with the 255lph pump?

 

[evil_eagle]

Well, I would double check to see if that block off plate on the EGR is nicely sealed up, are you using a metal gasket around the block off plate? And what type of material did you use for the block off plate?

Your fuel trims look about right for your given mods, your mid and high fuel trims are fairly close to 100% My old FWD talon back in the day would be at 125%-135% mid and high fuel trims. I would not be too concerned about those.

As far as your o2 readings, how old is your o2 sensor? with those fuel trims and .81 o2 volts my geuss would be you have a slugish o2 sensor that is going south on ya.

Now I forgot to ask you, did you gut the MAF at all? IE lower honey comb, or anything done to it?

Also what brand of fuel are you running? what octane level? The idaho pee water here sucks man, trust me I know, I fight it every day with my AEM EMS.

Do this, this is by far the best way to find out if it is real knock or not. Goto smokin hot deals on overland, get like 2 or 3 gallons of there 110 trick race fuel, dump it in the car and go log it again, if the knock goes away for good while running the race fuel then it is real knock your getting. If it still there after you put the race gas in then it is just false knock. Run low boost to start out with, this way you know for sure with that type of fuel and boost level you know in your heart you are not knocking, then log it and find out what happens.

And do not be scared to run race gas, a couple of gallons will not hurt the o2 sensor at all. I run c16 almost every week, that 116 octane leaded fuel, granted I change my wideband o2 about every 2 months or so, but I get the sensors for hella cheap, so its alittle more reinsuring for me anyway.

Let me know.

 

[boiseDSM]

The block off plate is sealed up tight.
I'm starting to think more and more that the culprit is the O2 sensor. My logs show dead places on an extremely light cruise or idling at a stoplight. It used to cycle at idle and now is just read -0.02, which I don't think means too much. It does have some miles on it too, about 50,000 I think.

The MAF is hacked. The honeycombs upstream of the sensor are still intact. The lower honeycomb is removed.

I'm running mostly 93 stinker. I am running boost directly to the wastegate without a boost ontroller. My old one stuck closed and I haven't had time to get a new one. So I think it really is phantom knock. I don't see any way 93 gas should knock so bad on 9 psi.

I will be getting a new O2 sensor soon. I'll let you guys know if it helps.

Evil_Eagle-> What are the symptoms of a dead O2 sensor? Could a dead one cause knock?

thanks again