legacycentral bbs

Why do these things always happen to me?

First, I was up in the NC mountains over spring break. I get up there fine, and don't drive for a day or two. I jump in to move my car to a different spot, and there is a persistent CEL while I move it. We proceed to clean all the nastiness that has accumulated over the past 16 years out of the engine compartment. It looks wonderful. When I go to start it up to check something else, I get a TPS code. I realize that I got water in the connection, and dry it out and reset the ECU.

For the next week or so, I occasionally get CELs on startup that go away if I turn it off then back on. Confusing, but I don't have time to screw with it since I'm coming up on exams.

Fast forward to today. I was on my way to lunch between classes, and I got a CEL that stayed on. Through some experimentation, I found that when the CEL was on, it would only run at about wastegate boost (~6-7 PSI). It came on and went off a few (3 or 4) times in just 20 minutes. I figured it had to be the pressure sensor/boost control solenoid. I get home and pull the codes to see which it is, and I get a laundry list.

Code 22: abnormal voltage from Knock Sensor
Code 23: abnormal voltage from Air flow sensor
Code 31: abnormal voltage from Throttle Sensor
Code 42: abnormal voltage from Idle switch
Code 45: Faulty pressure sensor or pressure exchange solenoid valve inoperative
Code 49: use of improper air flow sensor

Codes 22, 23 and 49 I am used to seeing, as they show up every time I pull codes. I figured these were just some it threw for a second, then realized there wasn't anything wrong.

I have a set of FSMs, so out they come. I decided to check all of these instead of just the code 45, just to be on the safe side.

First, I discard the idea of checking the knock sensor, due to its royal PITA position. I can see it since I cleaned my engine 2 weeks ago. It's the old grey style. I figure it is broken.

Next, I decide to check the MAF sensor next. I am getting incredibly off reading from 1 or 2 of the terminals on the connector, but ~correct reading from right at the ECU (which is a royal pain to get out).

I checked the TPS signals, and the resistance of the actual sensor was well below minimum levels. One of the terminals on the connector had 0 ohm resistance between it and the body when it should have had ~1 M ohm.

I am in the process of building a scantool, and hope to borrow my dad's laptop or a friend's to check out all my various parameters.


This sucks. Any ideas anyone? Is there something I'm doing wrong?

Nobody?

I just got done making the cable, and now I just need to commandeer a laptop. My dad is hopefully getting his switched out right now, so I should have some results tonight.....

Anything I should look for?

Anybody here?

Chances are only one of those codes is actually active when the CEL is on. Most of those probably happened when stuff got wet.

Next time the CEL is off, try clearing the codes. Then the next time it comes on you should have only one code. Hopefully.

I think there's a better-than-50% chance you'll find it's the knock sensor code.

I'll have to do that.

Would the knock sensor cause it to cut boost?

What do you mean by cutting boost? It would cause it to limit boost to wastegate-regulated levels, but it wouldn't cause it to cut fuel.

Code 45, however, would cause it to cut fuel to prevent the engine from goign on boost.

Limited to wastegate boost.

I haven't experienced fuel cut at all.

That adds credibility to the knock sensor theory.

i just double-checked, and it looks like the previous owner didn't replace it (like he did everything else).

Hmmmm...

Isn't the updated sensor supposed to have a white connector?

If so, it looks like this has an updated sensor.

Yes, it is.

It's still possible the sensor is damaged or that it has a poor connection to the block. You should take a few minutes to check it out.

A tip: Have one of those claw-type pickup tools ("magic fingers") handy. It'll make it easier to fish out the bolt after you loosen it, and you can use it to start the bolt as well.

What about a magnetic pickup tool? Would that affect the sensor in a detrimental way?

No, but they're not the same thing. Magic fingers allow you to grip the head of a bolt, and their shape allows the tool to function as a u-joint so you can thread a bolt into its hole, making the installation of hard-to-reach parts easier.

Not that you need to do it how I did... It just makes it easier, I think. My knock sensor swap took me less than 5 minutes.