legacycentral bbs

Hey guys,

I'm going to be tackling the problems with my car yet again this afternoon and I'm trying to put together as much information as I can before I start. So far I know that the ecu is showing the following codes:

11. crank angle sensor - no signal entered
22. knock sensor - abnormal output
23. airflow sensor - abnormal voltage
31. throttle sensor - abnormal voltage
42. idle switch - abnormal voltage
49. air flow sensor - use of improper air flow sensor

Is there any information in your manuals for precedures to test these sensors? Sorry, it's a lot of information that I need, but any help you two could provide to get my car back on the road would be greatly appreciated!!

All those codes came off at the same time ?

"You sure you not locked your feets in the wires ? "

Have you taken a close look at all your connectors ?

Number one thing when trouble shooting. Reset the ECU, drive the car and see what codes reoccur. It is very likely that all those parts are bad.

Okay, worked on the TPS last night while I was waiting for the codes to clear. After much frustration and a lot of cursing I was not able to adjust the idle switch properly. There was literally NO WAY to have the sensor read no connectivity at .0028in and connectivity at .0035, or whatever you had to do (I tried and reread the instructions several times). What I noticed when checking the TPS was that the resistance at the designated terminals on the sensor would read 24.something k-ohms when closed and 9.something when open. Something's definitely wrong with the sensor.

You can get the sensor to read the proper values......if the sensor is good.

It sounds like yours is not.

Anyone know the part number for the TPS?

I'm pretty sure all BC/BFs have the same throttle body, part number 22633AA060. They're really expensive new, though, so I'd suggest you go to a junkyard and pull one from another Legacy rather than buy one from a dealer.

I was able to find a whole throttle body assembly at a wrecking yard for $29.99+$15 shipping. Not too bad considering the TPS at the dealer was like $350!! Even subaruparts.com was selling them for $250!!

I ordered another knock sensor from subaruparts.com as well . . .

Assuming these two "fixes" don't work (these were the original codes to show up), it's onto the crank angle sensor and MAF sensor. Anyone know the part number for the crank angle sensor, just in case?

Oops, I mistyped -- above where I said throttle body I mean throttle sensor.

$45 shipped for a throttle body with sensor is pretty good. I hope that fixes your trouble.

The crank angle sensor code came back up after you cleared it? Usually that code appears if the engine stalls out or something like that; it's pretty rare for the sensor itself to just go bad.

Edit: hehe, this post makes my post count 555.

Yup, I let the ecu clear itself while I was working on the car. Battery disconnected pproximately 1.5 hours. Code still showed up upon start up, and it was a smooth, quick start.

I'm wondering how the engine can still run seeing as the ecu has no idea where in the rotation the crankshaft is in?

Congrats, vrg3!! This post makes 655 for me. I gotta stop posting so much.

Knowing the camshaft position it can infer the crankshaft position. It'll end up with a little less precision in ignition timing but it'll work well enough for you not to really notice.

Did you do the actual clear memory procedure, or did you just unplug the battery for 90 minutes? It's possible that the ECU's memory didn't actually get cleared even if it was unplugged. Did you check the codes immediately after clearing them but before driving, to make sure the codes actually did get cleared?

Yeah... I spend way too much time here. It's the first board I've ever been really active on. It's the great technical convo that makes me compulsively check for new messages every 15 minutes. Sigh.

Okay, so as long as the ecu's getting the camshaft signal it should be alright. I checked the codes before I started the engine after I supposedly cleared them. I did the battery disconnect method but I'll try doing the clear all codes method by hooking up the green connectors later today.

Also, since there's a MAF sensor code, shouldn't the car run in limp mode where the rpms are limited to under 3000rpms?

Believe me, I'm glad you check your messages every 15 minutes, cause I need all the advice that you can give me here.

boostjunkie: It is possible, just very hard. What happens is that when you tighten it down the sensor moves! You have to learn to compensate for it. The first time I did it I took a good 25 minutes just readjusting the stupid thing and I still did it wrong (used the wrong screw for the feelers)

I was doing the same thing. I used the top screw instead of the bottom one. But then I tried aligning the sensor with the bottom screw. Either both values showed zero resistance or both values showed some resistance. I realize it's a small margin of error between .0028 and .0035 but DAMN, I should be able to at least zero in on the proper alignment before I tighten it down, right?

Do you have any idea why the throttle sensor resistance moved BACKWARDS from what it should be? Instead of increasing resistance, it decreased in resistance as the throttle was opened.

Most (if not all) the error codes you read out of your ECU were stored in memory but were not active. I can tell you that it's very unlikely your MAF sensor is actually damaged; usually they just won't work at all if they really need to be replaced. Those codes got stored in memory when the ECU got a weird signal from the MAF at some point. I get that code every once in a while when I'm teaching friends to drive in my car, since they stall the engine out. When the engine stalls the ECU gets a little confused and so thinks that some of the data it's getting is wrong.

If the resistance reading seemed to move backwards, you're probably measuring it at the wrong terminals. The sensor is essentially a big resistor whose resistive element has a wiper on it. That wiper can move from one end of the resistor to the other end, as the throttle valve turns. You're probably measuring the resistance between the wiper and the wrong end.

I checked an double checked and triple checked to make sure the correct terminals were being tested.

You guys won't believe this strange GHOST IN THE MACHINE!!!

Started the car up just now and there was no CE LIGHT!! Let the car idle . . . everything seemed okay. Let the car warm up by driving it and noticed it was still driving fine. Decided to open her up to 1.15bar, everything checked out fine!! Even a/f readings were correct. It stalled out a couple times taking the car out of gear and letting the engine slow itself down, but otherwise it drives like before this ever happened!!

I'm going to swap out the throttle body anyways, since most likely the TPS will be properly adjusted on the TB (barring any incidents with shipping). I'm sure the random stalls are due to the TPS being misaligned, and after yesterday's fiasco, I don't want to go through that hassle again!! Besdies, I already paid for the throttle body setup and I can't return it.

That being said, does anyone want the updated knock sensor I got from subaruparts?

If yours is the old one, I'd swap it anyway. The old ones (grey connectors) had a tendency to crack and then become oversensitive. You might have escaped that thus far, but it can't hurt to replace it, you did buy it.

That was one of my replacements in my frantic quest to get rid of my hesitation...

personally.....I'd just swap the TPS.....not the whole throttle body. You may be opening up an entirely new can of worms by doing that.

Why is that? I figure that the TPS that's matched to the throttle body would be more ideal? What are some of the possible issues that could arise from the use of a different throttle body?

I knew you were going to ask this........

On the early throttle bodies, as you have found out, the throttle switch is in the TPS, and that is what needs to be adjusted properly.

If you're getting a used TB, there's no guarantee that the switch will be set right. Someone could've messed with it, or over time, the signal got tweaked. In either case, it's not good to assume the new one would be set perfectly.

Also, there is differences in TB's between the early years and later. The early ones had like a needle valve adjuster to adjust the amount of air bypassing the throttle plate. The later (92+ probably) didn't have that bypass. I've been meaning to check, but most likely there is different part #'s for the TB as well as the IAC valve. Most likely everything has been superceded with the later model stuff.

You can check out this thread for more info and pics/links http://www.bbs.legacycentral.org/viewto ... =black+goo

My idle still is not where it should be after my TB fiasco last year. I've moved the TPS around and re-adjusted it no problem. I'm probably going to try and find a IAC valve from a 92-94 legacy and swap it, and see if that helps things.

I was unaware of the differences. FWIW, I got a TB/TPS from a 93 turbo legacy. I'll see about swapping over the sensor and such . . .

I'll have to check out that link as well. But for now, everything still checks out fine *shrug* No CE light (24 hours and counting).