CEL and Code 24

[syncrosubie]

Hi,

 

I've got a 1993 EJ 22 in my 1991 syncro westy...it's been 10k miles and one good year on the engine conversion installed by a talented professional

(unfortunately I'm far enough away from him that I haven't taken the

van back.)

 

Based on the advice of 3 trusted mechanics, I am starting to believe

that my Code 24 (IAC) has nothing to do with the IAC.

 

Condition:

 

Under periods of no-throttle descending (takes a few seconds), be it

65 mph in 4th gear or 20 mph in 2nd gear, the CEL lights up. It's

always a Code 24.

 

Once the CEL is lit, I get a hesitation, or cut-out, at full

throttle. (If the CEL is NOT lit, no full-throttle hesitation

occurs.)

 

Also, the CEL will re-set itself at the next stoplight or stop

sign...or anytime you put the clutch in and let the revs drop to

idle (well the revs drop beyond idle - gives me a little idle dip

and then shuts off the CEL and returns to normal idle - this idle dip seems to be a common subaru-vanagon issue)

 

I think that it has done this basically since the conversion was

installed...but until I installed a VSS this summer, I didn't

understand the relationship between the full-throttle cutout and the

CEL being lit. (The CEL was almost always on without the VSS so it

was kind of meaningless.)

 

The van may run a little worse when the CEL is lit...but it's not

something my wife would notice!

 

What I've Done:

 

Tried the electrical portion of another IAC - no change

 

Tried to hook up BF scan tool but my ECU wasn't recognized.

 

I thought I tested the TPS at one point and it tested fine.

 

Trusted mechanics have suggested oxygen sensor (I presume it was new

with conversion)and throttle position sensor.

 

I've posted this to the subaru vanagon list, but am hoping that one of you subie fans might be able to help out!

 

thanks much!

 

Tom

[syncrosubie]

Bump....getting a few tidbits from the subie vanagon folks.

 

Anybody know if the injectors shut down when coasting down hill?

 

TPS, VSS, and RPM involved in this situation? Or is the VSS independent?

 

thanks,

 

Tom

 

Hi,

 

I've got a 1993 EJ 22 in my 1991 syncro westy...it's been 10k miles and one good year on the engine conversion installed by a talented professional

(unfortunately I'm far enough away from him that I haven't taken the

van back.)

 

Based on the advice of 3 trusted mechanics, I am starting to believe

that my Code 24 (IAC) has nothing to do with the IAC.

 

Condition:

 

Under periods of no-throttle descending (takes a few seconds), be it

65 mph in 4th gear or 20 mph in 2nd gear, the CEL lights up. It's

always a Code 24.

 

Once the CEL is lit, I get a hesitation, or cut-out, at full

throttle. (If the CEL is NOT lit, no full-throttle hesitation

occurs.)

 

Also, the CEL will re-set itself at the next stoplight or stop

sign...or anytime you put the clutch in and let the revs drop to

idle (well the revs drop beyond idle - gives me a little idle dip

and then shuts off the CEL and returns to normal idle - this idle dip seems to be a common subaru-vanagon issue)

 

I think that it has done this basically since the conversion was

installed...but until I installed a VSS this summer, I didn't

understand the relationship between the full-throttle cutout and the

CEL being lit. (The CEL was almost always on without the VSS so it

was kind of meaningless.)

 

The van may run a little worse when the CEL is lit...but it's not

something my wife would notice!

 

What I've Done:

 

Tried the electrical portion of another IAC - no change

 

Tried to hook up BF scan tool but my ECU wasn't recognized.

 

I thought I tested the TPS at one point and it tested fine.

 

Trusted mechanics have suggested oxygen sensor (I presume it was new

with conversion)and throttle position sensor.

 

I've posted this to the subaru vanagon list, but am hoping that one of you subie fans might be able to help out!

 

thanks much!

 

Tom

[nipper]

hrmmm

 

Injectors may shut down with a closed throttle, thats easy enogh to check with a rigged test light.

Now i dont know how you are wired in, but you are not OBII, your obdI, which means there are diagnostic connectors you need to hook up to get the codes. Those are usually under the dash near the gas pedal, but in your case, they aint there.

How old is the 02 sensor? Another possability is the knock sensor and ignition timming.

 

nipper

[syncrosubie]

Thanks, Nipper.

 

The OBDI port is intact and attached to the ECU under my back seat. I've tried to use the scan tool out of cornell but thus far it has not recognized my ECU...a subievanagon wiring harness modifier (Tom Shiels) made and tested my communication cable so I think that's good.

 

I believe the O2 sensor is new with the conversion...but if I can't get any data out of the scan tool that seems like the next thing to replace...if it doesn't work I'll have a spare. My thought is the O2 sensor is seeing something it doesn't like...perhaps the ECU thinks it's idling when in reality it's just coasting...so it's getting too much gas? Not necessarily a problem with the O2 sensor...

 

What would tell the ECU it's coasting? RPM from cam or crank sensor? or VSS?

 

What are you thinking re the knock sensor triggering the the coasting CEL?

 

again, THANKS!

 

hrmmm

 

Injectors may shut down with a closed throttle, thats easy enogh to check with a rigged test light.

Now i dont know how you are wired in, but you are not OBII, your obdI, which means there are diagnostic connectors you need to hook up to get the codes. Those are usually under the dash near the gas pedal, but in your case, they aint there.

How old is the 02 sensor? Another possability is the knock sensor and ignition timming.

 

nipper

[Legacy777]

The injectors don't shut down. They may go to a minimal duty cycle, but they don't shut off.

 

You need to listen to your ECU. It's throwing a specific engine code for a reason. The OBD1 computers aren't super smart like the OBD2 stuff, so the only reason it'd throw a code is if it's getting a resistance outside its normal range, or the IAC isn't responding like it's supposed to be.

 

I'd suggest checking all the wiring from the IAC valve to the ECU to confirm it's good. I'd also hook up a volt meter to the power source of the IAC valve to make sure power is not fluctuating. From there's I'd get a meter that can measure duty cycle, and watch what the duty % is on each of the open & close leads coming from the ECU. Or try and get the BC-BF scan tool to work.

 

Can you post the part # for your ECU? I'll check it out and see what year it's from.

 

Maybe your ECU is on the fritz...

 

Josh

[nipper]

Theere is no general reader for OBDI cars, just specific readers. You do have some limited diagnostics ability by connecting the diagnostic connectors togehter and going through a specifc procedure (ill look later if you need it).

 

nipper

[syncrosubie]

Thanks, Legacy777. I will get the ECU number and post it...but my van is having some front end work done and I probably won't see it for a week.

 

Should the IAC do anything when coasting downhill at say 60 mph (3200 rpm ish in my van)? If I put it neutral and coast...no code thrown. It seems to me that perhaps it thinks it's idling when in fact it's coasting? But in general it idles fine...

 

Thanks for the other suggestions....power source voltage to IAC seems easy enough...would doing this in the driveway be adequate? or would you monitor while driving and specifically when the CEL is triggered?

 

I hope to have better luck with the BC-BF scan tool.

 

I could probably get a spare ECU from the guy who did the conversion. I have unplugged and replugged all the connections several times...

 

-Tom

 

 

 

 

 

 

 

 

The injectors don't shut down. They may go to a minimal duty cycle, but they don't shut off.

 

You need to listen to your ECU. It's throwing a specific engine code for a reason. The OBD1 computers aren't super smart like the OBD2 stuff, so the only reason it'd throw a code is if it's getting a resistance outside its normal range, or the IAC isn't responding like it's supposed to be.

 

I'd suggest checking all the wiring from the IAC valve to the ECU to confirm it's good. I'd also hook up a volt meter to the power source of the IAC valve to make sure power is not fluctuating. From there's I'd get a meter that can measure duty cycle, and watch what the duty % is on each of the open & close leads coming from the ECU. Or try and get the BC-BF scan tool to work.

 

Can you post the part # for your ECU? I'll check it out and see what year it's from.

 

Maybe your ECU is on the fritz...

 

Josh

[Legacy777]

Yeah, the IAC valve will still have a duty % amount when coasting.

 

You can check it while idling, however it'd provide you the most information if you checked the voltage when the problem/CEL was occuring.

 

Josh

 

Thanks, Legacy777. I will get the ECU number and post it...but my van is having some front end work done and I probably won't see it for a week.

 

Should the IAC do anything when coasting downhill at say 60 mph (3200 rpm ish in my van)? If I put it neutral and coast...no code thrown. It seems to me that perhaps it thinks it's idling when in fact it's coasting? But in general it idles fine...

 

Thanks for the other suggestions....power source voltage to IAC seems easy enough...would doing this in the driveway be adequate? or would you monitor while driving and specifically when the CEL is triggered?

 

I hope to have better luck with the BC-BF scan tool.

 

I could probably get a spare ECU from the guy who did the conversion. I have unplugged and replugged all the connections several times...

 

-Tom

[sea#3]

This may help you . http://www.graphics.cornell.edu/~v/b10scan/

 

Just a question

1) What was the ECM out of standard or automatic ?

 

SEA#3

[syncrosubie]

Thanks SEA#3.

 

I've got the cable to go from the ECU to my old Win95 laptop. And I downloaded the cornell program.

 

But the program didn't recognize my ECU.

 

My van is having a little front end work done so I can't tell you what ECU is in it...but I will report back when I get the info.

 

thanks,

 

Tom

 

 

This may help you . http://www.graphics.cornell.edu/~v/b10scan/

 

Just a question

1) What was the ECM out of standard or automatic ?

 

SEA#3

[syncrosubie]

Got my van back from the front end work.

 

Installed a new to me, supposedly good, IAC. Performs exactly like my old one....CEL on no-throttle descending, etc. bolsters my confidence that IAC is NOT the problem.

 

Tested resistance on both IACs - both at spec.

 

Tested idle portion of TPS - again right at spec.

 

Tested ECU connector with ignition on at pins 1 and 2 of the biggest ECU connector - spec said 6V on one and 7V on the other. I got 12.5 V on both - not convinced I'm doing this right.

 

Did not test wiring between ECU and IAC - needed another pair of hands.

 

ECU had the following info:

 

F9

22611 AA931F

JA18000 RK3 2206

Automotive Electronic Control Systems, Inc.

 

Still can't get Scan tool to work. Will try my buddies laptop and try my laptop on his van.

 

Assuming wiring between ECU and IAC checks ok - and I think it will since the system works most of the time - Not sure where to go from here...02 sensor is cheapest part to buy and try.

 

Tempted by a used/rebuilt ECU.

 

slightly tempted by a new IAC

 

slightly tempted by a new TPS.

 

Should I try the local subaru dealer or independent for some scan info - do they have software to monitor realtime conditions?

 

thanks,

 

Tom

[nipper]

scantool will not work. period. You are an obdI car. Scan tool is for OBDII only. To read obdI you need to follow the proceedures in the manual or get the very expensive dedicated code reader:

 

connect the test mode connector (i do beleive these are green)

Turn the ignition to on

Make sure the check engine light is on

Observe the check engine light. Make sure it lights

 

(book is alittle odd at this point)

 

dissconnect the test connectors (car still on)

Observe the flashes and count them the first long flash will designate the the first digit fo the code, short flashes will designate the second digit.

 

 

i can scan and email the pages to you.

 

nipper

[syncrosubie]

Thanks Nipper.

 

My understanding is the cornell.edu BC/BF free-ware scanning software will work on OBDI...and should work on my ECU.

 

I've got codes...that's the code 24 that I keep getting that won't go away! And doesn't seem to be the IAC.

 

thanks,

 

TG

[nipper]

Thanks Nipper.

 

My understanding is the cornell.edu BC/BF free-ware scanning software will work on OBDI...and should work on my ECU.

 

I've got codes...that's the code 24 that I keep getting that won't go away! And doesn't seem to be the IAC.

 

thanks,

 

TG

 

im loopy from meds, start this again iin the morning so we can play with it.

[Manarius]

Thanks Nipper.

 

My understanding is the cornell.edu BC/BF free-ware scanning software will work on OBDI...and should work on my ECU.

 

I've got codes...that's the code 24 that I keep getting that won't go away! And doesn't seem to be the IAC.

 

thanks,

 

TG

That software will and has worked on the 1990-1994 Subaru Legacy's. Unfortunately, you're going to have to figure out what exactly is wrong with your connector so that you can get the voltage thing figured out. I think if you do that and read that the ECU sees, that will be the insight into what is wrong.

 

Although, your ECU may just be crazy. Sometimes I think mine is.

[OB99W]

The problem could be due to a wiring or voltage supply issue (or possibly a real IAC fault), resulting in an IAC code and temporary shutdown of only one injector. You might want to read the last section (page 30, "Legacy and Impreza Engines with No Injection Pulse #1 Cylinder") of the following:

http://endwrench.com/current/spring04pdfs/InsiderInfo.pdf

[Cougar]

I did a little research and here is what my info says about the ECU for a '93 Legacy.

 

The air control valve has 3 pins. Two of the pins tie to wires that go to ECU pins 1 and 2 of a 26 pin connector. The third wire ties to the common lead of the injectors. Pin 3 of the control valve is tied to a black wire that ties to pin 1 of the ECU and pin 1 of the valve ties to pin 2 of the ECU.

 

How does that match up with what you have?

[Legacy777]

The ECU is from a 92 MY Legacy. Vrg's scan tool should work. You may want to email him.

 

As for the IAC wiring, the spec is different for AT & MT. The MT spec is 12v on pin 2 of the F47 connector....(which goes to 0v after 1min)......and 0v on pin 1 of the F47 connector.

 

Make sure you have pin 1 of the ECU going to pin 3 of the IAC valve, and vice versa, make sure you have pin 2 of the ECU going to pin 1 of the IAC valve.

 

I really don't think your IAC valve or TPS is the issue. I'd suspect either your wiring, or the ECU itself.

[syncrosubie]

Thanks folks!

 

I confirmed the wiring - it's as you described, cougar. I checked continuity of both pins 1 and 2 of the 26 pin connector to the IAC - both were fine....0.3 ohms. The [pin 1(ECU) - pin 3(IAC)] and [pin 1(IAC) - pin 2(ECU)] relationships were correct.

 

I confirmed my previous results - ignition on - Pins 1 and 2 to ground both give ~12.5 Volts. My buddy's van (a 90 AT motor/ECU -not sure if that matters, but wires were the same color) tested the same. Legacy777 - are you sure about your last post i.e. the 0 V at pin 1? How would I know if I have a MT or an AT ECU? My understanding is the ECUs in the 92s work for both.

 

I used my laptop, the scantool, and my cable and it worked PERFECTLY on my buddy's van. So...it seems like my ECU doesn't want to talk to the scan tool. Wiring to the ECU is as shown on the cornell.edu website.

 

I'm close buying an ECU...the subie diagnostic directions point me there. And the fact that I can't talk to the ECU makes me suspicious of it.

 

22611 AA931F ... does the F mean anything?

 

Does anybody know anything about ECUDirect.com?

 

Or other suggestions?

 

still tempted by O2 sensor and ignition relay as these are relatively cheap parts that could be remotely involved, I guess?

 

THANKS for you continued thoughts!

 

-TG

 

 

 

 

The ECU is from a 92 MY Legacy. Vrg's scan tool should work. You may want to email him.

 

As for the IAC wiring, the spec is different for AT & MT. The MT spec is 12v on pin 2 of the F47 connector....(which goes to 0v after 1min)......and 0v on pin 1 of the F47 connector.

 

Make sure you have pin 1 of the ECU going to pin 3 of the IAC valve, and vice versa, make sure you have pin 2 of the ECU going to pin 1 of the IAC valve.

 

I really don't think your IAC valve or TPS is the issue. I'd suspect either your wiring, or the ECU itself.

[sea#3]

Thanks folks!

 

I confirmed the wiring - it's as you described, cougar. I checked continuity of both pins 1 and 2 of the 26 pin connector to the IAC - both were fine....0.3 ohms. The [pin 1(ECU) - pin 3(IAC)] and [pin 1(IAC) - pin 2(ECU)] relationships were correct.

 

I confirmed my previous results - ignition on - Pins 1 and 2 to ground both give ~12.5 Volts. My buddy's van (a 90 AT motor/ECU -not sure if that matters, but wires were the same color) tested the same. Legacy777 - are you sure about your last post i.e. the 0 V at pin 1? How would I know if I have a MT or an AT ECU? My understanding is the ECUs in the 92s work for both.

 

I used my laptop, the scantool, and my cable and it worked PERFECTLY on my buddy's van. So...it seems like my ECU doesn't want to talk to the scan tool. Wiring to the ECU is as shown on the cornell.edu website.

 

I'm close buying an ECU...the subie diagnostic directions point me there. And the fact that I can't talk to the ECU makes me suspicious of it.

 

22611 AA931F ... does the F mean anything?

 

Does anybody know anything about ECUDirect.com?

 

Or other suggestions?

 

still tempted by O2 sensor and ignition relay as these are relatively cheap parts that could be remotely involved, I guess?

 

THANKS for you continued thoughts!

 

-TG

I checked your ECM # and it doesn't designate whether it is Auto or Standard

The only way to tell is with a VIN # (which I presume you don't have)

What may be happening is that it might not be getting a input signal that it is in gear and stalling,

Because of that you are getting a code as well

 

Maybe Legacy 777 could get you the I/O for the ECM on the neutral switch or the PRNDL switch and see what you have

SEA#3

[Legacy777]

Legacy777 - are you sure about your last post i.e. the 0 V at pin 1? How would I know if I have a MT or an AT ECU? My understanding is the ECUs in the 92s work for both.

 

That's what the factory manual said, but who knows....it has been wrong in the past. If you have not done so, or don't have the wiring diagrams. I'd recommend you d/l these scans of the factory manual from me

http://www.main.experiencetherave.com:8080/subaru_manual_scans/1992_FSM_(Engine_%26_Electrical)/ It's about 9.5 mb.

 

22611 AA931F ... does the F mean anything?

 

No the F doesn't mean anything.

 

I'd recommend you post a WTB post in the parts shed on www.bbs.legacycentral.org I'm sure you can find an ECU of there for A LOT cheaper then what you're going to find at ECUdirect.

[Legacy777]

The 92 ECU's went in both AT & MT cars....it doesn't matter. What does matter is that the ECU is correctly pinned so it knows whether it should be controlling the engine for "automatic mode" or "manual mode". It does this via the MT/AT identifier pin on the ECU. On the 92's (or at least the 92 ECU I've dealt with) if that pin is grounded, the ECU is in MT mode. If that pin is floating (nothing connected to it) the ECU is in AT mode.

 

A lot of this info is in my AWD & 5spd swap write-up and may be worth reading if you haven't.

www.surrealmirage.com/subaru/swap/electrical.html

[syncrosubie]

Thanks Legacy777.

 

Not to beat a dead horse...but I'm looking at page 75 (page 202 of 335 of PDF) of your scanned 1992 Service manual ... says 6 V and 7V for non-turbo (I get 12.5 v at both). Am I missing something? This seems to be the only thing that is out of spec...

 

Any reason I care if the ECU is in M/T or A/T mode? I think I can figure that out with all the info you scanned.

 

Thanks for your patience! I paid good money for this conversion so I wouldn't have to learn all of this stuff! Oh well...good for the brain I guess!

 

-TG

 

 

 

 

 

 

 

The 92 ECU's went in both AT & MT cars....it doesn't matter. What does matter is that the ECU is correctly pinned so it knows whether it should be controlling the engine for "automatic mode" or "manual mode". It does this via the MT/AT identifier pin on the ECU. On the 92's (or at least the 92 ECU I've dealt with) if that pin is grounded, the ECU is in MT mode. If that pin is floating (nothing connected to it) the ECU is in AT mode.

 

A lot of this info is in my AWD & 5spd swap write-up and may be worth reading if you haven't.

www.surrealmirage.com/subaru/swap/electrical.html

[Cougar]

It sounds like the ECU pins for the valve are tied to a open ended circuit and are ok. The communication lines appear to be bad though so replacing the ECU may fix that.

 

It sounds like the lead that runs between the valve and the injector common lead may have something to do with this. I am guessing that it is a return lead for the valve circuit somehow.

[syncrosubie]

I ordered an ECU from a subaru - suzuki dismantler. $150. Will let you know how it goes.

 

edited - found legacy's discussion of MT/AT pin...will investigate what I got.

 

Printed out the engine electrical wiring diagram - but I can't figure out whre the wires go when they hit the small numbered boxes (i.e. Section 6-3 (0504), page 19) numbers 1 thru 11. Seems like I need a Fold-out?

 

Keeping my fingers crossed that I can plug and play!

 

thanks,

 

TG