legacycentral bbs

I'm the new owner of Kelly's old Radio Flyer, the red '91 SS.

Here's what I knew when I bought it.
http://bbs.legacycentral.org/viewtopic. ... 25#p325525

So after a 1000 mile adventure in one day that included inability to boost and accelerate, losing the clutch pedal, no marker lights and a horrible alignment wobble, I'm back home with the car. Time to fix the problems, starting with the big one; the engine cuts out badly over a certain throttle position.

I've replaced the coil and igniter, no fix. Checked spark plugs, burning clean on good plugs. I tried unplugging the knock sensor after learning that a previous owner had a similar issue that was fixed with the knock sensor unplugged; no fix. I ran the car in D-Check mode. There are no vacuum/boost leaks. Nothing I have tried has affected the cutting out in any way. I still have a Perfect Power 6 harness wired into the car, but the problem is apparent even with the bridging plug in, which is how I've been running it and testing things.

Problem description:
At first we thought it was throttle position dependent. That is not true. I hooked up a boost gauge to it, and it is manifold pressure dependent, meaning that anytime I press the throttle enough to make the car come near positive pressure, it cuts out. I can make it do this in neutral, or in any gear. As soon as the needle on the boost gauge approaches 0 on it's way towards positive pressure, the car dies. Still has spark, but no fuel. I can drive it staying in vacuum and it runs great. It idles great. It gets great gas mileage. I made a 500 mile drive with it staying in vacuum the whole way yesterday and got around 25-26mpg. The cutting out is an absolute fuel cut. As long as I keep the throttle open enough to make it go near atmospheric pressure, the car just drops fuel and dies. In neutral, I can rev it, but as soon as I open the throttle all the way vacuum drops to 0 and the engine cuts fuel and dies.

It has several codes, some caused by me and some caused by missing/inoperative sensors. I can't get the codes to clear. I've left it with the battery unplugged for over an hour, and tried the full D-Check reset, which may not work because I'm lacking a speedo cable.

Code 22 - Abnormal knock sensor voltage - I caused this one by unplugging the knock sensor to try to fix the problem.

Code 23 - Abnormal MAF voltage - I caused this as well, by unplugging the MAF to verify it's functionality.

Code 31 - Abnormal TPS voltage - Caused by myself in testing.

Code 35 - Canister purge inoperative - There is no charcoal canister. The lines are looped on the purge valve, which is still plugged in. How can I make this go away?

Code 44 - Wastegate duty solenoid valve inoperative - No BCS on this car, it is running off of wastegate pressure right now and I have no BCS, none came with the car. Could this cause a problem?

Code 45 - Pressure Sensor Inoperative - This may be the cause of my problem. There was no vacuum/pressure line going to the two sensors on the passenger's side strut tower. I ran a manifold vacuum/pressure line to them, but the problem persists. I can't clear the code though, It seems

Code 49 - Improper Airflow Sensor - I don't know why this code is showing, maybe because I unplugged the MAF? It has the proper stock aluminum MAF.

I need some help! Right now my plan is to take care of all the codes that I can and hope that one of them is causing my problem. I'm off for now to drive in D-Check and see what codes it throws while driving. I will update with current D-Check codes when I get back.

EDIT:

Drove in D-Check until only three codes were blinking.

Code 33 - VSS - Duh, no speedo cable. Will be fixed.

Code 44 - Duty solenoid valve inop. - No BCS in the car. I need to get one if only to make the computer happy. Or trick it somehow?

Code 51 - Abnormal signal from neutral switch - Don't have the harness for this in the car. I need to wire this up.

Only the Code 44 should affect my fuel cut I would think. I believe all of the previous codes remain stored as well if I just check the codes with the black connector under the dash.

Help!

There's a few things....

I'd start with the codes you got in d-check.

Get the speedo working before you do anything ANYTHING further! I can not stress this enough. The stock ECU does some incredibily screwing stuff without a speed signal.

Regarding the wastegate solenoid, get a dropping resistor from an automatic trans car and plug it into the wastegate solenoid connector. The dropping resistor will even bolt onto the back of the bracket on the passenger strut tower.
http://www.main.experiencetherave.com/s ... P_7903.JPG

Neutral switch shouldn't cause the issues you're having, but I'd get one wired in.

Once you fix those, clear the codes, and run another check of the codes and d-check.

Legacy777 wrote:There's a few things....

I'd start with the codes you got in d-check.

Yeah, I figured those are the real problem. I just can't clear the others because those three prevent D-Check from clearing the ECU.

Legacy777 wrote:Get the speedo working before you do anything ANYTHING further! I can not stress this enough. The stock ECU does some incredibily screwing stuff without a speed signal.

I agree. Going to the junkyard tomorrow after work to pick up a few things, I'll make sure to have the speedo cable be a priority. I'm starting to doubt that it is causing the issue, but it is a top priority for sure.

Legacy777 wrote:Regarding the wastegate solenoid, get a dropping resistor from an automatic trans car and plug it into the wastegate solenoid connector. The dropping resistor will even bolt onto the back of the bracket on the passenger strut tower.
http://www.main.experiencetherave.com/s ... P_7903.JPG

Where is this located on an automatic trans car? Do N/A auto Subarus have this as well, or only turbos?

Legacy777 wrote:Neutral switch shouldn't cause the issues you're having, but I'd get one wired in.

Yeah, will do. I might need to in order to let D-Check clear the rest of the codes. I also have the parallel cable hooked up to the SSM interface, I just need to get Vikash's scan tool onto a floppy and get my ancient laptop out there to monitor the ECU. This may give me some helpful hints as well.



Here's why I don't believe that the speedo cable is causing my issues. Rather, I believe it is related to the fuel system somehow:

I hooked the PP6 back up and zeroed Kelly's tables out. This made the car run the same as if it had the bridging plug installed. I watched the MAF voltage, and noticed that over a certain V was causing the car to cut out. The PP6 is using the MAF V as it's deflection (load) modifier. I started messing around with the timing and fuel at that point and above, just a little bit. I noticed that retarding timing had no effect except making the car rev slower until cutting out. Adding fuel via adding MAF V made it cut even worse. Subtracting fuel bit by bit made it run better! Now, I'm not one to tempt fate without a wideband (yet), but I got it to free rev at full throttle without cutting out and was building 1-2psi of boost in neutral as well! Very interesting... I pulled some timing just to be safe and went for a short drive, and verified that I can maintain 1-2psi in the first couple gears before cutting out. This bit of news makes me suspect something in the fuel system as being messed up. I zeroed the tables out again because obviously there's something wrong here, and tuning is absolutely not the way to fix it. I yanked the vacuum line off the fuel regulator, and verified that it has vacuum. With the line off the regulator, the car cut out and bucked very very badly on throttle opening. Line back on, car smooths out a bit and I can rev it at part throttle again. This is the first time I've had an unexpected change in how the engine stumbles by unplugging something. It's too dark today, but I will trace the FPR line and make sure it's on a true manifold vacuum/pressure source.

Seems fuel pressure/supply related to me now. I need to pick up a fuel pressure gauge kit and verify that my fuel pressure is within spec. Need to verify operation of the pressure regulator as well. I have stock pinktops and a stock MAF on the car. Seems odd to me that a fuel pressure problem would cause it to cut out so abruptly and completely at just under atmospheric pressure.

I feel that I'm really close to figuring out what's going on here. Hopefully I'll be proven wrong about my suspicions and a speedo cable will fix it, but if not I think I may be on the right track.

Thanks for the help so far.

Sounds like you're making progress. Glad to see that the car is getting the TLC it deserves .

A note about the FPR. The mechanical functioning of the FPR and injectors should be just fine from the experiences I had with them. When I swapped motors from my 93' I swapped the whole long block. I took of the intake manifold to swap over harnesses but then put it back as-was. From your PM I apparently missed something if you found a vac leak on the way to your home (hopefully it wasn't something too blatent and my mechanic abilities come into question ).

Feel free to change out the FPR and such, just a word of note they ran great when I had the motor, that's all.

And the dropping resistor is on the back of the bracket that holds the MAP sensor and BCS. Or that's where it was for me.

Thanks Lee, good to know the FPR and injectors are from your car and were working at the time of the swap. The vacuum leak was nothing too major, just had a clamp missing on the F-Tube and it was alternately spewing oil onto the top of the motor and sucking in unmetered air. I stuck a clamp on it and the top of the motor dried out and it ran a little better off-boost.

After sleeping on it, I tentatively suspend my suspicion of a fuel problem because all my negative numbers in the PP6 were doing was pulling MAF V, and it seems to cut out at a specific MAF V so I was in effect tricking it into boosting a little bit and making myself feel smart.

Priority 1 is to fix all the codes. Then, if it still doesn't run right, I'll focus on other things. Hopefully I'll be able to find the right dropping resistor, neutral/reverse harness and speedo cable on an N/A Subaru, that's all I have access to here.

Thanks for the input!

Yes the dropping resistor can be found on the non-turbo automatic cars. They will be over on the passenger side strut tower.

Alphius wrote:Priority 1 is to fix all the codes. Then, if it still doesn't run right, I'll focus on other things. Hopefully I'll be able to find the right dropping resistor, neutral/reverse harness and speedo cable on an N/A Subaru, that's all I have access to here.

This method is definitely going to be the best way to approach this.

Let us know how things are coming along as you work on getting the codes addressed.

Got a speedo cable, hooked up my reverse/neutral harness, and stuck a dropping resistor in place of the boost solenoid.

D-Check shows one code remaining. Code 44, for the wastegate solenoid still; although I do have the dropping resistor plugged in. I think I'm going to have to get a stock BCS to make D-Check get rid of that last code. Am I correct in my suspicion that D-Check will only reset the ECU if it detects zero problems? I still have all the same codes as before just by doing the regular code check procedure. Can't seem to clear them, and I have a suspicion they are causing my problem. I'm leaving the battery disconnected overnight and through tomorrow until I get home from work to see if that will clear the codes.

I also ran b10scan and watched all the parameters. The ECU sees all the different inputs successfully, and all the values make sense. When I go over atmospheric pressure, the injector duty cycle drops to 0.000 and the car stalls. Everything else looks normal.

I'm pretty stumped.

EDIT:

Slept on it.

Code 45 - Faulty sensor or pressure exchange solenoid valve inoperative - Prevents abnormal supercharging pressure using "fuel cut" in relation to engine load

I think that may be my problem, since it does seem load dependent after watching the scan tool last night. When I first got the car those two sensors had no vacuum line going to them. Problem is, I can't clear the code because D-Check still sees the BCS problem.

I left the battery disconnected all night, and it won't be connected until I get home from work this evening. Maybe it will clear all my codes for me.

EDIT2:

Evidence:
http://bbs.legacycentral.org/viewtopic. ... 18#p210518
http://bbs.legacycentral.org/viewtopic. ... 793#p16793

The dropping resistor should be alright. Check the resistance of the dropping resistor. It may be too low and may be causing the problem.

The resistance of the wastegate solenoid is around 20 ohms. So if you got a resistor that is too low, it still may throw a code. What year and model subaru did you get it off of?

As for clearing the codes, it would be best if you can try and not have codes come up in d-check. If you get codes in d-check, you may not be able to clear the memory.

I'd recommend trying to connect both the green & black connectors together to clear the memory. If you try that process and it won't let you clear the codes, than there's an active code (which you already know you've got) and you need to try and fix it first.

You may also want to check the wiring from the wastegate solenoid to the ECU to make sure it's good, and that you have 12v on one of the pins of the wastegate solenoid when the ignition is on.

I will check the resistance of the resistor this afternoon. I got two because I found a couple different plug styles depending on the year of the car and I wanted to make sure I had one that would plug in. The one that's in there now (and plugged right in) is from a '92 L Auto Wagon. I have one with the wrong plug from a '97 L Auto that I can test the resistance of and wire it in if the other one is bad.

Got to fix the active codes to clear the memory. I need to get rid of the Wastegate Duty Solenoid Valve code. I may go to Radio Shack and get a resistor if I can't make the Subaru pack work. Checking the wiring to the BCS connector will be done as well.

Do you know what the difference is between green+black connected and just green connected for D-Check mode?

I have a feeling if I get this stuff fixed I'll be golden and the car will run right.

Both dropping resistors measure 13 ohms. One terminal on my plug has 12v. Other one has continuity to the ECU. Letting it sit unplugged overnight did reset my codes, I've now only got the code 44 for the BCS remaining. The smallest resistor I have around the house right now is a 10k, which I'm guessing will not work. I'm about to look up the FSM tolerance for the BCS and go buy a resistor to try to fool the ECU.

With just code 44, I still have the fuel cut exactly the same as before.

EDIT:
Just went and tried D-Check with a 50-ohm and 100-ohm resistor, I'm still getting BCS codes. I think I need to get a stock BCS. Dropping resistor did not work either.

EDIT2:
Lee is going to send me an extra BCS. Hopefully it fixes the problem. I should know by the end of the week.

My car did that exact same thing when there was a line off the bcs. Plugged the line back in and ran fine.

Jeff

yeah....try putting the BCS back in. That's funny that the dropping resistor isn't working. Unless they changed the tolerances in the ECU for different years....that's the only reason I can see it wouldn't have worked.

Anyway, let us know how it runs when you get the BCS.

Bad news...

Still throws a code 44 even with the BCS plugged in. BCS measures 13 ohms, and was working fine on Lee's car. I even ran the vacuum lines to it like stock and drove it in D-Check until only the one code 44 was blinking. Still cuts out over a certain load.

So, the power wire to the BCS that shares power with the Pressure Exchange Solenoid has 12v with the key on. The other wire has 0 resistance all the way to the ECU on the correct pin. There is infinite resistance between the two wires, so they aren't shorted to each other anywhere in the harness. In D-Check with the key on/car off, all the solenoids are supposed to click. The Pressure Exchange Solenoid does click, yet the BCS does not.

This doesn't make any sense, because the wiring checks out but it is acting like there is a broken wire to the ECU or something similar. Is it possible my ECU is messed up? I'm running out of ideas and things to test. The good news is that if I can figure this out, most likely the dropping resistor will in fact work fine to fool the ECU.

Gah, this is frustrating.

EDIT:

Also, with b10scan running on a laptop I can see that the ECU is continually commanding different duty cycles on the BCS, with or without the BCS actually plugged in. Also, I'm gonna clear the codes by unhooking the battery again tonight and see if the code 44 comes back with everything hooked up like it should be. Grasping at straws here, but maybe it'll work then. Probably wishful thinking.

EDIT2:

Morning time now. ECU was reset overnight, then everything was hooked up this morning and BCS code came back within 2 seconds of starting the car. I double-checked all of the wiring and it is all correct. Last thing I can think of is to try to put either the BCS or dropping resistor into the wiring at the ECU under the dash. Could I just hook it to any old keyed 12v under there then hook the other side up to the BCS pin on the ECU?

Yeah that sounds a little funny.

What you described the scan tool doing is normal.

If you try the BCS at the ECU and it still doesn't work, then I'd suggest replacing the ECU. If it does work however, I'd suggest double checking your 12v source at the BCS with engine running to make sure it's not dipping low or cutting out.

Also, you can try running 12v through the wire at the BCS that goes to the ECU and hook it up to a light bulb or something, and see what voltage you get at the ECU connector (disconnect the ECU connector when testing this). The wire may show 0 continuity with a meter, but when you put more amperage through it, like a solenoid or light bulb, the wire may have issues due to degradation.

OK, I left the stock BCS wiring attached to the ECU. I then used a tap connector and ran my own line off of the wire at the ECU for the signal to the BCS and ran the other line to the fuse panel for power to the BCS.

I blew 20A fuses after starting the car. I tried twice, and blew two fuses. Each time, it started with the Check Engine light off then it flicked on after about 2 seconds of idling. Figuring that there is a short to ground somewhere in the harness, I then snipped off the stock wiring so my BCS wiring from the ECU to the BCS and then to the fusebox was the only BCS wiring attached to the ECU.

It stopped blowing fuses. Still does not work, however.

My supposition is that the short to ground in the BCS line has somehow killed the part of the ECU that controls the BCS. Therefore I should try another ECU.

Ugh. Any other suggestions?

That sounds like a plausible scenario.

Trying a different ECU sounds like the next step.

Trying to find another ECU. Harder than I thought it would be.

In the meantime, I've made the discovery that the BCS line in the ECU is grounded. All the time. Even with the key off and no power to the ECU. I can slide the connector in and out on the BCS and feel it click, however when it's plugged all the way in no clicky. I'm more and more confident that this is the problem.

Need an ECU...

It works!

Thanks to Blackbart the new ECU fixed it. Old one must be shorted internally to ground on the BCS line.

Small problem, the dropping resistor I am using instead of the BCS gets hot. Too hot to touch. Should this be OK, or should I rework my system to use the actual BCS?

Woohoo! Success!

, Awesome! Frank saves yet another Legacy .

I would suggest using the stock BCS if the dropping resistor is getting that hot. It's probably getting hard on the wiring and eventually the resistor will fail - or so I would think.

I know using the BCS is my plan. Just out of curiousity, does the car throw a code with the resistor and BCS disconnected completely?


And congrats on getting it sorted out. It's good to see the Radio Flyer running strong again. Now you just need injectors and bigger turbo .

I rewired my connector so it is under the dash, I mostly did that to isolate any wiring problems in the stock wiring. I'll have to rewire it out to under the hood if I use the stock BCS instead of the resistor, because that BCS clicks like crazy anytime I approach boost. It would be very annoying under the dash. I think I'll run it with the resistor for a few days and use my infrared thermometer to see exactly how hot it gets and whether or not the heat is detrimental to it's lifespan.

Using some electrical calculations, the resistor is dissipating around 15 watts of energy depending on the alternator's output. Based on the size and weight of the resistor, I'd say it's probably rated for a bit more than that, and should be fine. I'll just run it for a while like it is now.

It does throw the code and run terribly again if neither the BCS or resistor are hooked up.

Maybe in a couple paychecks I'll pick up a wideband, turbo, injectors and MAF and get to tuning so I can chase some STi guys around. Until then, 10psi with what I've got now is plenty of fun.

Thanks for the help guys! Especially Legacy777. You've all been a great resource so far.

Use the resistor. It should be fine. That's what it's job was originally designed to do. It's tied into the trans to absorb duty cycle to soften shifts.

Mounting the resistor to the metal bracket will help in dissappating heat.