Time to play "Which sensor's broken?"!!!
 

Howdy everybody. I've been having issues getting my SVX to run properly after letting it sit for probably 6-10 months. I'm having a new issue with some pretty unique characteristics that I'm hoping somebody can ID the cause of.

The car starts every time...better than it has in a while thanks to a new battery and clean terminals all around. It idles pretty well when it's cold, also...a bit rough but I think that's related to my 5MT mount more than anything else.

The problem starts as the engine gets warm. The idle slowly drops and drops and drops until it's just barely running (400-500RPM). If I blip the throttle it'll come back to life, then slowly settle lower and lower and lower. Eventually it will stall. Sometimes when I blip the throttle it just dies immediately.

I pulled the diagnostic connector to see what codes I have, and this is when it got a bit more interesting. As I held the engine at 3000 RPM it would stay there no prob for about 15-20 seconds, then it would completely cut power, RPM would drop, and no response from the a-pedal to revive it. About 1-2 seconds later if I was still on the gas it would respond again for about .5sec, rev up, and then cut out again. If I keep my foot in the gas it'll keep cutting in and out in this sequence seemingly indefinitely. If I let my foot off the gas it will stall.

I've been digging through old threads and the closest thing I can find is the Cam Position Sensor. Is that a valid match to these symptoms?

I have a small crack at the base of one of the vacuum hose connections coming out of the end of the main air intake tubing on the driver's side, but I can't believe that would cause the cutout and stalling. It seems like a definite sudden ECU action of cutting all power for 2 seconds, then giving it back for .5 sec, then cutting it, giving it back, etc. A definite limp mode.

Your help is very much appreciated!

Re: Time to play "Which sensor's broken?"!!!
 

alternator?

Re: Time to play "Which sensor's broken?"!!!
 

interesting you say that because i have other odd things going on in connection to the alternator. Why was that your thought?

Here's another thing that happens that points at the alternator:

When I disconnect my alternator wiring mod, everything goes dead. My car needs the alternator wiring mod's alternator + to battery + wire in order to function.

Re: Time to play "Which sensor's broken?"!!!
 

I would check your alternator grounds then if it only works when doing the wire mod.

Re: Time to play "Which sensor's broken?"!!!
 

I have a simple little thought that may help

If you have a car battery charger, charge the battery right up and then re connect the battery to the car and see what happens, if everything's cool for a couple of hours and then gets all screwy it's probably the alternator.

In the next month or so, I'm changing, the crank, cam and cam sensors, my car idles pretty low with no stalling but with other issues, will let you know what happens.

Re: Time to play "Which sensor's broken?"!!!
 

I've checked the alternator output, and it's @ 14.4V, so it seems healthy, but I could watch it when the engine cuts out and see if anything is changing.

Also, where are the cam position sensors?

The frustrating part is that I can't rev it long enough to pull codes. I may just give up and take it to Subaru.

what would cause a full engine power cut like I'm seeing? Seems like the ECU is trying to protect something.

Re: Time to play "Which sensor's broken?"!!!
 

There are 2 Crank sensors right behind the alternator and one Cam sensor next to the battery but it seems the 2 crank sensors behind the alternator are actually 1x Crank and 1x Cam but it's still called "Crank". It's just the same part number as the Cam sensor next to the battery.

Keep in mind 14.4v is at the higher border line.

Check MAF

Re: Time to play "Which sensor's broken?"!!!
 

Is the 14.4v with the alternator wire mod? When I had mine on it was at 14v without it was 13.8v

Re: Time to play "Which sensor's broken?"!!!
 

More symptoms to work with:

When it stalls I hear a relay clicking on and off, and the cooling fans are turning on and off.

Today I went to drive it around and see how long it would work properly. When I started it it was idling horribly. I got under the hood and disconnected fuel injectors, and none of the passenger side injectors caused any change in the idle. So, I turned off the car, pulled the front passenger side plug to see what brand they are and to see if there was anything noticeable about its tip (NGK R plugs). Then I realized I can't tell anything by the plug tip =) So, I put it back in, and re-started. It idled fine the second time!

So frustrating.

The second time I let it idle until it died. The idle speed just slowly gets lower and lower and lower until it dies. Once it gets below about 500rpm, if I tap the throttle to revive it, it'll sputter and die.

Re: Time to play "Which sensor's broken?"!!!
 

I am fairly sure you will find that you have a faulty water temperature sensor, or associated wiring. Don't mess about, check the fault codes. Easy to do, instructions here in the how too stuff. If I am correct, you will get two long flashes and one short repeating, i.e. 21.

Re: Time to play "Which sensor's broken?"!!!
 

like I said, I can't pull the codes because I can't rev for 40sec. Maybe if I do it RIGHT when it's cold I can rev for 40sec before it dies.

Would the coolant sensor explain why this only happens when it's warm?

Re: Time to play "Which sensor's broken?"!!!
 

I have exactly same symptoms. I'm confused though...you talk about having to rev your car for 40 secs? The test I did was simply put one of the blue leads into the smaller black conector ( top, 2nd from left ), then turn on ignition till the Check Light starts blinking. Counting the code mine was '32', o2 Sensor. Am I doing it right? Is this revving one you talk about something different?
I have also been told that I should clean my MAF Sensor first as it can sometimes be the problem.
Incidently, I too looked at the Alternator as an issue. When I sprayed connection points with WD40, it did lessen symptoms somewhat.

Re: Time to play "Which sensor's broken?"!!!
 

Yes this is the correct process, I'm not sure where he got the idea he needs to rev the engine:confused:

Re: Time to play "Which sensor's broken?"!!!
 

Plugging in without revving pulls codes from memory. You are correct. Revving performs a MPFI diagnostic check if I'm not mistaken.

Regardless, I made some interesting progress tonight that goes back to what somebody mentioned above.


Alternator.


I started the car and it was idling fine @ 14.4V from the alternator + terminal. I watched alternator voltage when it started sputtering and it was down at 12.6V. So, it seems like the alternator is kaput. I wonder why it would die as it got warm?

I'll double check tomorrow, but it seems like the alternator puts out 14.4V when cold, but not when warm.

I still have the issue that when I remove the alternator + to battery + portion of the alternator wiring mod, everything goes dead. This would imply that one of the white wires running from the alternator + terminal is broken somewhere, right?

Thinking just about the alternator, does anybody have any feedback ast o the mechanism that could be causing this?


Thanks

Re: Time to play "Which sensor's broken?"!!!
 

If the sputtering reduced the idle rpm substantially, that could lower the voltage output from the alternator... Normally, I think.

Keith:cool:

Re: Time to play "Which sensor's broken?"!!!
 

Maybe :lol:

Keith:cool:

Re: Time to play "Which sensor's broken?"!!!
 

Yes keith,

The cart is before the horse. :lol:

Sputtering of the engine will cause the alternator to run slow (P.S. Below normal idle speed.) and beyond the point that it will produce sufficient voltage, so as to be above battery voltage. It is not a case of the alternator getting warm, it is the engine temperature rising to a point whereby the engine falters and as a result the alternator slows and stops charging. :p ( P.S. On this basis the alternator is not the cause of the problem.)

The fault whatever it is, if it has existed, will be retained in memory. You do not have to even start the engine before reading codes, much less rev it. Check for codes after simply turning on the ignition as per instructions. ;) (P.S. A check on the codes should provide an answer.)

Hooooray, Trevor.

Re: Time to play "Which sensor's broken?"!!!
 

Yea...that thought occurred to me last night after I posted. I'll check and see which comes first...the low idle or the low voltage. I'll also check it by seeing how it runs while connected to another vehicle via jumper cables. The reason my brain was zeroing in on the alternator was because of what the Autozone guy, of all people, told me. He said that when they run their alternator tests some of them will idle fine and put out the proper voltage, but their procedure is to run them @ 2000RPM under load. He says that some will gradually drop voltage until they don't supply enough to run the vehicle properly. So, if I hook the SVX to a healthy car via cables it should at least change the behavior. That would allow me to at least focus on the charging system for further diagnosis.

BTW, where are the OE alternator grounding points? Still trying to figure out why the car goes dead without that alternator mod cable (alt + to battery +)

Re: Time to play "Which sensor's broken?"!!!
 

The alternator is bolted to the engine, the engine is grounded, as will be proven by the starter working.

It would appear that when the mod. was carried out, the original wires connected to the main terminal on the alternator were possibly disconnected. This would leave only the alternator to battery wire to complete the required circuit.

Alternatively, the wire battery to fuse box may not be connected to the battery, again breaking a circuit.

Why do you insist that the alternator could be causing an idling problem? :confused:

You state in post #1 :-

I pulled the diagnostic connector to see what codes I have, and this is when it got a bit more interesting. As I held the engine at 3000 RPM it would stay there no prob for about 15-20 seconds, then it would completely cut power, RPM would drop, and no response from the a-pedal to revive it. About 1-2 seconds later if I was still on the gas it would respond again for about .5sec, rev up, and then cut out again. If I keep my foot in the gas it'll keep cutting in and out in this sequence seemingly indefinitely. If I let my foot off the gas it will stall.

There is no diagnostic connector to pull (disconnect ?), when checking for codes. If you disconnected something you could have caused your “more interesting” problem. :D

You also state “The problem starts as the engine gets warm.” Surely this is the route cause of your fault. ;)

I now leave you to it. ;)

Re: Time to play "Which sensor's broken?"!!!
 

Sorry...didn't pull any connector. Pulled codes. I guess I meant I pulled the connector and wires out from the panel. Didn't actually disconnect anything.

I'm not insisting that the alternator is at fault...just at a bit of a loss and trying to troubleshoot what could be causing my problems. In doing so, I'm thinking about what has changed since it was working properly.

New spark plugs - could crappy plugs cause this (NGK R plugs)?
New timing belt - it's too smooth when it's running for timing to be off. I also triple checked that when I had the engine out for work.
Rebuilt alternator - What are symptoms of failure? Why is the circuit broken when I disconnect the alternator wiring mod + to + cable?
New valve cover, cam, and rear seals - Don't see how these would relate to my problem.

Based on what Auto Zone explained, I feel that I need to check the alternator output voltage while revving. If the voltage isn't steady and trails off even with revs as they said it would if the alternator was bad, then I have a good suspect. I will also try running the vehicle while attached to a jump as that would give me good info as well. The first may show good alternator voltage and the second may have no effect to stalling, but they're both free and easy and will give me confidence in the rebuilt alternator. However, if the alternator shows a voltage drop while revving then it's pretty obvious what's going on, and if the jump keeps the vehicle running smoothly then it's virtually solved. The only question then is why it occurs only when warm.

I cleaned the IAC by removing the TB and spraying with carb cleaner. The AAC valve has been working fine all along. My battery is brand new. It starts great every time and idles fine until it's warm. I can still try the engine temp sensor, but if the alternator doesn't check out, then why spend the $30 on the temp sensor?

Thank goodness I went all in and bought a new daily driver before I did all the work I did on the SVX!

And the engine revving diagnostics are based on the method from the ancient Kinetic Concepts web page document. I pulled the codes from memory and got vehicle speed sensor #2 (33), P switch (51), and N switch (52). If I'm not mistaken, these should all be caused by the MT.

So, thanks for the input, I'm not trying to be stubborn with the alternator. I'm just running out of logical options and the replace and check method gets costly fast.

I'll update after doing some more testing, probably tomorrow night. If anybody has further ideas I'd like to hear them!

Re: Time to play "Which sensor's broken?"!!!
 

Always keep in mind that under certain circumstances some of the codes could be ones stored for a while, could have already been corrected, and not cleared. I haven't time to read everything here today so if this is already covered, sorry!

Keith:cool:

Re: Time to play "Which sensor's broken?"!!!
 

O.K. all understood. ;)

Recently I checked for codes on my SVX and got a strange read out. Clearing and second check a couple of days later, produced logical codes.

It could be well worth your while to run a second check as this would not be at all arduous. Give the gremlins the chance of a second opinion. :D

Re: Time to play "Which sensor's broken?"!!!
 

OK...got it all figured out :)

1st - alternator tested OK @ AZ and it also held voltage even as vehicle was sputtering and dying

2nd - I followed the 2 white wires from the alternator into the fuse box and found that my fusible link was actually broken and the car was only running because of the alternator wiring mod + to + wire bypassing the fusible link. Put in a new link and now I don't need the wiring mod to keep power.

3rd - checked my passenger's side coils and plugs to determine why they occasionally misfire. The coils look fine. Plugs look OK, but I'm not sure what to look for. They appear to possibly have some oil on them. Could that be from not doing the valve cover gasket properly, or pinching it or something? I suppose the oil could be flooding the cylinder somewhat causing a misfire. Still not sure on that one:confused:

4th - unplugged the engine coolant temp sensor (two wire sensor) and it ran fine. I had noticed that the fans NEVER came on with that plugged in so I figured I'd unplug it and see if anything changed. Well, alot changed. Fans came on and engine didn't stall.


So, Trevor, you win the grand prize! Temperature sensor it is!

Re: Time to play "Which sensor's broken?"!!!
 

Sad to resurrect this thread, but the stalling has NOT been cured :confused: But, a friend at a local shop hooked the SVX up to an OBD1 reader and it gave a crank angle sensor code.

So, could that be the issue? To re-summarize what has happened so far

1. Start and idle cold no problem, but slowly drop RPMs as car warms up eventually stalling once temp is reached
2. Start and idle horribly rough with a random assortment of fuel injectors or spark plugs not working. For example, one time none of the right side bank fuel injectors were working. Turned it off, turned it back on, and all was well. Today none of the spark plugs were working. Turned it off, waited a bit, and turned it back on...all was well.
3. Start and idle cold no problem, begin to stall, but save itself when I unplug the Coolant Temp sensor
4. Start and idle cold no problem, begin to stall, proceed to stall even when I unplug the coolant temp sensor.

As you can see...a random assortment of stall or rough idle scenarios.

Looking for more input...

Re: Time to play "Which sensor's broken?"!!!
 

Yeah, absolutely. Check the wiring first though, after I removed my alternator I discovered the insulation right at the CAS connector had become brittle and split open, exposing the bare wire. I fixed it using some brush-on wire insulation before it became a problem. A malfunction with the CAS will DEFINITELY stall your engine... thats gotta be it.

Remove the sensor and inspect it, sometimes they can collect a lot of crap on the magnetic end, you may be able to clean it. From the intermittent nature of your problem, take a close look at the connection too, it seems like theres a hairline fracture of the wiring, or the connector lead, or just a corroded connection.

All in all, there is a 90% probability that this is your problem, so check it, maybe clean it, or replace it if everything else looks good. The ECU is extremely reliant on the CAS for fuel and spark management, as it provides the ECU with the position signal for top-dead-center.

Re: Time to play "Which sensor's broken?"!!!
 

Hmm just a couple thoughts/suggestions. First on newer cars it isnt really recommended to pull the alternator as that can cause a voltage spike that has the potential to blow circuits, fuses, or various electrical components that dont like spikes in voltage. This is more precautionary than anything but felt it would be good to state.


2nd- As has been stated the CAS would definitely account for your problems.

3rd- Have you tried inspecting and testing the MAF? What about cleaning it with MAF cleaner?

4th- I for the longest time had a bypass air control solenoid code on the old engine and it would cause poor idle, and random stalling. Something to think about but prob. not your issue.

5th- When i put the most recent engine in the pass. side bank of injectors were not injecting fuel. The injectors were clogged and tapping on them while the engine was barely running cleared them out. Again prob. not your problem but not hard at all to tap the injector while its running to see if it changes anything.

6th- I believe a while ago someone was battling the electrical connector in the center of the firewall that sends the signal to each side of the engine for either fuel or spark cant remember.

Little long winded but I figured it couldnt hurt to give some suggestions as I know how crappy it is to deal with this. I am guessing you are a 5spd. Have you got a lightened flywheel or is it stock? Just a thought that with a lightened one like i have if you come off the gas totally quickly it will just barely catch the rpms in time to stop it from stalling. I tweaked my driving habits and got the stall issue much more in control. Again sorry for the length. Good luck with your search

Re: Time to play "Which sensor's broken?"!!!
 

also, do you have a digital multi-meter? If not that is a very valuable tool to have in general and especially on these cars as they are getting to the age that wires are gonna start getting very brittle.

Re: Time to play "Which sensor's broken?"!!!
 

Check the ground on the back of the manifold!
double check it

A way left field but possible cause for these systems could be restricted exhaust. It's almost textbook symptoms of it. You could have a plugged cat or something. It's highly unlikely but possible.

Just giving new ideas.

Also, I have never seen an SVX with an ECT failure (i know it happens, I just haven't seen it) but i've seen a bunch of 2.2l cars do it and usually they had a MIL on and they would be hard to start cold, and the fans would run at high speed at any time the IGN was on or run.

Good luck!

Re: Time to play "Which sensor's broken?"!!!
 

Thanks for all of the input everybody.

Next question...which of these would still allow the engine to run smoothly sometimes. Today I started it and it ran great until it got up to temp...then it stalled. Seems like the exhaust, grounds, injectors, and other hardware issues wouldn't allow it to ever run right. This leaves me thinking sensor (maf, crank, cam) or possibly intermittent wiring issue.

Probalby gonna bite the big one and order the crank and cam sensor :(

Re: Time to play "Which sensor's broken?"!!!
 

Hmm... I would clean the maf, get a multimeter and try to test the sensors out before you start to throw parts at it. Just my .02.

Re: Time to play "Which sensor's broken?"!!!
 

will do tomorrow. Can somebody link me to the MAF test procedure. I know the cam/crank sensors require an oscilliscope to check, so I'll just have to wait for my temporary replacements to arrive from Tom soon.

Are all 4 of these sensors in a default tuning map when the vehicle is warming up?

Seems like it couldn't even idle cold without a crank angle sensor as it'd have no basis for timing. What about cam position sensors...can it idle cold without those? How about MAF? If my memory serves me correctly, it can idle cold without MAF input.

Re: Time to play "Which sensor's broken?"!!!
 

The car wont run at all without a MAF input.... it needs to know how much air the engines ingesting in order to supply the right amount of fuel. A bad MAF would act up cold or hot. Crank and cam sensors can be visually inspected, and also tested to some degree with a multimeter, using a metal object paseed back and forth in front of the magnet and looking for a voltage jump.

You have already thrown a code for the Crank sensor, I dont believe some of the others have read through the entire post or that would be the first recommendation made, to inspect the sensor AND wiring/harness, clean, and replace if the cleaning and inspection didnt help solve the problem.

Re: Time to play "Which sensor's broken?"!!!
 

Just read this and sounds familiar...

Recently I had to do a knock sensor job so I rinsed the warm engine down with some cool water... Well the water worked its way into the crank angle sensor wiring... Car would idle perfectly, but would not rev over 1600RPM. No codes, no lights just loss of ignition... Check that harness for resistance from the angle sensor connectors all the way back to the bulkhead harness connector at the back of the engine. If you find a wire with high reisistance, it is likely the harness dealing with some signal contamination

Tom

Re: Time to play "Which sensor's broken?"!!!
 

great...thanks for the tips. I'll dig in with my digital multimeter tomorrow :) Kinda excited to do a little diagnosis!

Re: Time to play "Which sensor's broken?"!!!
 

checked the cam and crank sensors and connectors. found 2.7k ohms resistance across rhe sensors and .87k ohms across the connectors. i know i can`t really test the sensors by resistance but i did anyway =) swapped the position of the 2 sensors that are the same and wrapped electrical tape around where the wire insulation was broken @ the connector for the sensor on the side of the engine. still stalls when warm. starts again no prob though. runs for a few sec then stalls. ugh

Re: Time to play "Which sensor's broken?"!!!
 

im sorry but you try changing the MAF right??

Tom

Re: Time to play "Which sensor's broken?"!!!
 

yea...but I don't know how to know if either of the 2 I have are any good.

Re: Time to play "Which sensor's broken?"!!!
 

If the symptoms are the same with either one installed you can pretty much eliminate the MAF itself as a cause.

Quick question, you DID replace the coolant temp sensor with a new one yes?

The fact that the car runs fine cold but starts acting up when it gets warm really seems like a timing or mixture problem. One possibility I came up with while racking my brain going over your thread, is maybe the ECU is constantly detecting a cold engine via the CTS and not going into closed loop. The ECU adjusts timing and mixture based on the temperature signal it receives from the CTS, and that may be why it runs great when cold but after the engine warms up it starts giving you problems.

If you have already replaced the CTS, test it with a multimeter (cold and hot), and also check the harness connector. If yours is like mine, the wires bend almost 180 degrees from the wire loom to the CTS connector because of where the wires come out of the loom, making a wire problem very possible.

Another thing could be the Auxilary Air Control valve sticking, I believe this valve provides the extra air to increase your cold idle speed. If this valve does not shut when the engine warms up, it will make the engine run too lean causing idling and stalling issues. The bypass air control valve (aka IAC valve) could also be a problem, if its clogged, once the auxillary air control valve closes when the engine reaches normal temp theres nothing else to supply air to the engine without the throttle being pressed, which of course will stall it.

Hope this helps, if I can think of anything else ill post it up, keep us informed of any more tests or symptoms you have.

Re: Time to play "Which sensor's broken?"!!!
 

Thanks for the additional input. I'll try swapping the MAF again to make sure. I did install a new Coolant Temp Sensor bought from Subaru. What voltage/resistance do I want to measure on that thing?

It definitely seems like a timing/mixture issue as the engine switches into closed loop operation, and I would assume that unplugging the CTS throws it into a default map that runs well. But even this trick didn't work last time (Monday).

I cleaned the IAC, and I think I checked the aux air control valve also. It's been acting up so long I can't even remember what all I've checked. However, it stalls even if I have the throttle open to 2k rpm or so...not just at idle. So I don't think it's that. It's something that controls mixture/timing at all times, not just idle.

Here's to hoping the cam/crank position sensors or MAF will fix it.

I'll keep you posted

Re: Time to play "Which sensor's broken?"!!!
 

you have checked the timing marks to make sure the belt didn't slip right? If they are aligned properly, I would pull the LH gear off and make sure it hasn't backed the bolt out enough to allow the cam and gear alignment to become trashed (have seen this before more than once)...

Tom