The Top DSM Community on the Web

For 1990-1999 Mitsubishi Eclipse, Eagle Talon, Plymouth Laser, and Galant VR-4 Owners. Log in to remove most ads.

Please Support JNZ Tuning
Please Support STM Tuned

2G Help Revive My Talon - Doctor, Doctor, Give Me the News

This site may earn a commission from merchant
affiliate links, including eBay, Amazon, and others.

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

shinzon

15+ Year Contributor
350
72
Sep 11, 2006
St. Petersburg, Florida
Some background: During the summer of 2015, after nearly a decade of daily driving, my '96 TSIAWD (stock with the exception of a 14b swap) developed a nasty exhaust leak so I parked it. It ran fine otherwise. Sometime in early 2016, I attempted to start the car, but realized the battery harness was severely corroded. After purchasing a new harness from Mitsu (one of the last in the country at the time), I took the car apart, life got in the way, and I never put it back together.

So about a month ago I started working on the car again. I drained the fuel, replaced the fuel filter and all other liquids and filters, new bpr6es plugs (.28) and NGK wires, finished the harness swap and she fired up.

Symptoms: The car now starts, idles, and moves under it's own power, but it gets real unhappy when you start trying to make boost: bucks, and it backfires when you lift. If you stay below 5psi or so, the car is drive-able, but I've noticed that as soon as the engine warms up, and the ECU enters closed loop, it starts to stutter, run/idle very rough, and struggle not to stall.

The car still has the exhaust leak that made me park it. Might be a bad turbo to manifold gasket, cracked turbine housing on the 14b, or cracked manifold, or bad manifold gasket. Seems to be coming from that area.

It's throwing CELs for Fuel Pressure Solenoid, Wastegate Solenoid, and Crank Position sensor

What I've checked: I suspected Coolant Temp Sensor so I hooked up EVOScan and logged it, but it seems to be reporting temperature correctly. MAF sensor, intake air temp sensor, and barometric sensor seem to be giving sane vales. CPS reads zero while the car is running (this is what it's supposed to do right?), but I see it change value when starting. I only have a many-year-old NGK narrowband O2 sensor, so I can't ascertain much about a/f. I did notice a couple counts of knock after crossing the boost threshold at half throttle.

I've also done a compression test. Warm with WOT, results were: 4)192, 3)195, 2)195, 1)192. All four cylinders were high (maybe carbon deposits in the cylinders of a factory built 150k mile engine?), but they were almost bang-on in relation to each other.

Where should I be focusing my attention?
 
Last edited:
Could be a bad o2 sensor or an exhaust leak so bad that it's screwing with the o2 sensor reading. This is besides all the cels it's throwing at you.

Yeah, I was thinking a pre-trubo exhaust leak would suck fresh air through the turbine and give a false lean condition reading to the ECU. Running rich would correlate with the black build-up around the bases of the new spark plugs when I took them out for the compression test. I was hoping to get the car running well enough to get a decent drive-cycle on it so I could see what the fuel trims were doing.

As for the CELs, the solenoids should be irrelevant, but the CPS is a bit of a mystery for me. I was under the impression that a bad/dying CPS would cause an intermittent no-start...

Thanks for the reply.
 
New information: I'm embarrassed to say that after 12 years of DSM ownership, I didn't think to do a boost leak test first thing. System was leaking badly from a torn o-ring on the MAP sensor and I could barely pressurize it to 10psi. I'll update with the remaining symptoms (if any) once I get this o-ring, and a few leaking vacuum lines sorted out.
 
Well, an update: After fixing all of my boost leaks, the engine ran worse: made a piff, piff, piff noise when running and didn't have enough power to move.

Seemed like the timing was off. Since I was throwing a crank sensor code, and the ECU uses the crank sensor to set timing advance, I knew what I had to do next. I set about the unholy task of tearing down the front side of the engine. My hands swollen and bruised, my knuckles bleeding I finally got the lower timing cover off and inspected the CPS.

Well, I discovered why it wasn't sending good pulses to the ECU: there as a metal shaving stuck to the magnet on the sensor. Where the shaving came from I have no idea. I saw nothing loose under the timing cover, nothing marred. I can only assume I introduced that shaving when I changed the CPS and timing belt components about 10k/four years ago.

More worryingly: the back of the sensor looks like this:
https://imgur.com/a/jdCeC

Looks like something may have rubbed into, or melted it, but I have no idea what could have done that to the backside...Also I have no idea what the RTV like goop is. Melted plastic?

I have heard of crankwalk taking out the CPS, but I would assume a walking crank would make the ring impact with the armature of the sensor, not do damage to the backside...

I'm stumped here. Any help would be greatly appreciated.


Edit:

Solved it: I am monumentally stupid. After looking at a diagram in the service manual, four years ago I installed the CPS upside down, so that instead of the disc spinning between the armature, the metal plate was making contact with the back of the sensor. I'm amazed the disc spinning against the back of the sensor was enough to provide pulse to the ECU, and that the motor ran for perfectly for 10k miles before the sensor gave up.

Feel free to make fun of me. I'm flabbergasted by my own stupidity.
 
Last edited:
I assume you will replace the plate also. Look at its oriemtation carefully when installing. A backwards plate is disastrous. I also wasn't paying attention one time when bolting things down. Destroyed my plate and sensor quick like.
 
I assume you will replace the plate also. Look at its oriemtation carefully when installing. A backwards plate is disastrous. I also wasn't paying attention one time when bolting things down. Destroyed my plate and sensor quick like.

The plate itself looks surprisingly unscathed, no obvious marring or anything like that. In any case, I have an extra one laying around that I'm throwing on and I'll be sure to triple-check the orientation.
 
She still ain't right.

While I was in there, I took a couple of days to replace all of the seals on the front side of the engine. After 150k and almost 22 years, they were well passed it. As an aside, someone turned me on to Lisle 58439, a blind (shaft-type) seal puller. That tool made for the quickest, most hassle-free seal replacement job I have ever done. The front main, the front balance shaft, the oil pump, and the cams--all came out without a fight. Highly recommended to anyone doing this job.

As I said further up-thread, before I tore it apart, it was throwing a crank sensor code. I changed out the CPS, and I got it back together enough that it will start while supported with a jack, and with the front case off--it's still misfiring, but it's better than it was before I tore into it and replaced the CPS. Before it seemed like the timing was off, now it seems like it's just missing one or more cylinders.

It instantly turns the check engine light on after resetting the ECU and gives me a p0300, a multiple cylinder/random misfire. After letting the engine run for a few minutes in its hobbled state, it popped up a P0335, crank position sensor, as well. I checked for spark and cylinders 1 and 4 are not getting spark; 2 and 3 are.

I quadruple checked: all timing marks are still lined up, the crank signal ring is installed in the correct orientation, and I installed the CPS the right way around this time.

This persistent crank sensor code perplexes me. I guess the next step is to swap with a known good coil from my other Talon to make sure I'm not dealing with two separate problems. Beyond that, I'm at a loss.

What should I be looking at from here?

On the plus side, it doesn't seem to be leaking oil passed my new seals at least...
 
Last edited:
Make sure your plug wires are in the correct order. Mechanically speaking these cars are pretty simple, the computer just determines injector pulse timing.

They're brand new NGK wires, but I'm going to break out my multi-meter after work tomorrow and check their resistance, and test the coil.

The ignition system is where my knowledge is falling short, in particular what it uses to determine when to fire the coils. The service manual states:

...The engine control module determines which ignition coil should be controlled by means of the signals from the camshaft position sensor, which is incorporated in the camshaft, and from the crankshaft position sensor, which is incorporated into the crankshaft. It also detects the crankshaft position in order to provide ignition at the most appropriate timing in response to the engine operating conditions. -- Vol. II - 8A-69

So, according to this, it uses both sensors to determine when to fire the 1/4 coil or the 2/4 coil. What I can't seem to find any information on is HOW the ECU uses this signal to decide which to fire, and if either one sensor or the other being bad would cause only one coil to fire.

In any case, I'm going to back probe the CPS connector to attempt to follow the diagnostic procedure for P0335 outlined in the manual, so I'll have more information tomorrow.
 
They're brand new NGK wires, but I'm going to break out my multi-meter after work tomorrow and check their resistance, and test the coil.

The ignition system is where my knowledge is falling short, in particular what it uses to determine when to fire the coils. The service manual states:



So, according to this, it uses both sensors to determine when to fire the 1/4 coil or the 2/4 coil. What I can't seem to find any information on is HOW the ECU uses this signal to decide which to fire, and if either one sensor or the other being bad would cause only one coil to fire.

In any case, I'm going to back probe the CPS connector to attempt to follow the diagnostic procedure for P0335 outlined in the manual, so I'll have more information tomorrow.

Cylinders 4 and 1 are on the first coil on the bracket. It could be that the coil has gone bad. If you have a spare coil now would be the time to try it

Are you sure you put the crank trigger plate on correctly and not backwards.
 
The cam sensor determines when to fire the coils, you could have a bad ignition transistor.

I hadn't considered the transistor. After doing a quick search on here, it seems many people with a bad transistor reported both 1&4 with no spark. Thanks for the info, and I'll add it to the list I'll check this afternoon.

Cylinders 4 and 1 are on the first coil on the bracket. It could be that the coil has gone bad. If you have a spare coil now would be the time to try it

Are you sure you put the crank trigger plate on correctly and not backwards.

I have a running '97 AWD I can swap parts from so I'm good on spares to try and diagnose this, LOL.

I am 100% sure. I marked the orientation before removing the original, and the trigger plate I replaced it with was old enough that it had discoloration/oxidation around where the crankshaft balance shaft belt sprocket was on the back, and the smaller main timing belt sprocket was on the front. Really easy to match up.

I only looked at it for a few minutes, but is it actually possible to put the plate on backwards and still have the timing marks line up? Seemed like with the way it's keyed, flipping it over would move the timing notch from the 10 o'clock position at TDC to 7 o'clock... Wouldn't aligning the dot on the crankshaft sprocket and the notch in the plate always ensure you put it on right?

Thanks for the responses. I'll post back later this evening after I run some tests.

Edit: Oh, and a word of caution. I had originally bought a Dorman 917-024 aftermarket trigger plate before I found my spare OEM in my parts cabinet. If anyone goes down that road, inspect the trigger plate thoroughly. The one I received came slightly bent/bowed out of the box. Had I put that on the car, it'd have murdered my new CPS.
 
Last edited:
The cam sensor determines when to fire the coils, you could have a bad ignition transistor.

*DING DING DING* We have a winner.

Tested all four wires and they were well below 22kohm, so they're good. Tested the coil, and the secondary coils were high enough to be out of spec, but the primary's were bang on in range.

Just to be sure, I swapped over the coil from the '97 and I got the same result: no spark on 1&4. I then swapped over the power transistor and she fired right up with all cylinders. I confirmed that I have spark on all four.

P0335 has yet to return, but according to the manual, it'll take a few starts where the problem is present before it'll illuminate the CEL. I'm going to spend the rest of the afternoon starting the car, and letting it sit and cool to see if I can encourage it to throw the crank sensor code again. If not, I should be good to put the front side of the engine back together tomorrow.

Thanks for all your help guys. I'll update the thread with any issues that remain once I get it put back together.
 
Last edited:
Glad I didn't rush to put it back together. It runs fine now, starts right up every time you turn the key, but after a few cold starts p0335 is back again. Looks like I'll be back probing the CPS connector tomorrow to confirm what I suspect.

I also noticed that the passenger fan is coming on when you first start the engine while it's cold. Not only is it coming on when it shouldn't, but it looks very weak: turning slowly, maybe half or a third of the usual RPM. When the engine heats up it ramps up to full blast. Similarly, if I jump it from the battery it comes on fully.

If I'm not mistaken, a fan that behaves like this is a sign of a dying ECU. A phantom p0335 code with no symptoms (i.e. the car starts) is another if I'm not mistaken.

When I was plugging in my code scanner and turning the key, I had my head down in the driver's footwell and I heard clickclickclickclick coming from inside the center console when i turned the key to ON. If I'm not mistaken, a dying ECU will send weird signals to the fuel pump relay causing it to click repeatedly when you turn the key.

Also of note now that I'm thinking in this direction: I thought the IAC had seized because when I first gave the car power again after sitting so long the IAC was rattling and jittering. I unplugged it and was going to return to replace that once I got the car running.

So I suspect my ECU is dying. Tomorrow I'm going to pull it out and look for bad caps. I've had an unsocketed EPROM ecu I scored in a junkyard laying around for years so I'll be able to swap it out if need be.

Does anyone have any information they can share on dying ECUs? Any of this sound familiar?
 
Last edited:
The 2g didnt have nearly the cap problems that the 1g did, have you checked out your output voltage from your alternator? how about the output from the coolant temp sensor? is your ac or defrosters turned on? that will bring the coolant fan on.
 
Ive been chasing my iac and ecu for a week now before getting the balance of the tps/bss/tss correct, and every issue you have described is common with ecu failure

Iac driver failure is a sign of cap failure

Clicking mfi relay is also a sign the driver from the ecu isnt functioning correctly

Pull the ecu out and inspect the board. It should be very evident where corrosion and rust spots have fouled the board

You could also just have a bad mfi relay, the iac is controlled by that relay as well

Can you hook into the obd2 port to view live data to determine what the ecu is seeing, coolant temps, voltage, etc
 
I'll give you guys a proper reply later, but I slipped out this morning before work and I have some new information:

I performed the test procedure for p0335 outlined in the service manual on page 14A-160

Measured at the harness side:
#3 (red) and ground with key in ON position: Batt Voltage. Checks out good.
#2 (blue) and ground with key in ON position: 4.97v. Manual says between 4.8-5.2v so this checks out.
#1 (black) and ground with key in ON position: Continuity. Checks out.

Back-probed wire #2 (blue) with the harness connected
Measured .001v with key in ON position. (no spec given for this).
Measured ~1.80~2.50 with engine cranking. Manual says this should be between 0.4-4.0v so that checks out.
Measured ~1.90~1.91v with the engine idling. Manual says this should be between 1.5 and 2.5v so this checks out.

So what I can gleam from this is that the power (MFI relay), ground and signal wires on the harness side are good, and that the new CPS is outputting within spec. The manual states that if the CPS passes the cranking and idling tests, then the ECU is the fault.

Looks like I'll be tearing out the ECU after work.

Thanks again for all the help guys. I'll report back later.
 
Last edited:
Confirmed, bad ECU.

Yanked out the ECU, cracked it open, and it looks like I have three bad caps:
You must be logged in to view this image or video.


Probably caused by sitting outside (though covered up) with a sealed cabin in the Florida sun for two-and-a-half years.

Popped in my EPROM ECU, the car started right up, and has yet to throw a crank sensor code. I'm going to try and coax it to come back on like I did yesterday. Hopefully, if all goes well, I should have this car back together and back on the road this weekend.

I'll update the thread with any further issues.
 
Well she's back at it again.

I drove it to work for a week to see how it'd fair; got through most of the week fine, but started to act up on Thursday.

It started running really rough, threatening to stall as I sat at a light. I tried to get into boost and it felt like I was hitting fuel cut.

I had planned to do a boost leak test this morning, because I thought I heard what sounded like something in the intake let go when I shifted the last time i drove it, but as soon as I started the car this morning it popped up a CEL: P0100. I also noticed that the factory boost gauge on the dash was doing nothing when I revved the engine.

I followed Inspection Procedure 43 on page 14A-204 to check the MAF circuit at the harness side:
- 4 (red) & ground with the Ignition ON: Battery Voltage - checks out good.
- 3 (blue w/ white) with Ignition ON: 4.97v. Should be between 4.8-5.2 so checks out.
- 5 (black) & ground with Ignition ON: Continuity. Checks out.

I followed the diagnostic procedure for P0100 on page 14A-147:
- Backprobed 3 (blue w/ white) & ground with the engine idling: 3.02v. Should be between 2.2-3.2v so checks out.
- Backprobed 7 (red w/ white) & ground with engine idling: .96v. Should be between 0-1v so checks out.
- Backprobed 7 (red w/ white) & ground with engine at 2krpm: 0.96v. Should be 6-9v. So this is no good.

The manual says that if it fails the 2krpm test then to check connector B-53 (the big one) at the ECU:
- Backprobed 19 & ground and got 7.47v. Should be between 6-9v so this is good.

At this point the manual says to check the connector at the ECM and the MAF for damage. They both seem good so I'm looking at another dead ECU.

I did notice that when I did the no-good 2krpm test, if I raised RPM to ~3500 (the boost threshold) it jumped to about 7v. I suspect the MAF actually passed this test and the discrepancy between my results and the book is that my engine has a 14b on it, not the quicker spooling T-25. If the MAF passed this test, the ECU is the fault per the book.

So to confirm my suspicion, I pulled the MAF from my '97 and got the no change on the symptom. Looks like my EPROM ECU is bad too.

I'm going to pull it out and open it up and see if providing power after sitting dormant for 8 years caused it to leak capacitor fluid, or otherwise break in an obvious way.

Beyond that, what could kill an ecu repeatedly?

Edit: Things I've recently confirmed good through diagnostic procedures from the manual: CTS, CAS, CPS, IAC, and now the MAF. Also, though irrelevant to this problem, the temp sender for the dash and pressure sender for the oil gauge.
 
Last edited:
A Tale of Two ECUs is about to become a tale of three. It's nowhere near as bad as my original, but the capacitors on this ECU are bulging too:

You must be logged in to view this image or video.

You must be logged in to view this image or video.


I think I may just contact ECMTuning and have them repair and socket this ECU instead of chasing down another.

Bad grounds, check the ones under the kick panel on the passenger side.

Wasn't aware there were any important grounds down there. I'll check it out and clean them off.
 
Last edited:
Support Vendors who Support the DSM Community
Boosted Fabrication ECM Tuning ExtremePSI Fuel Injector Clinic Innovation Products Jacks Transmissions JNZ Tuning Kiggly Racing Morrison Fabrications MyMitsubishiStore.com RixRacing RockAuto RTM Racing STM Tuned

Latest posts

Build Thread Updates

Vendor Updates

Latest Classifieds

Back
Top