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 ExtremePSI
Please Support ExtremePSI

2G Down a cylinder or two, then OK?

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.

daughters-car

10+ Year Contributor
52
20
May 18, 2012
Celestine, Indiana
This new car is killing me! Just fixed a whole list of issues, the last of which was getting the cruise control working and thought it was finally good to go. Then last night while teaching my daughter how to drive a stick shift, it started running terrible. We are working on a 97 Spyder GST with a manual transmission and awd swap.

We drove to town with the car running fine, stopped for food and when we started the car after it was running horrible with the check engine light on. I checked for obvious issues like loose hoses, connectors, plug wires, but didn't see anything obvious. The car was barely drive-able, definitely down 1 or 2 cylinders. I took over driving duty to limp us home. I noticed the tach and speedometer were both reading incorrectly (low), the wideband was pegged lean, and the car would spit black smoke, flames and sparks if you got into the throttle too much! We made it home and I broke out the ODB scan tool and got only one code - P0300 random misfire. Well duh! I could tell that. It was like zero degrees out, so I cleared the code and went in the house to put on some warmer clothes so I could try to diagnose without freezing to death.

After changing clothes and going back to the garage, the car started up and ran perfect! I tried wiggling connections at the CAS, TPS, ignition coils, PTU, injectors, and temperature sensor with no effect. Took the car for multiple test drives and could not duplicate the issue again. Then while letting it sit and idle, the check engine light came back on, but the car was still running fine. This time it had multiple codes, P0300, P04?? (something to do with evap), P0505, and P1105. I decided to give up for the night at that point.

The bad tach signal was leading me toward the PTU, but this car already has a new PTU, coils, plug wires, and spark plugs. I have found many wiring issues in this car while fixing other issues, so another wiring problem wouldn't surprise me either. I did some searching and see several instances where a PTU was only intermittently bad and did cause multiple codes similar to those I am now getting. Would a bad PTU also cause the speedometer to read wrong?

I have a spare PTU that I can swap, but I don't want to try that until the car acts up again. That way if it fixes the issue, I'll know that was the problem. I think I'll carry the spare PTU and tools to change it with me for a while.

Any other suggestions to diagnose this problem that magically fixed itself? My daughter wants to take this car back to college (4 hours away) with her this coming weekend and I don't trust it right now.
 
What were the wire colors you fixed; do you remember?

Yeah, it's weird. My '97 has a brittle, delicate harness whereas my '96 has no issues, even after sitting neglected for a few years. Both cars have seen the same environmental conditions, so one would think they'd be in similar condition. Maybe a previous owner sprayed some sort of chemical on the '97s harness while cleaning the engine at some point, maybe the engine bay got really hot at some point, who knows...



Yup, the Fuel Pressure Solenoid gives the ECU control over fuel pressure. On a hot start the ECU uses it to adjust fuel pressure to prevent a hard start when the first few jets of fuel atomize when injected into a hot cylinder.

The engine harness in my 98 looks and feels brand new. Even the electrical tape and loom is soft and malleable. Never seen an engine harness like it before honestly

You must be logged in to view this image or video.
 
What were the wire colors you fixed; do you remember?
It was the wire to pin 4 of the IAC, I would call it white with a blue stripe. It looks like the wiring diagrams call it gray with a blue stripe. I fixed all the wires on the knock and temp sensors.

The car threw the check engine light today while running errands at lunch - P1105 fuel pressure up solenoid. Still starting, running, and driving just fine. Looks like there must still be something wrong somewhere because I know that sensor is good and it passes all the diagnostic checks in the book. It's not really sensitive enough to know that it's not connected to the vacuum lines right? My guess is another intermittent issue with the wiring somewhere.
 
It was the wire to pin 4 of the IAC, I would call it white with a blue stripe. It looks like the wiring diagrams call it gray with a blue stripe. I fixed all the wires on the knock and temp sensors.

The car threw the check engine light today while running errands at lunch - P1105 fuel pressure up solenoid. Still starting, running, and driving just fine. Looks like there must still be something wrong somewhere because I know that sensor is good and it passes all the diagnostic checks in the book. It's not really sensitive enough to know that it's not connected to the vacuum lines right? My guess is another intermittent issue with the wiring somewhere.


You would need to pull out the service manual for the diagnostic flow for trigger parameters. I'm guessing one of those parameters is knock in the first 1000 revolutions with coolant temperatures over 176 degrees

Hook the vacuum lines back up to the solenoid
 
I'm getting nowhere with this thing, not sure what to try to check next. So the manual I have only lists P1105 judgement criteria as coil surge voltage not detected when solenoid is triggered, so that's not much help. Following the test criteria, solenoid checks OK, solenoid connector (A79) has battery+ voltage when ignition is on - so that checks good as well. Next is where I thought I might be on to something, but instead believe I have found an issue or error with the manual's test criteria. It has you disconnect ECU connector B53 and check for battery+ voltage on pin 3 on the harness side. I get 5 volts, not full battery voltage. I spent hours chasing and researching this - traced wiring through the MFI relay and on and on. My conclusion is that the manual is in error and here is why. With ECU connector B53 disconnected, the ecu does not have ground (pins 13 and 26). Without proper ecu ground, the ecu can't pull the MFI relay on (by grounding MFI relay through ecu pin 38), so the MFI relay does not supply battery+ voltage to the system. The systems that I am trying to check get battery+ voltage through the MFI relay! So, the test procedure in the manual makes no sense to me. I traced wiring diagrams and see that the +5 volts that I am getting is coming from the other systems connected to the MFI relay (MAF, cam, and crank sensors). So after spending hours trying to track this down, I'm pretty sure it was a dead end waste of time because the procedure in the manual isn't accurate. With ECU connector B53 plugged in to the ecu the MFI relay works properly and supplies bat+ voltage to those systems as it should.

I'm going to connect the fuel pressure solenoid back into the vacuum lines as recommended and see if the p1105 still comes back or not. I think it may have had nothing to do with the original problem this thread is about that I am trying to diagnose. Since the p1105 code is the only issue that has re-occurred so far, I thought trying to diagnose it might lead me to discover whatever might have caused the original issue.
 
I'm getting nowhere with this thing, not sure what to try to check next. So the manual I have only lists P1105 judgement criteria as coil surge voltage not detected when solenoid is triggered, so that's not much help. Following the test criteria, solenoid checks OK, solenoid connector (A79) has battery+ voltage when ignition is on - so that checks good as well. Next is where I thought I might be on to something, but instead believe I have found an issue or error with the manual's test criteria. It has you disconnect ECU connector B53 and check for battery+ voltage on pin 3 on the harness side. I get 5 volts, not full battery voltage. I spent hours chasing and researching this - traced wiring through the MFI relay and on and on. My conclusion is that the manual is in error and here is why. With ECU connector B53 disconnected, the ecu does not have ground (pins 13 and 26). Without proper ecu ground, the ecu can't pull the MFI relay on (by grounding MFI relay through ecu pin 38), so the MFI relay does not supply battery+ voltage to the system. The systems that I am trying to check get battery+ voltage through the MFI relay! So, the test procedure in the manual makes no sense to me. I traced wiring diagrams and see that the +5 volts that I am getting is coming from the other systems connected to the MFI relay (MAF, cam, and crank sensors). So after spending hours trying to track this down, I'm pretty sure it was a dead end waste of time because the procedure in the manual isn't accurate. With ECU connector B53 plugged in to the ecu the MFI relay works properly and supplies bat+ voltage to those systems as it should.

I'm going to connect the fuel pressure solenoid back into the vacuum lines as recommended and see if the p1105 still comes back or not. I think it may have had nothing to do with the original problem this thread is about that I am trying to diagnose. Since the p1105 code is the only issue that has re-occurred so far, I thought trying to diagnose it might lead me to discover whatever might have caused the original issue.

The red power feed wire that is connected on pin 1 of the fuel pressure solenoid SHOULD have battery voltage, as that feed is directly from the MPI relay; that circuit feeds other components on the engine and elsewhere with 12v. The ECU is connected to the ground side of the FPS, and triggers it by grounding pin 2 on the FPS via a transistor on the ECU connected to pin 3 of connector B-53.

The other 5v you're seeing on the MFI circuit diagrams is the 5v reference voltage output by an internal voltage regulator in the ECU. It provides a stable, known voltage for sensors, like the TPS, CPS, CAS, etc. to measure against. The ECU triggers the MPI relay, and the MPI relay feeds 12v to the ECU which it steps down to 5v for the reference circuit with its voltage regulator.

If you only have 5v on pin 1 of the fuel pressure solenoid, then you've found your issue. The car shouldn't run if the MPI circuit is at 5v, but what you can do is check pin 3, the red wire, of the camshaft position sensor. You should have battery voltage here as well, and it's spliced off the same circuit as the FPS. If you do have battery voltage, then you probably have an issue at the FPS connector, or something wrong on the FPS leg of the MPI circuit (partial open). This could cause all sorts of issues with sensors, ignition, fuel delivery, etc. If you don't have battery voltage at the cam sensor, then you have a power feed problem somewhere upstream.

Edit:
I misread your post. If you have 12v on pin 1 at the FPS, and 5v at pin 3 on the harness side of B-53 with B-53 disconnected, then you have an issue with the ground leg of the circuit between the FPS and the ECU, or the FPS connector itself. This is assuming that the FPS itself passed the diagnostic procedure referenced at the top of the flow chart.

I don't have my '97 Vol I in front of me, but the check in the '96 manual seems correct. It first has you check pin 1 on the harness side with the FPS connector disconnected to confirm that the MPI relay is sending 12v with the key on. If you have battery voltage here you're confirming that the MPI relay and the wiring between it and the FPS are good. Next it has you check pin 3 on the ECU harness for voltage with the harness disconnected. Battery voltage here confirms a good ground leg between pin 2 of the FPS and pin 3 on the ECU side of the harness, where the transistor which grounds the FPS to actuate it connects. With the harness connected here you'd have 0 volts (the ECU is grounding the circuit), but with it disconnected you should have battery voltage from the MPI relay--obviously, the FPS needs to be good and the FPS connector needs to be connected for this test.

The fp solenoid is only checked by resistance, there is no functionality test done on it by the ecu.

Somewhat right. According to the service manual, when the ECU actuates the FPS, it monitors voltage on that circuit to confirm that the FPS is functioning. It looks for a surge in voltage when the FPS is triggered, and if it doesn't see that surge it sets CEL p1105. A change in resistance would change voltage reading, you're correct, but it's not resistance the ECU is specifically looking at.
 
Last edited:
I don't think I explained the situation and tests very well in my post above. Basically, I think I found out why the test procedure in some versions of the manual are different than others. The test procedure in the version that I have does not work. If connector B-53 is disconnected at the ecu, the MFI relay does not function - that is what I found anyway. I think this is probably normal, since the ecu gets it's ground through pins 13 and 26 on connector B-53. If the ecu doesn't have ground because b-53 is disconnected, then it can't trigger the MFI relay because it does so by grounding the relay.

So, I've noticed through pictures some others have posted of test procedures in other versions of the manual, sometimes they have you supply ground to the MFI relay to trigger it on for testing procedures. I have also seen, like earlier in this thread, where the manual has you check current flow instead of voltage. I think I just figured out why.

Anyway, I think my car is working correctly because if the MFI is triggered on, I do have battery+ voltage on that circuit. If the MFI is not triggered on, then I have +5 volts, which from looking through the wiring diagrams is "back fed" from the other sensors I mention on that circuit.
 
Ah, yeah, I get what you're saying now. There is a ground on pin 13 & 26 of the B-53 connector, and they both ground to #3 by the cruse control box on the firewall. Ground 3 is where the knock sensor, CPS, PTU and VSS ground, and I hadn't noticed the ECU grounds there as well...nice find. The ECU has multiple grounds, and I'm not 100% sure which is required to pull relays to ground.

Simple way to confirm would be to disconnect B-53, turn the key on, and see if you have voltage coming directly out of the MPI relay or elsewhere on the MPI circuit (which it sounds like you tried).

So, if the book procedure is wrong, a better way to test the ground leg of the FPS circuit would be to take the ECU out of the equation by either jumping the load side of the MPI relay, or by providing it a ground before doing the pin 3 test. One could also simply just jump the ground side of the disconnected FPS connector to battery positive and confirm 12v on pin 3 of disconnected B-53.

Edit: wrong see below. All that said, I still don't think your 5vs are backfed from the reference circuit. From my understanding, the MPI relay feeds the ECU's regulator for the reference circuit. I could be wrong, but this is my take on power-flow: with the ignition on, the ECU "wakes up" by seeing power from the ignition switch. When the ECU is on, it grounds the MPI relay, and power from the relay is then fed back into the ECU on pins 12&25 of connector B-53. It's from these two pins that the ECU derives the power it needs to switch components high, and also for the internal 5v regulator for the reference circuit. Even if you're right and the ECU needs the grounds on B-53 to energize the MPI relay, the 5vs you're seeing cannot come from the reference circuit, because the reference circuit is not being fed power when B-53 isn't connected.

If you're seeing 5vs, something else might be feeding into the MPI circuit (or I'm wrong and the MPI doesn't provide power for 5v ref. I've never seen it documented anywhere for DSMs, but it's similar for other makes).
 
Last edited:
Since I'm working on my '97 today, to satisfy my curiosity, I ran some tests to confirm your results. My good Fuel Pressure Solenoid is on the '96 currently, so this car has a known bad FPS on it, and is throwing a P1105.

With the key on, I measured battery voltage at pin 1 of the FPS, the power feed from the MFI relay. This checks out good.
With FPS connector connected, and B-53 at the ECU disconnected, I measured pin 3 at the harness side of B-53 and found 6.97v. Interesting.

Next, to see what disconnected B-53 did to the 5v reference circuit, and get another reading on what it does to the MFI relay circuit, I moved to the Mass Airflow sensor.
As a control, key on, MAF connector disconnected, B-53 connected:
Pin 3 (Blue/yellow), the 5v reference from the ECU for the MAF, I measured 5.0v
Pin 6 (Red/Blue), the 5v reference for the IAT sensor, I measured 4.72v.
Pin 4 (Red), the battery+ feed from the MFI relay, I measured Battery Voltage.

Key on, MAF disconnected, B-53 disconnected:
Pin 3 (Blue/yellow), the 5v reference for the MAF, I measured 6.8v.
Pin 6 (Red/Blue), the 5v reference for the IAT sensor, I measured 6.72v.
Pin 4 (Red), the battery+ feed from the MFI relay, I measured 6.54

Next I disconnected everything from the ECU.
Key on at the FPS connector:
Pin 1 (red): 1.2v. Still voltage on the MFI circuit...

At B-53 with key on and the FPS connector connected:
Pin 3: 1.2v.

With key on at the MAF sensor:
Pin 3 (Blue/yellow) I measured 0v. Open.
Pin 6 (Red/Blue) I measured 0v. Open.
Pin 4 (Red) I measured .20v.

For one last test, with B-53 disconnected, and the FPS connector disconnected, I jumped battery positive to pin 2 on the FPS harness. At pin 3 of B-53 I measured battery voltage.

So, what can I gleam from this:

First and foremost is that you are correct, the diagnostic procedure for p1105 in the manual is wrong. With B-53 disconnected, the MFI relay is not triggered, and therefore it's impossible to get 12v on pin 3 of a disconnected B-53. A quick search didn't result in anyone mentioning an incorrect procedure for the FPS tests, so looks like you've brought us new information some 20 years on.

Second, I confirmed that same same voltage present on the 5v reference circuit is present on the MFI circuit with B-53 disconnected. As you suggested, the 5v reference might be backfeeding through the the MFI circuit when the ECU is ungrounded.

Thirdly, I was talking out of my ass in the above post. Unsurprisingly, Mitsubishi is doing things a little differently than I've seen on other makes. Though the MFI circuit feeds back into the ECU, it does not power the 5v reference regulator. This power likely comes from the ignition feed, and should be present independent of the MFI relay state. I could have 100% confirmed this by disconnecting the MFI relay, and measuring for 5v reference at the MAF, but I didn't feel like tearing apart my center console today.

Lastly, for posterity and for future searchers, a better diagnostic procedure for a P1105 would be:
  1. Check the fuel pressure solenoid itself via the method described in the service manual.
  2. With key on and the FPS connector (A-79) disconnected, measure pin 1 (red). You should see battery voltage here. If you do that means the MFI relay, ECU's control of the MFI relay, and all the wiring upstream of the FPS are good. If you don't it means you have a bad FPS connector, bad MPI relay (but that'd cause a whole host of other issues), bad wiring between the MFI relay and FPS, or bad ECU (ECU not triggering the MFI relay, but again there'd be other issues).
  3. With connector B-53 at the ECU disconnected for safety, and the FPS connector disconnected, jump a wire from battery positive to pin 2 (blue/red) on the FPS connector. Measure pin 3 (blue/red) at connector B-53 and you should see battery voltage. If you do then you know the ground connection between the FPS connector and ECU connector B-53 is good; if the previous steps were good, it implies the ECU is the fault (not triggering the FPS). If you don't see battery voltage then there is a problem on the ground leg (blue/red wire) of the FPS circuit, either at the FPS connector, the B-53 ecu connector, or the wiring between the two.
 
Last edited:
motomattx said:
The fp solenoid is only checked by resistance, there is no functionality test done on it by the ecu.
Click to expand...
Somewhat right. According to the service manual, when the ECU actuates the FPS, it monitors voltage on that circuit to confirm that the FPS is functioning. It looks for a surge in voltage when the FPS is triggered, and if it doesn't see that surge it sets CEL p1105. A change in resistance would change voltage reading, you're correct, but it's not resistance the ECU is specifically looking at.

The same thing a resistor will do in the circuit. Hundreds of us have replaced that solenoid with a resistor and have never had a cel since.
 
The same thing a resistor will do in the circuit. Hundreds of us have replaced that solenoid with a resistor and have never had a cel since.

Yup, that's why I said you were somewhat right; still: the ecu is looking at voltage, not resistance. Adding the appropriate resistor to the circuit brings the voltage into the range the ECU expects so it doesn't set a CEL. So you were right in effect, and I was just clarifying.
 
The manual is actually explaining how the driver for the circuit becomes saturated, its more than any tech would ever need in reality, having a 2yr degree in industrial electronics I can of course fully understand what they are explaining, its just a little much for the average person, the translation from japanese to english is why some of these things are in the book the way they are.
 
The manual is actually explaining how the driver for the circuit becomes saturated, its more than any tech would ever need in reality, having a 2yr degree in industrial electronics I can of course fully understand what they are explaining, its just a little much for the average person, the translation from japanese to english is why some of these things are in the book the way they are.

I hadn't thought about the translation, good point. What do you think their original intention/method for the p1105 pin 3 test was before the botched translation? Guess I need to track down an Evo3 service manual, LOL.
 
Oh, I think that what they printed is technically whats happening in the ecu LOL, its just not the kind of info that a tech here in the states needs to know, I see alot of Japanese manufacturers get tech crazy like that in the service manuals, we just want to know how to diag the problem, we dont need a rundown on the hex code in the ecu's rom and what line of code it uses to set the cel for a particular item.
 
Thank you shinzon for taking the time to do that testing and verifying the MFI is not triggered with B-53 disconnected! That certainly helps me to know what I observed on this car is not abnormal.

Something that had not yet occurred to me, all of my testing on the FPS has been cold. It could be possible for it to test good cold, but go open when hot. I need to do some more testing - I wonder how low the temp on the wife's oven goes...
 
Sure as heck, the FPS goes open (infinite resistance) when heated up. I put it in the oven set at 170 and used an infrared thermometer to check the temp every few minutes while also checking the resistance. It goes open when it gets to about 120 degrees. It reads 39.5 ohms at room temp. I've got two other spare solenoids I'll test and make sure they're still good when hot and swap one if those onto the car. I'm sure that will solve the P1105 code and I can get back to focusing on why it lost two cylinders.
 
Sure as heck, the FPS goes open (infinite resistance) when heated up. I put it in the oven set at 170 and used an infrared thermometer to check the temp every few minutes while also checking the resistance. It goes open when it gets to about 120 degrees. It reads 39.5 ohms at room temp. I've got two other spare solenoids I'll test and make sure they're still good when hot and swap one if those onto the car. I'm sure that will solve the P1105 code and I can get back to focusing on why it lost two cylinders.

The fuel boils in the rail and dumps the pressure. The fpr solenoid prevents this

I explained how the fpr solenoid causes the car to run extremely poorly like it's only 2 cylinders on hot starts on page 1.

hook up the solenoid
 
Hey chrysler kid if the fpr sensor is bad would it cause a rough idle for 30 seconds or so on a warm start an hour later? My car seems to stumble for a few seconds then if i hold the gas steady to keep my idle then its fine after that.
Not trying to jack the thread just curious if thats what i have goin on.
 
Hey chrysler kid if the fpr sensor is bad would it cause a rough idle for 30 seconds or so on a warm start an hour later? My car seems to stumble for a few seconds then if i hold the gas steady to keep my idle then its fine after that.
Not trying to jack the thread just curious if thats what i have goin on.

Sounds like the classic DSM hot start issue. I run into it too if I get the engine good and hot, and then let it sit for like 45-60 minutes, then try and start it again. It'll run rough like it's missing a cylinder or two with the smell of unburned gas, and it clears up after a few moments when you apply throttle. While the FPS was supposedly designed to alleviate this condition, even though I have a functioning FPS, I still run into this issue under certain circumstances.

Interestingly, I very rarely run into this issue on the '97, but it's consistent on the '96. I've seen it suggested it's actually a bug in Mitsubishi's code for the ECU. Maybe they made revisions to the code that lessened it.
 
I've put a good 500-600 miles on this car now without anymore trouble. Planning to take it to my daughter @ college this weekend when we visit her for her birthday. Still no idea what made it run so bad that one night. Hopefully it was just one of the many wiring issues or ground points that I fixed while trying to pinpoint the issue and it will not act up again!
 
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