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 ExtremePSI

MPI Relay, No Power to ECM side, can i bypass inner diode?

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.

BeaverItout

Probationary Member
7
1
Jul 31, 2018
Stevens Point, Wisconsin
Resurrecting a 92 1G that has been a relic of the yard for the past decade. Previous owner almost killed me when i figured out the fuel pump was shorting out in the tank with a hot wire with no inline fuse coming off the fuse panel to the pump to ramp up pump voltage. tank was shot, pump was shot. took a while to find a new tank and installed a 255 walboro once i did this last week. was not getting spark. did a bunch of wire chasing and figured out i was not getting supply power to my ecu. Turns out only the fuel side of the mpi relay was switching on for fuel, but the other half was not switching on to turn on the ecu. I am fairly confident with wiring but still hate doing it and there is much to be learned, but i traced the problem to what looks like a diode inline with the ecu side of the relay. when i jump 12v past the diode the little electromagnet inside engages the contacts, but when i apply voltage to pins 10 & 8 on the relay i get nothing. obviously the diode is bad, but my question is how important is that diode in place? cant find much on it but what i am guessing is its in place to eliminate back voltage surge to the ecu incase the motor dies?? wondering if i can bypass the diode via a jumper or will it cause more harm than its worth down the road. Once i finally figured out i had no voltage there i closed the contact points so i could get it started long enough to move the vehicle to a spot where i can legitimately work on it. It runs for a bit and dies, and throwing code 44, but ill save that for another discussion, thanks!

Suspect diode
 

Attachments

  • IMG_20180731_201837680.jpg
    IMG_20180731_201837680.jpg
    261.5 KB · Views: 140
Last edited by a moderator:
Resurrecting a 92 1G that has been a relic of the yard for the past decade. Previous owner almost killed me when i figured out the fuel pump was shorting out in the tank with a hot wire with no inline fuse coming off the fuse panel to the pump to ramp up pump voltage. tank was shot, pump was shot. took a while to find a new tank and installed a 255 walboro once i did this last week. was not getting spark. did a bunch of wire chasing and figured out i was not getting supply power to my ecu. Turns out only the fuel side of the mpi relay was switching on for fuel, but the other half was not switching on to turn on the ecu. I am fairly confident with wiring but still hate doing it and there is much to be learned, but i traced the problem to what looks like a diode inline with the ecu side of the relay. when i jump 12v past the diode the little electromagnet inside engages the contacts, but when i apply voltage to pins 10 & 8 on the relay i get nothing. obviously the diode is bad, but my question is how important is that diode in place? cant find much on it but what i am guessing is its in place to eliminate back voltage surge to the ecu incase the motor dies?? wondering if i can bypass the diode via a jumper or will it cause more harm than its worth down the road. Once i finally figured out i had no voltage there i closed the contact points so i could get it started long enough to move the vehicle to a spot where i can legitimately work on it. It runs for a bit and dies, and throwing code 44, but ill save that for another discussion, thanks!

Suspect diode

Let me pile on with a similar problem I am wrestling with right now. I have replaced a 4g37 with a '92 N/T 6bolt 4g63. The 4g63 tested and was running great before I yanked it out and dropped it into my '90 Laser. I also rewired the '90 Laser to be compatible with the 4g63 ECU.
Once I hooked everything up, I went to start the engine and it turned over but would not start. I energized the A55 connector and the fuel pump whirled happily. My next step was to check the EFI relay. I energized terminal #10(+) and #8(-) and got no power between terminal #4 and #5:
1G basic ECU MPI circuit function

The relay appears to be toast. I have three questions:
1) Since the EFI relay was working before the transplant, I wonder if I made a mistake and somehow shorted the relay when I rewired the car. Could someone please check my work here:
rewire 4G63 el. connectors to fit 4G37 connectors

2) If something in the harness is shorting the EFI, what can be done to prevent the new EFI relay being shorted as well?

2) Can the damaged/faulty relay be fixed?
 
Last edited:
One more thing regarding measuring 12V on pin #4 and #5 in this MPI relay debugging scenario:
1G basic ECU MPI circuit function

Is it power (12V) across pins #4 and #5 or is it 12V between pin #4 and battery ground as well as 12V between pin #5 and battery ground? In essence, pin #4 and pin #5 become power source.
The way I read the diagram in the link above is that when the coil is energized, circuit is closed, the power goes to both pins at the same time and should be measured between each pin and the battery ground.

Which is it? If I measure voltage individually between each pin and the battery ground, I get 12V on each one of them but the fuel pump does not turn on (no whirling for 2-3 seconds) when the ignition is ON.

Edit: I performed additional MPI tests to see if something else is fried and found nothing:
1) leads from pin #9 to battery(+) and pin #6 to battery (-) i got continuity b/w pins #3 and #2. Then removed lead from pin #6 and got no continuity (infinity).
2) installed lead from pin #3 to the battery(+) and pin #7 to battery(-) and checked for voltage across pin #2 and the battery ground - got 12V. Upon removing lead from #7, got no voltage.
 
Last edited:
Sounds like you have the same problem i do. Something probably spiked voltage and took out the diode. When pins 10(+) and 8(-) are energized they run power to the ecm via pins 4(+) and 5(+) to actually get it to turn on and route power to the necessary sensors. the relay is repairable, but you need to carefully take off the outter bronze housing, locate the ecm side and apply positive voltage past the the diode with pin 8 to battery ground and see if it closes. If so, bad diode. I also took a business card and closed the contact myself before i tested past the diode to see if the ecm would power on and the car would start. that was the problem. Your boost guage should bounce to zero once the ecm receives the power to turn it on from pins 4 and 5. The diode in mine was labeled G1J. That diode seems to be rather uncommon. I found that either an NTE116 or 1N4005 would be more common, suitable replacements. You will need to first make note of which side the grey bar is facing before you start unsoldering the diode (diodes only allow the rated current to travel one way). Then desolder and replace. I figured there was a reason why mine blew in the first place and a blown diode will cost me 50 cents or less and save me more costly repairs to the ecm in the long run.
 
Last edited:
One more thing regarding measuring 12V on pin #4 and #5 in this MPI relay debugging scenario:
1G basic ECU MPI circuit function

Is it power (12V) across pins #4 and #5 or is it 12V between pin #4 and battery ground as well as 12V between pin #5 and battery ground? In essence, pin #4 and pin #5 become power source.
The way I read the diagram in the link above is that when the coil is energized, circuit is closed, the power goes to both pins at the same time and should be measured between each pin and the battery ground.

Which is it? If I measure voltage individually between each pin and the battery ground, I get 12V on each one of them but the fuel pump does not turn on (no whirling for 2-3 seconds) when the ignition is ON.

Edit: I performed additional MPI tests to see if something else is fried and found nothing:
1) leads from pin #9 to battery(+) and pin #6 to battery (-) i got continuity b/w pins #3 and #2. Then removed lead from pin #6 and got no continuity (infinity).
2) installed lead from pin #3 to the battery(+) and pin #7 to battery(-) and checked for voltage across pin #2 and the battery ground - got 12V. Upon removing lead from #7, got no voltage.
Though with your car being NT you wont have a boost guage to bounce to zero. Try what i stated above and it hopefully will solve your problem.
 
Though with your car being NT you wont have a boost guage to bounce to zero. Try what i stated above and it hopefully will solve your problem.

I have taken the cover off and started testing and looking into how the relay functions inside. What I see is that pins #4 and #5 come off the same plate; therefore, if pins #10(+) and #8(-) are energized, the coil closes the circuit, it is physically impossible for pins #4(+) and #5(+) not have power at the same time. Since the ECM receives the power to turn it on from pins 4 and 5, the CEL coming on for 3-4 seconds should be an indication that the ECM is powered ON, checked and functioning. Would that not be the case?

Furthermore, given all that, the correct way to test voltage on pins 4(+) and 5(+) should be between each pin and battery ground...and that checks out in my case. Initially, I looked for voltage between the pins, which I now believe was incorrect.

My understanding is that power for the fuel pump winds up on pin 2(+) on the MPI relay, i.e. if pin #3(+) and #7(-) are energized, the power is sent to pin #2(+) and directly to the fuel pump. Needless to say, even though I get 12V on pins 4 and 5, that still does not wake up my fuel pump.
Would a faulty diode not affect both the ECM and fuel pump at the same time? Since I can only see one diode and my ECM appears to be fine judging by the CEL coming ON for a couple seconds when ignition is turned ON, I am kinda thinking my diode should be fine but cannot say for 100%.
 
Last edited:
I have taken the cover off and started testing and looking into how the relay functions inside. What I see is that pins #4 and #5 come off the same plate; therefore, if pins #10(+) and #8(-) are energized, the coil closes the circuit, it is physically impossible for pins #4(+) and #5(+) not have power at the same time. Since the ECM receives the power to turn it on from pins 4 and 5, the CEL coming on for 3-4 seconds should be an indication that the ECM is powered ON, checked and functioning. Would that not be the case?

Furthermore, given all that, the correct way to test voltage on pins 4(+) and 5(+) should be between each pin and battery ground...and that checks out in my case. Initially, I looked for voltage between the pins, which I now believe was incorrect.

My understanding is that power for the fuel pump winds up on pin 2(+) on the MPI relay, i.e. if pin #3(+) and #7(-) are energized, the power is sent to pin #2(+) and directly to the fuel pump. Needless to say, even though I get 12V on pins 4 and 5, that still does not wake up my fuel pump.
Would a faulty diode not affect both the ECM and fuel pump at the same time? Since I can only see one diode and my ECM appears to be fine judging by the CEL coming ON for a couple seconds when ignition is turned ON, I am kinda thinking my diode should be fine but cannot say for 100%.
The pump should only kick on when the engine starts cranking. I had what I thought was a similar problem as well. Installed a new pump, had it switching on and then lost it .Took the pump back out and reallize I didn't tuck the wires in the tank well enough and the sender pinched the wire. The diode is only to protect the ECM side. I was initially not getting spark and started testing my CAS and crank sensor and reallized I didn't have 5v at either which lead me back to my ECM.
 
The pump should only kick on when the engine starts cranking. I had what I thought was a similar problem as well. Installed a new pump, had it switching on and then lost it .Took the pump back out and reallize I didn't tuck the wires in the tank well enough and the sender pinched the wire. The diode is only to protect the ECM side. I was initially not getting spark and started testing my CAS and crank sensor and reallized I didn't have 5v at either which lead me back to my ECM.

This is a 1990 Laser with a 1992 4g63 and ECU in it. When the ignition was turned ON, one could always hear the fuel pump whirring for a few seconds. I tested the pump by energizing the A55 test connecter and the pump is working. However, when I turn the ignition ON now, the pump does not wake up and when I start cranking, no fuel is delivered.
So I started tracking the problem. Since I can hear the fuel pump operating when I energize A55, I can narrow the problem down to the MPI Relay or its control circuit is faulty. I checked for battery voltage to the MPI relay on terminal #10 and do get 12V. If it's not the MPI relay I dont know what else it could be...other than the harness which came from the '92 Talon and had to be modified for the 1990 connectors, most notably the C67 and C58 connectors.

I should add that I just dropped the 4g63 in and am trying to start the motor for the first time...and the first problem I have run into is the fuel pump not turning on when the ignition is ON.
 
This is a 1990 Laser with a 1992 4g63 and ECU in it. When the ignition was turned ON, one could always hear the fuel pump whirring for a few seconds. I tested the pump by energizing the A55 test connecter and the pump is working. However, when I turn the ignition ON now, the pump does not wake up and when I start cranking, no fuel is delivered.
So I started tracking the problem. Since I can hear the fuel pump operating when I energize A55, I can narrow the problem down to the MPI Relay or its control circuit is faulty. I checked for battery voltage to the MPI relay on terminal #10 and do get 12V. If it's not the MPI relay I dont know what else it could be...other than the harness which came from the '92 Talon and had to be modified for the 1990 connectors, most notably the C67 and C58 connectors.

I should add that I just dropped the 4g63 in and am trying to start the motor for the first time...and the first problem I have run into is the fuel pump not turning on when the ignition is ON.

Terminal 10 on the mpi relay has nothing to do with the fuel pump, only ECM powerup. you are going to be looking for power on terminal 3 with your ignition on. completely separate relays, just under the same roof.

my fuel pump does not engage until the motor starts to crank.
 

Attachments

  • MPI1.gif
    MPI1.gif
    25.5 KB · Views: 1,475
Last edited by a moderator:
... you are going to be looking for power on terminal 3 with your ignition on...

that's absolutely correct. Dang, too obvious, hiding in plain sight. I missed it. Needless to say, I do get power on pin #3.
The diagram shows pin #9 gets power if shift is in "park" or "neutral" for automatic tranny. I cannot make out when pin #9 gets power with 5sp tranny.

Questions:
  1. How does pin #9 get power with 5sp tranny?
  2. Is the engagement of fuel pump regulated by the "fuel pump control" module shown in the diagram? If yes, where is it?
  3. What is the model of your MPI? Mine is MD119202 / E8T07071. Could that be the difference in behaviour? I should add (clarify) that the 1.8 4g37 was priming before cranking and that's what I expected here too. Perhaps the 1g 4g63 with stock ECU does not prime until cranking.
  4. Is it a fact (or rumor) that with ignition ON the CEL goes on while ECU runs an internal check of all circuits and turns OFF when ECU is found in proper and working order?
 
Last edited:
Questions:
  1. How does pin #9 get power with 5sp tranny?
  2. Is the engagement of fuel pump regulated by the "fuel pump control" module shown in the diagram? If yes, where is it?
  3. What is the model of your MPI? Mine is MD119202 / E8T07071. Could that be the difference in behaviour? I should add (clarify) that the 1.8 4g37 was priming before cranking and that's what I expected here too. Perhaps the 1g 4g63 with stock ECU does not prime until cranking.
  4. Is it a fact (or rumor) that with ignition ON the CEL goes on while ECU runs an internal check of all circuits and turns OFF when ECU is found in proper and working order?

I am starting to answer my own questions here: Apparently the answer to question #3 is that my 4g63 with stock ECU is not priming until cranking.
I unplugged the fuel line from the fuel rail and observed at what point it starts squirting gas...absolutely nothing happens (no fuel comes out) until I start cranking.
I am getting the feeling I am looking for solutions to my problem in the wrong thread...:banghead:
... onto checking for a spark and voltage on my coolant sensor...
 
I am getting the feeling I am looking for solutions to my problem in the wrong thread...:banghead:
... onto checking for a spark and voltage on my coolant sensor...
My ordeal here is finished. Turns out that some time during lifting the engine from one car and parachuting into my other car, I ripped one of the leads to the ECT/CTS sensor out and was not getting any signal voltage (~5V). Why one tiny wire telling the ECU what temperature the coolant is before cranking prevents the engine from running is beyond me. That's just not right. I'll let somebody else tell us why that is so important for the ECU to know. I need to figure out now why the 4g63 ECU is not working with my tach...
 
Hey beaver to answer your question no you can not bypass that diode it would back feed power thru that unit and fry anything that is directly related to its function. AKA ecu, fuses, fuel pump. A diode keeps power from back feeding in the other direction.
 
Why one tiny wire telling the ECU what temperature the coolant is before cranking prevents the engine from running is beyond me. That's just not right. I'll let somebody else tell us why that is so important for the ECU to know.

I need to figure out now why the 4g63 ECU is not working with my tach...

With the CTS disconnected the ECU thinks the temp is like -55 degrees. Since fuel doesn't vaporize very well at that temp the ECU injects a lot of it so that some will and ignite but since it really warmer much more does and floods the engine.

If you trying to use a 91+ ECU in a 90 car you will have tach problems. More so if you are also using the 91+ PTM. There are several threads on the various ways to address the issue.

As to the diode. It's there to keep from back powering the electrical via the MPI circuit. I'm not sure it's an issue with the DSM but many Mitsubishi cars use that relay and it might have been required in the first. It's good to have and I'd just replace it if I had a bad one.
 
With the CTS disconnected the ECU thinks the temp is like -55 degrees. Since fuel doesn't vaporize very well at that temp the ECU injects a lot of it so that some will and ignite but since it really warmer much more does and floods the engine.

If you trying to use a 91+ ECU in a 90 car you will have tach problems. More so if you are also using the 91+ PTM. There are several threads on the various ways to address the issue.

As to the diode. It's there to keep from back powering the electrical via the MPI circuit. I'm not sure it's an issue with the DSM but many Mitsubishi cars use that relay and it might have been required in the first. It's good to have and I'd just replace it if I had a bad one.

Thanks for clarifying issues re CTS - now it makes sense.
Re diode - since I am dealing w/ a lot of issues installing 1992 4g63 into 1990 4g37 setup, I tried to figure out how the MPI relay works by reading the diagram and it looked like the ECU would burn out if direct 12V current was applied to the fuel pump instead of the A55 port - which is what I was going to do to make sure it works. Dont quote me on it thou - that was my impression.

Your comment re ECU/cluster comes in prescient time as this is what I will be attacking now - I had to deal with 1992 tranny issues: the 1990 linkages, cable bracket and shifter was not compatible with the 1992 tranny setup. I should've kept the '90 tranny and slap it on the 4g63...here comes 15y.o. wisdom left here to the rescue:
90 shift linkage vs 91-94 shift linkage

The way this is going, there will be no 1990 setup left in this car...but at least the engine runs and all the gears in the tranny go in like hot knife into butter - smooth!

I also noticed my radio stopped working. The clock works which tells me the power is getting to it; however, the radio and tape deck is dead. Is this issue somehow related to my swapping in the 1992 engine harness? Did I make a mistake along the way in rewiring connectors or something like that? It's a low priority but I wonder if it is all related?
 
I tried to figure out how the MPI relay works by reading the diagram and it looked like the ECU would burn out if direct 12V current was applied to the fuel pump instead of the A-55 port - which is what I was going to do to make sure it works. Don't quote me on it thou - that was my impression.

You'll note that the MPI and Fuel Pump sides of the relay get power from different points (circuits) in the electrical system, the MPI via the MPI fuse directly from the battery and the Fuel Pump via the Ignition Switch and Ignition Fuse. The path between the two circuits is via the battery and the only direct connection between the fuel pump power and the ECU is through the occasionally undocumented ECU pin 13 which is set up to sense when battery voltage is applied to it. While it's possible to burn up parts of the ECU by applying battery voltage directly to pins that's not one of them.

I also noticed my radio stopped working. The clock works which tells me the power is getting to it; however, the radio and tape deck is dead. Is this issue somehow related to my swapping in the 1992 engine harness? Did I make a mistake along the way in rewiring connectors or something like that? It's a low priority but I wonder if it is all related?

The Clock is connected to the unswitched power (room fuse?) to the radio and ECU, IIRC. Look at the other switched power to start.
 
Thanks for clarifying issues re CTS - now it makes sense.
Re diode - since I am dealing w/ a lot of issues installing 1992 4g63 into 1990 4g37 setup, I tried to figure out how the MPI relay works by reading the diagram and it looked like the ECU would burn out if direct 12V current was applied to the fuel pump instead of the A55 port - which is what I was going to do to make sure it works. Dont quote me on it thou - that was my impression.

Your comment re ECU/cluster comes in prescient time as this is what I will be attacking now - I had to deal with 1992 tranny issues: the 1990 linkages, cable bracket and shifter was not compatible with the 1992 tranny setup. I should've kept the '90 tranny and slap it on the 4g63...here comes 15y.o. wisdom left here to the rescue:
90 shift linkage vs 91-94 shift linkage

The way this is going, there will be no 1990 setup left in this car...but at least the engine runs and all the gears in the tranny go in like hot knife into butter - smooth!

I also noticed my radio stopped working. The clock works which tells me the power is getting to it; however, the radio and tape deck is dead. Is this issue somehow related to my swapping in the 1992 engine harness? Did I make a mistake along the way in rewiring connectors or something like that? It's a low priority but I wonder if it is all related?

Sorry i ghosted for a bit. Got sidetracked working on other stuff for a bit and had to take a break on the dsm. I replaced the diode on the ecm power side of the mpi relay, bench tested fine, put into the car and..... nothing. hopefully i can find a little extra time this week to finally figure it out. The car did come with an HKS turbo timer and it looks like they used the correct harness for it, but im guessing my problem lies somewhere in that mess of wires. When i close the contacts on the relay, boost guage zero's out and it will start, run for 20 seconds or so and die. Guessing my FPR is shot, so trying to snag an aftermarket at the moment. Will update when get somewhere.
 
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