7.4mpi just started running rough and with gas coming out exhaust

nola mike

Vice Admiral
Joined
Apr 22, 2009
Messages
5,362
My injector plugs are numbered and connected correctly. 2, 4, 6 and 8 on the starboard side.
I will see if the voltage readings ever change.
I will also look to see if any colors match. Maybe the banks are really different than just evens and odds. The spark plugs were definitely only noticeably black on the even side.
Certainly sounds like a mechanical issue if 2 injectors from each side are fired separately from the other 2. But the voltage thing doesn't make sense then
Agree the difference in voltage might be the static state of the ECM, don't know, just guessing. Would be interesting test to remove on of the injectors and see if the injector would fire just with the key ON. Might try on/off a few times to see if things change
That might be helpful
 

dubs283

Vice Admiral
Joined
Jul 27, 2005
Messages
5,322
Mine has the unused regulator but I don’t recall any fuel in the vacuum hose.
The damper can fail without fuel present at the vacuum line. I think the only test aside from installing a new part is to rig a fuel line to that side of the motor/fuel rail that you can manually regulate.

I've had this issue once and iirc that was one of the tests performed. It may mean destroying the fuel rail to rig the test. It was many years ago so my memory has faded a bit. Probably at my wits end by that point and basically knew the rail needed replacing so creating the testing circumstances was moot
 

Ozshadow

Seaman Apprentice
Joined
Jun 23, 2013
Messages
46
Should both j1-1 and j1-17 have power when the ECM is removed and the key is on?

This is a MEFI 3
 
Last edited:

Ozshadow

Seaman Apprentice
Joined
Jun 23, 2013
Messages
46
This is good practice. Keep going, next is key off check continuity on injector leads to ecm/ ground.

It's rare for an ecm to go bad. It does happen but you need to rule out all possibilities besides spending big $ just to find out there's a broken/melted/unconnected wire or something else causing issues

I forget, have you investigated fuel pressure/regulatory stuff?
Bumping up the resistance setting, all read about 293 resistance Key off.
 

alldodge

Moderator
Staff member
Joined
Mar 8, 2009
Messages
42,054
B+ = Bat voltage
MEFI-3 J1-1 is B+ injector driver and J1-17 is B+ injector driver
Key ON or Running should be B+

Yes should have 12V because it gets B+ thru the injectors to the pins. The injectors won't reduce the voltage next to nothing because there is no current draw

Think I would look into sending the ECM out to have it tested. OBD Diag or Boats Unlimited
 

ScottinAZ

Master Chief Petty Officer
Joined
Jun 25, 2009
Messages
831
B+ = Bat voltage
MEFI-3 J1-1 is B+ injector driver and J1-17 is B+ injector driver
Key ON or Running should be B+

Yes should have 12V because it gets B+ thru the injectors to the pins. The injectors won't reduce the voltage next to nothing because there is no current draw

Think I would look into sending the ECM out to have it tested. OBD Diag or Boats Unlimited
I think I would have to agree with you. IIRC, most of the ECU's out there control the GROUND side of the injectors, so current is not running through the ecu. If you have one side of the injectors "off" and the other side is "on" either you have the extremely bad luck of catching it right at the firing point, OR you have a short to ground, essentially putting the injectors at 100% duty cycle (always on). If I read the above responses correctly, when you crank the engine, one side will flash (as it should) and the other side is on full time (100% duty cycle again). since the switch is on the ground side, and the ecu is that switch, I would be very strongly considering a bad ecu.

Before I condemn the ecu however, I would be probing the ecu side of the injectors (ground side) for a simple wiring short to ground. Doesnt take much chafed insulation to cause major issues.
 

Ozshadow

Seaman Apprentice
Joined
Jun 23, 2013
Messages
46
I will send it to OBD. I know the banks fire in a different order but the odd side holds 12v on both injector leads until it grounds and fires.

The even side is 12v on the power side and 8v on the ECM side which makes it act sorta like a ground. The noid light is on solid but dim. All it takes is one injector connected to see it. It can be any on the even side.

The only codes in the ECM are from me disconnecting the sensors.

The only wire with a worn spot was retapped and wasn’t shorting.

At this point I am sending it out. Thanks for all the help. I’m just hoping something hidden in the magic box is the clear answer and is repairable.
 

nola mike

Vice Admiral
Joined
Apr 22, 2009
Messages
5,362
I will send it to OBD. I know the banks fire in a different order but the odd side holds 12v on both injector leads until it grounds and fires.

The even side is 12v on the power side and 8v on the ECM side which makes it act sorta like a ground. The noid light is on solid but dim. All it takes is one injector connected to see it. It can be any on the even side.

The only codes in the ECM are from me disconnecting the sensors.

The only wire with a worn spot was retapped and wasn’t shorting.

At this point I am sending it out. Thanks for all the help. I’m just hoping something hidden in the magic box is the clear answer and is repairable.
This still doesn't make sense if the injectors aren't on an even/odd circuit. Please keep us updated.
 

dubs283

Vice Admiral
Joined
Jul 27, 2005
Messages
5,322
This still doesn't make sense if the injectors aren't on an even/odd circuit.
Correct. I'm still leaning towards a short to ground on the problem side of the engine wiring harness.

OP: your continuity test sounds like you checked through the harness still connected to the ecm? You need to check between the injector leads and a ground with the harness disconnected, key off/on
 

Ozshadow

Seaman Apprentice
Joined
Jun 23, 2013
Messages
46
The only other thing I can think of is the opposite side of the ECM - not the wires going to the injectors. Maybe a pin isn’t being powered or receiving full power? Maybe a relay?
 

nola mike

Vice Admiral
Joined
Apr 22, 2009
Messages
5,362
Correct. I'm still leaning towards a short to ground on the problem side of the engine wiring harness.

OP: your continuity test sounds like you checked through the harness still connected to the ecm? You need to check between the injector leads and a ground with the harness disconnected, key off/on
Where? I don't see a common wire that would only affect 1 side.

Edit: unless splice 110 and 111 are physically close to each other, and you have all 4 of the affected wires right next to each other and chaffed
 

Ozshadow

Seaman Apprentice
Joined
Jun 23, 2013
Messages
46
Where? I don't see a common wire that would only affect 1 side.

Edit: unless splice 110 and 111 are physically close to each other, and you have all 4 of the affected wires right next to each other and chaffed

I think this post explains what we we all missing. See #5 by John Marzetti of Boats Unlimited. My engine is an L29.


He explains the unique injector banks.
 

alldodge

Moderator
Staff member
Joined
Mar 8, 2009
Messages
42,054
Important - injector banks for L29 is 1,3,5,7 / 2,4,6,8. Injector banks for for all other MPI models are 1,4,6,7 / 2,3,5,8. MEFI
BUP does know his stuff and is a skilled Tech. The above is not in any manual I've ever seen but I'm sure its correct. This would explain the issue and I'll make note of it for future reference

Does look like its your ECM or a short
 

Ozshadow

Seaman Apprentice
Joined
Jun 23, 2013
Messages
46
FINAL UPDATE - fried injector driver. Not repairable.

How difficult is removing the entire harness between the ECM and injectors? I know the short injector harness come off easily. I want to look at every inch of wire before I put in a $$$ ECM.
 
Last edited:

alldodge

Moderator
Staff member
Joined
Mar 8, 2009
Messages
42,054
The motor harness goes everywhere so it will need to be disconnected from everything
 

Ozshadow

Seaman Apprentice
Joined
Jun 23, 2013
Messages
46
The motor harness goes everywhere so it will need to be disconnected from everything
How should I test the harness with a multimeter from the injector to the ECM connector to be absolutely certain I don’t have something that will cause a problem for the new $2500 ECM?
All injectors read 12.9 resistance from what I recall - all were the same.
I am just second guessing everything since I can’t afford a mistake.
 

dubs283

Vice Admiral
Joined
Jul 27, 2005
Messages
5,322
You need to test for continuity between the injector leads and whichever pins connect accordingly to the ecm. Oem wiring diagram will be needed to figure out which pins need to be tested
 
Top