Probably not this simple, although I will chime in. Took our 340 to our regular mechanic 2 weeks ago to have drives serviced. Had plugs and wires changed on port motor, did stbd last year, and when they put it back in water and we went for test run that motor, port, would not get over 2800 rpms. Took it back to the marina and it turned out to be a bad set of NEW plug wires. Not sure how they tested them. There were no errors codes, alarms, nothing from the Vessel View. It just said "System ok".
So I asked what they did to figure it out. They switched plug wires with the other motor and the problem followed. Changed the wires and all good. Guess question would be are you absolutely positive you are firing on all cylinders? With no codes, it has to be something fuel or firing related.

 

This photo from Post #81 is what is troubling me about the engine wiring. I can't tell what sensors these wires go to but in retrospect.....if one of these connections goes to a discreet sensor.......you will have a problem.

 

Those wiring pigtails usually go to analog sensors ie water temp or oil pressure.

 

Are you able to get a graph of the data from idle to the 2000 RPM point? I'd be curious if you hammer it down does it go slightly higher then down to 2k as if the system is controlling to that point. If so, it really has to be a safe mode. It's not likely a fuel flow or air flow issue would go to exacly 2000 rpm. I'm with outhers, just because the scan tool shows no safe modes (or faults), it may not be decoding the bits/bytes for those fields.

 

Check this out....page 126 of MEF1-1 manual.

Notice: The sensors that trigger Safe mode do not throw DTCs (It was 1996). You would need a tester (which there is a very direct note about at the top of the first page) that could see the voltages on those pins (J1-19, J1-20, J-21, J-22, J1-4 and J1-6). Even then, you would still need to know which sensors the ECM was expecting to see.

About the only way to solve it is go to each location where a sensor is supposed to be and locate it/check the wiring for continuity back to the ECM. Fortunately there are only six of them and three should be on the engine and two should be on or near the drive. The last one is on the helm if it is installed.

 

This thread is like a Masterclass. WOW! The wealth of knowledge on CSR is really something!

 

Prior to sending the injectors out for cleaning I paid a local mechanic $300 to plug the boat into his $5000 computer diagnostic device. None of the discrete switches were showing as being triggered and we confirmed that it did not have itself in safe mode. The mechanic stated that looking at the computer diagnostics he could not find anything wrong with the engine and his suggestion was to send the injectors out for testing as it seemed like a lean running condition and possibly one or more injectors were bad.

I think the injector trigger duration being insufficient is likely the best lead but what could possibly cause that.

The loose wires mentioned in a previous post I was able to determine per the wiring schematic that the grey one is supposed to be loose. It’s a hookup for a Mercruiser diagnostic tool of some sort for getting a reading on a tachometer. The other 12 gauge wire doesn’t exist in the schematics that I have. Testing it with my voltmeter doesn’t reveal any voltage but if I ground it engine alarms begin to sound nonstop. All the sensor wires are otherwise accounted for and are plugged in to their appropriate sensors.

 

...

I think the injector trigger duration being insufficient is likely the best lead but what could possibly cause that.

The loose wires mentioned in a previous post I was able to determine per the wiring schematic that the grey one is supposed to be loose. It’s a hookup for a Mercruiser diagnostic tool of some sort for getting a reading on a tachometer. The other 12 gauge wire doesn’t exist in the schematics that I have. Testing it with my voltmeter doesn’t reveal any voltage but if I ground it engine alarms begin to sound nonstop. All the sensor wires are otherwise accounted for and are plugged in to their appropriate sensors.

Put an o-scope on the injectors and see if the PW matches the ECU display. If they do, it's not that. The PW being displayed is just based on map/baro/rpm/temp...

 

I would do the following now : Tie up the boat carefully and perform a stall test ( wot with boat moored to the dock ) and look how high it revs .

If the engine is limited by injector on duration ( so fuel flow ) i expect it to rev a bit lower than 2000 with a stationary boat due to higher resistance .

If it limits here also at 2000 exactly i suspect somehow of a hidden safe mode triggered by the ECU .

No need to say : take care for security and your clamps / mooring . Beyond that this test is free of charge and i think will answer what i mentioned above.

 

Did someone already ask - is the throttle lever opening the throttle body and TPS all the way??

 

Per the data throttle position sensor is reading that it is open (and shows closed at idle) so I would think it's not the problem. I suppose the shaft could be rotating and the blade not opening but not likely to occur....

 

I had the '92 SW, did the end of the '89 which was MAF package, then the '90 which was the 1st speed density with the L98. The '92 was not much different controls wise. Drove those on test trips in probably 20 states around the country (hot/cold/alt/sea level)... OP sorry for the derail!

 

@Ericinmich & @PlayDate (and others), love your post. I like to start with basic's and maybe I missed it but did @Hoplite808 ever get back to us with actual mechanical vacuum gauge readings that I asked about before? Again excuse me if I missed it, but it is very important to me.
This is tough, but at this point, on the internet, with limited info, i.e. what @Ericinmich spoke of such as graphed live data and scoping the injectors (and any and all other suspect sensors or actuators) I'm going to recommend inducing codes. If we at any point suspect the harness, or any connection between the ECM and any sensor or actuator I would not rely on the Ohm meter/resistance test, as we are only using a 9V battery with little to no respectable amperage draw. I'm not really a fan of voltage drop test on some circuit's either. I prefer load testing each and every circuit of my suspect sensor or actuator after failing an induced code test.

 

Okay, so the screen shots I shared were pretty much all the data that is available from the Fox Marine unit via the app on my phone which I screen shot. There is another screen that will allow me to reset codes but that is pretty much it. The guy I can pay $300 to, who can run a more advanced diagnostic will not give me a printout. He only would let me view live data. It showed that none of the discrete switches were triggered and it did allow induction of codes which he went through and reset and could find no faults. This is why I spent more money sending the injectors cross country to be cleaned and serviced. However this did not remedy the problem. Also I can either have the Fox Marine unit plugged in or the engine in service mode not both.

 

I would give Service Mode a try. I do expect it to create an idle issue but you should be able to handle that. It basically extends the IAC to close the throttle plate and drop the idle rpm’s.

In Service mode….it should ignore any ECM limits but you will be the first to know..

Anything above 3,500 rpm and I will buy the first round!

 

Also, in Service Mode the automatic spark advance feature is disabled which means it may be sluggish to spin up. You could counter some of it by dialing in a few more degrees of advance.

 

Another basic question - @Hoplite808 replaced the injectors with some purchased off Ebay. Unclear, did he send the original injectors for cleaning and testing or the Ebay ones?? Suffice it to say - he had the issue before and after injector changing/cleaning.
The reason I ask - are the Ebay injectors (if that is what is installed) the correct ones for his application? If for example the Ebay injectors are 35 LB/Hr and the OEM are 60 LB/Hr then an RPM problem will ensue and you'll never know with any scan tool. The easiest way to determine is to see what the ECM is driving as a duty cycle at the stall RPM. If it is driving a 20% or less duty cycle at that 2000 RPM then we could probably remove the injectors as the problem. The O-Scope idea would be the tool for this.

 

Tom,

We went through the injectors diagnostics and finally convinced the OP to spend the money to send the OEM injectors to a shop on the mainland. This was reinforced by a local mechanic that the OP paid to diagnose the engine. The mainland shop basically found nothing wrong with the injectors (Post #223). OEM injectors are 24lbs/hr.

When I enter his engine parameters in my handy dandy fuel injection calculator it matches his Injector On duration at idle and 2000 rpm. It predicts that the Injector On duration should be around 13ms at 4000 rpm.

At 27 years old this MEFI-1 system is an antique but I was fortunately able to find a manual for it. Buried in this manual is how RPM reduction is triggered and how the ECM responds to it. Effectively anyone one of six "discreet" sensors can trigger it. Once triggered "half the fuel injection system is disabled" to keep it below 2000 rpm. Up to the trigger point....everything looks fine. At the trigger point.....everything advances normally except for the fuel injectors. I'm not sure how disabling half the injectors would show up on any monitoring device. Nominally.....the max Injector On pulse is 15ms or so they could be limiting the pulse or physically turning off a bank of injectors which was a popular trick in the 1990s.

Back in the 1990s we called this "limp home mode" in fuel injection systems. Over time the ECMs became far more capable of reporting and identifying problems taking most of the mystery out of them.

Short of having access to a Rinda, we are putting our best thoughts on the table. I hope Service Mode allows us to get a breakthrough. While details are scarce.....Service mode doesn't seem to care about any soft set engine problems and relies on internal table information to run the engine. We will see.

 

The limitation of what the fox can see is the limitation of the MEFI-1's ability to broadcast the parameter. Bluetooth can show more parameters than the MEFI bus.

 

I'm not sure this is the case. A bluetooth scanner can only provide data that exists on the ALDA bus, which is the same data available to any scan tool. That said, every scan tool that is looking at the bus may not be provide all the data, or the data in the correct format.

 

Hoplite, it was still good to have the injectors rebuilt. Based on how you had different results when swapping injectors earlier, it was very logical to think there was a problem with the injectors. Having them rebuilt eliminated a variable. You didn't waste your money on that one - chalk it up to preventative maintenance.

Reading through the various tables is not something I have as much experience with as some of your other "helpers", but I keep going back to how changing your injectors caused different results. I wonder if the wiring/connectors could be an issue?

As a cheap workaround to check if the injectors are actually firing... what about using a mechanics stethoscope on them? Or at least a long-handled screw driver or wooden dowel.

 

Could you consider my suggestion at post #274 - the 'stall test' ?

I guess we all agree we want to learn if the boat is 1) simply lacking power by obviously limited injector on time command ( fuel flow restricted )
OR
2) if something triggers the ECU to retard at 2000 rpm in gear .

Reving the boat in gear moored ( securely !!! ) should really bring the answer without any injector pulling .

If case 1) is true i would expect the engine to reach noticeable less than 2000 rpm,s now .

If the engine also here revs to 2000 it must be case 2) .

It is for me the most simple way to really narrow it down further.

 

Could you consider my suggestion at post #274 - the 'stall test' ?

I guess we all agree we want to learn if the boat is 1) simply lacking power by obviously limited injector on time command ( fuel flow restricted )
OR
2) if something triggers the ECU to retard at 2000 rpm in gear .

Reving the boat in gear moored ( securely !!! ) should really bring the answer without any injector pulling .

If case 1) is true i would expect the engine to reach noticeable less than 2000 rpm,s now .

If the engine also here revs to 2000 it must be case 2) .

It is for me the most simple way to really narrow it down further.

First....it is his boat and he can test it any way he wants. I believe your idea has merit but at this stage either the ECM is turning off half of the injectors at 2000 rpm or it is not. Revving up a drive and engine in a marina slip would be exciting to watch but I'm pretty sure the marina owners would be concerned about the stress on the docks/piers and the impending disaster if a line broke.

It would make it possible for the OP to run the boat by himself safely if he could secure it which is why it has merit.

 

If half the injectors were turning off... wouldn't that be quite obvious based simply on the sound (stumbling) of the engine?

 

There's no way half of the injectors are turning off!

 

I have in this discussion one thing i struggle to understand :

when for whatever reason ( limp mode or anything else ) the fuel delivery to all cylinders is limited by ECU but simultany the user WOT,s the engine by physically fully opening the throttle body ( no dts here which controls the throttle body , just a physical connection via linkage ) - is here not a massive lean situation created ??? Way too much air sucked with too little fuel sprayed in ???

How does a limpmode , guardian or however we call it work in a non dts mpi ???

 

I have in this discussion one thing i struggle to understand :

when for whatever reason ( limp mode or anything else ) the fuel delivery to all cylinders is limited by ECU but simultany the user WOT,s the engine by physically fully opening the throttle body ( no dts here which controls the throttle body , just a physical connection via linkage ) - is here not a massive lean situation created ??? Way too much air sucked with too little fuel sprayed in ???

How does a limpmode , guardian or however we call it work in a non dts mpi ???

Safe mode by shutting of injectors to a bank will make it a 4 cyl, and the other 4 become air pumps, no burn at all so no lean event.

As far as how a V8 runs in that mode, ask anyone with a GM AFM engine, it makes a different sound through the exhaust but it's still smooth. The transition can usually be felt though.

 

Not sure if this has been mentioned. I am coming from the auto industry, so similar. Many times when these crazy issues occur and can't be diagnosed, it turns out the computer was the issue. If this was mentioned before, sorry.

 

While we are waiting for the Service Mode verdict to come in, there is another issue that he raised that still is bothering me.

OP responded in his post #261: The other 12 gauge wire doesn’t exist in the schematics that I have. Testing it with my voltmeter doesn’t reveal any voltage but if I ground it engine alarms begin to sound nonstop.

It seems to be a safety sensor wire of some kind that is normally open....maybe for engine temp given its location?

 

Gee, guys...
A thread gets this long and people rarely read the whole story but give the OP a break.
He has gone thru head issues, machinist issues, mechanic issues, and is trying to just get it fixed.

First of course he "hears it" it's weak its only putting out about 50% power.
That doesn't mean it's going to be knocking or sputtering or otherwise noticeable by ear.
These were first generation ECM, they were MEFI, NOT "MEFI 1" because there was not concept yet for 2, 3, 4, 555.
They are extremely basic, very little self diagnostics.
There is NO Guardian mode, that was not introduced until v3

The "Power Reduction" mode is clearly Off
This mode is NOT triggered by the ECM itself, this is triggered by a number of optional external "switches"
High temp:No, Low Oil Pressure:No, Low Oil level:No, a physical switch could also be connected, a "Get home mode"
The E-Stop Input:Off (think fire bottle shutdown)
I have NEVER seen any of these MEFI 1 or 2 options actually used on a boat, early boats from Carbs days had alarms for these items already.

The other external trigger is if the Distributor Ignition Module cant communicate it's status to the ECM, the ECM will stay in a startup mode, I described this way back in the thread.

Yes the engine will run "fine" (but 50% power) with half the injectors out.
These early ECMs had only two injector driver circuits, four injectors on each circuit.
But it is NOT left bank, right bank, it's two on each side. Roughly in firing order.

So as was stated before it becomes a fairly smooth four cylinder engine.
But no 3.7L engine is going to get that boat on plane.

We are back to WHY.
Why low power, Why limited open time,
There are things some of us could do the the OP just doesn't have the tools for.

At this point we know the injectors are good, fuel pressure is good, basic spark is good.
I would break out my oscilloscope and be actually checking the output of the two injector circuits and see if that square wave matches what the computer THINKS it is doing. One injector driver circuit may be bad. But I own one, not many people do.

If I were going to throw a few parts at it I would start with the Ignition Module, only $95 dollars from Sierra
Sierra 18-5107-1 Marine Ignition Module - because it may not be sending the correct signal back to the ECM.

Then at this stage I would just send out the ECM, MEFI 1 are non-existent new.
Buying some E-Bay unit for hundreds or thousands is a crap shoot.
Mercruiser had a kit to use the MEFI 4 but it needed an adapter cable & programming, which I cant find anywhere today.

https://obd2allinone.com/ecmrepair.asp - test and repair them for $100 (test) to about $600 repair.
This is due to the fact it is rarely the computer that goes bad but rather power transistors.

@Hoplite808, hang in there !

 

Gee, guys...
A thread gets this long and people rarely read the whole story but give the OP a break.
He has gone thru head issues, machinist issues, mechanic issues, and is trying to just get it fixed.

First of course he "hears it" it's weak its only putting out about 50% power.
That doesn't mean it's going to be knocking or sputtering or otherwise noticeable by ear.
These were first generation ECM, they were MEFI, NOT "MEFI 1" because there was not concept yet for 2, 3, 4, 555.
They are extremely basic, very little self diagnostics.
There is NO Guardian mode, that was not introduced until v3

The "Power Reduction" mode is clearly Off
This mode is NOT triggered by the ECM itself, this is triggered by a number of optional external "switches"
High temp:No, Low Oil Pressure:No, Low Oil level:No, a physical switch could also be connected, a "Get home mode"
The E-Stop Input:Off (think fire bottle shutdown)
I have NEVER seen any of these MEFI 1 or 2 options actually used on a boat, early boats from Carbs days had alarms for these items already.

The other external trigger is if the Distributor Ignition Module cant communicate it's status to the ECM, the ECM will stay in a startup mode, I described this way back in the thread.

Yes the engine will run "fine" (but 50% power) with half the injectors out.
These early ECMs had only two injector driver circuits, four injectors on each circuit.
But it is NOT left bank, right bank, it's two on each side. Roughly in firing order.

So as was stated before it becomes a fairly smooth four cylinder engine.
But no 3.7L engine is going to get that boat on plane.

We are back to WHY.
Why low power, Why limited open time,
There are things some of us could do the the OP just doesn't have the tools for.

At this point we know the injectors are good, fuel pressure is good, basic spark is good.
I would break out my oscilloscope and be actually checking the output of the two injector circuits and see if that square wave matches what the computer THINKS it is doing. One injector driver circuit may be bad. But I own one, not many people do.

If I were going to throw a few parts at it I would start with the Ignition Module, only $95 dollars from Sierra
Sierra 18-5107-1 Marine Ignition Module - because it may not be sending the correct signal back to the ECM.

Then at this stage I would just send out the ECM, MEFI 1 are non-existent new.
Buying some E-Bay unit for hundreds or thousands is a crap shoot.
Mercruiser had a kit to use the MEFI 4 but it needed an adapter cable & programming, which I cant find anywhere today.

https://obd2allinone.com/ecmrepair.asp - test and repair them for $100 (test) to about $600 repair.
This is due to the fact it is rarely the computer that goes bad but rather power transistors.

@Hoplite808, hang in there !

I believe he has already replaced those parts in Post #1.

In regards to ECM focus....the theory is that the ECM is going into RPM Reduction mode and that fits almost exactly with what he is experiencing. As to what is triggering it.....we just don't know....yet. We just know the injectors On duration are not advancing when everything else is. I personally don't believe the Fox Marine tool is able to determine the RPM reduction mode without seeing all the discreet sensors. It only sees two out of six.

I know the OP has done a complete top end rebuild and the title of the thread says it all. It was all of our collective wisdom who told him to have the injectors serviced....and while it was a good thing to do....it didn't solve the problem.

I'm hopeful we are on the right track with the ECM but the real challenge is being able to override a set parameter in it. @Ericinmich has first hand knowledge of the system and I am hopeful that Service mode will do that.

I appreciate how much money and time he has invested in this engine but this is a tough problem to solve without having your hands on the engine and most importantly the right diagnostic tools to work with.

 

If it is indeed lacking fuel, how about “adding” fuel when you hit the 2,000 rpm limit.

get a helper and a can of carb cleaner with a straw. Spray it into a vacuum hose or even the throttle body and see if you can exceed the 2,000 rpm limit under load.

Access isn’t easy in a marine application, you may have to get a little creative in how you get the carb cleaner into the intake.

 

Why not replace all six sensors? That would only leave the computer in this scenario. Now I can officially be labeled as a parts cannon.

 

That certainly would be one approach and may be the answer but I doubt he has all six installed.

About the only two things that fit the problem now are the ECM/sensors and a timing chain that jumped a tooth.

He is getting closer to where the problem is.

 

Here is some data if this helps from my starboard 454 horizon, idle, 2000 rpm then 3600

 

You need to check your alternator and grounds
Voltage is low