Mercruiser 4.3L MPI 2004 Hard Starting

[mat5751]

Hello All

I have twin Mercruiser 4.3L MPI V6, 2004 (OW065549 and OW065554) with 230 hours each that are hard to start when hot/warm.

They will turn over but will take several minutes to fire. When they do eventually start, the exhaust smells rich for a few moments.

The workaround is to start in neutral at WOT then quickly throttle down. Both engines start straight away with this workaround.

The issue occurs in both engines, after they have been sitting for between 5 minutes to about 2 hours after use.

Both engines start straight away when cold.

The engines were recently serviced, show no fault codes when plugged into the computer and compression was checked and is good.

I have found the Service Bulletin 99-7 (http://www.boatfix.com/merc/bullet/99/99_07.PDF) regarding vapor lock. I do not believe that it is however as I have tried, all without any improvement at all:
- blowers
- idling after running for several minutes
- opening the engine hatch
- I have never used fuel with any ethanol in it.
And, when I do try to start at WOT, it starts straight away.

I have replaced the IAC values on both engines but no improvement.

I suspect that it may be one of the following:
- Temperature Sensor - even though there is no fault code, the sensor may have some fault which tricks the computer to think that the engines are warm (when they are not) and to inject more fuel than necessary. I think this because the hard starting is cured with WOT starts.
- MAP Sensor - again for the reasons above, an incorrect temperature reasoning causing more fuel being injected than necessary.
- Low pressure Fuel Booster pump - I don't think so, because the engines start fine at WOT
- Fuel regulator - Maybe if the regulators are impacted by temperature......

Before I start throwing parts at it, any thoughts?

Any assistance would be great.

Thanks all in advance.

 

[alldodge]

engines start fine at WOT

If this only happened after an hour or two then it could be vapor lock, but happening with only 5 minutes and also having heavy fuel, I would lean to a fuel issue.

An issue to keep in mind for me is they are both doing it

Check the fuel pressure at the rail, should have 43 psi at 1800 rpm.

Also turn the key to ON without starting and read pressure, the pressure should hold for a little while without dropping. If it drops off quickly then could be injectors or regulator

 

[Scott Danforth]

Check your fuel pressure to see if you have a leaking injector

 

[alldodge]

When this is happening and before you move throttle to WOT, look at the tach and see if you can see at least 300 rpm

 

[achris]

When were spark plugs replaced? Have you had the Diacom software hooked up see if the ECU is reporting any problem (past or present)?

Chris....

 

[tpenfield]

:welcome: to iBoats . . .

As Chris said, the computer really needs to be connected while this is happening to see the actual values of the sensors.

Based on what you described, the ECU will have an injector pulse while starting that is about 5X the pulse when the engine is running at idle and at its operating temperature. Your movement of the throttle to WOT changes the Throttle Position Sensor (TPS) so the ECU is probably responding with a different injector pulse, presumably one that is sufficient to start the engine.

My guess is that the engine does not 'know' it is in starting mode (i.e. not running) when it is warm. As to which sensor(s) would indicate that to the ECU, it is hard to say without the computer.

You mentioned that both engines are hard to start, and there typically is a data link between the engines to give them a master/slave relationship. The computer will also indicate such.

If you plan on doing your own maintenance on the engines, you may want to get the Rinda/Diacom software now, as it will pay for itself right away vs. paying a shop to look at the engines.

 

[mat5751]

Thanks all very much for your posts. I appreciate your efforts.

@ alldodge and Scott Danforth - I will check the fuel pressure. Alldoge, Ill also check and see if it gets 300 RPM

@ achris - I have checked my invoice for the last service in August and I got (paid for at least...) a new set of NGK BPR6EFS15 plugs, I am not sure if they were gapped correctly. As for faults, I have VesselView and no active faults show. I recently had another marine mechanic connect his computer (not sure if it was Diacom), it also said no current faults exist. As for past fault codes, I dug out an old print out and the following codes existed historically (before they were cleared at the service in August). Some of which can be explained by the lack of some sensors fitted (and a since replaced broken neutral switch on the port engine) but I am not sure about some of the others.......

PORT HISTORIC CODES PCM555:

0. TRIM CKT HI
1. STEER CKT LO
2. SEA TMP CKT HI
3. FUEL LVL CKT HI
4. PITOT CKT LO
5. NEUTRAL OVERSPE
6. BAT VOLTS LO
7. ECT TMP CKT HI
8. FUELPUMP RLY CK
9. BLK PSI CKT LO

STARBOARD HISTORIC CODES PCM555:

0. TRIM CKT HI
1. STEER CKT LO
2. SEA TMP CKT HI
3. FUEL LVL CKT HI
4. PITOT CKT LO
5. BLK PSI LO
6. BAT VOLTS LO
7. GUARDIAN
8. OVERSPEED
9. EST1 OPEN

I have the detailed senor results for the above codes at the time/hours they happened if required.

Thanks all.

 

[mat5751]

@ tpenfield - Hi Ted - Great to be aboard. Sorry, I didn't see your post before I replied to earlier to alldodge, Scott Danforth and achris. I agree, it does seem like a sensor could be causing the issue.
I am not physically up to doing my own maintenance and have relied on the marina to do so till now - but this one has them stuck.
Any chance you could share some info on what they should look for (on Diacom) so that I can point them in the right direction please? All I am being told by the marina is that there are no error codes......
Thank you

 

[Scott Danforth]

a good mechanic will simply look at the diacom data and compare it to what should be there.

a bad mechanic will simply plug in and say - no error codes.....i dunno

 

[havoc_squad]

a good mechanic will simply look at the diacom data and compare it to what should be there.

a bad mechanic will simply plug in and say - no error codes.....i dunno

+1 to this.

If the injectors are not leaking, I think a clear sign of a sensor or computer issue would be the fuel trim levels being stuck at a cold engine situation, much like what would happen if the choke was stuck on or a fuel enrichment circuit was stuck on for a warmed up engine.

Otherwise if fuel trim isn't the problem and no injector leaks on a warm engine idling, I would lean towards throttle body not clean, causing a rich/flooding condition if it is more than the engine computer will allow to compensate using the IAC.

 

[scoflaw]

There is no fuel that passes through the throttle body. If no codes are being set it's a fuel problem. The 555 is quite savvy about sensor and electrical issues

 

[alldodge]

Thanks all very much for your posts. I appreciate your efforts.

@ alldodge and Scott Danforth - I will check the fuel pressure. Alldoge, Ill also check and see if it gets 300 RPM

@ achris - I have checked my invoice for the last service in August and I got (paid for at least...) a new set of NGK BPR6EFS15 plugs, I am not sure if they were gapped correctly. As for faults, I have VesselView and no active faults show. I recently had another marine mechanic connect his computer (not sure if it was Diacom), it also said no current faults exist. As for past fault codes, I dug out an old print out and the following codes existed historically (before they were cleared at the service in August). Some of which can be explained by the lack of some sensors fitted (and a since replaced broken neutral switch on the port engine) but I am not sure about some of the others.......

PORT HISTORIC CODES PCM555:

0. TRIM CKT HI
1. STEER CKT LO
2. SEA TMP CKT HI
3. FUEL LVL CKT HI
4. PITOT CKT LO
5. NEUTRAL OVERSPE
6. BAT VOLTS LO
7. ECT TMP CKT HI
8. FUELPUMP RLY CK
9. BLK PSI CKT LO

STARBOARD HISTORIC CODES PCM555:

0. TRIM CKT HI
1. STEER CKT LO
2. SEA TMP CKT HI
3. FUEL LVL CKT HI
4. PITOT CKT LO
5. BLK PSI LO
6. BAT VOLTS LO
7. GUARDIAN
8. OVERSPEED
9. EST1 OPEN

I have the detailed senor results for the above codes at the time/hours they happened if required.

Thanks all.

While some of these are just circuits not connected, the others should throw a code. Since this is historical, they must have been fixed since they don't show on your display or the other scanner used

 

[havoc_squad]

There is no fuel that passes through the throttle body. If no codes are being set it's a fuel problem. The 555 is quite savvy about sensor and electrical issues

Define what you mean by fuel issue, if fuel trim is correct for the engine temp, no leaking injectors, correct fuel to match RPM, and lack of oxygen during idle is occuring, that would be air problem, not fuel.

Your fuel injectors can be putting out the right fuel amount but if it's starving for air, this issue can and does happen.

Yes, a motor can stall and run rough at idle because of lack of air and not throw a code.

A key sign would be the computer data of how much it is compensating with the IAC at idle when warmed up. There is a limit to how much the computer allows.

 

[scoflaw]

How could the TB be starving the engine of air. It's a wide open passageway with a butterfly. The 555 adjusts the fuel output according to the air from the throttlebody. The MAP and TPS dictate the injector pulse and width.

OP needs to get back to basics. Check out your fuel delivery system.

 

[Scott Danforth]

And diacom would show if the TPS or MAP is reading in the correct range, however a sensor going south or a bad connection such as a bad grond can drift the signal just enough for the motor to run like crap, yet not throw any codes

The fact that batt low was triggered on both ECMs would have me looking at diacom and a volt meter looking at all the connections

 

[tpenfield]

@ tpenfield - Hi Ted - Great to be aboard. Sorry, I didn't see your post before I replied to earlier to alldodge, Scott Danforth and achris. I agree, it does seem like a sensor could be causing the issue.
I am not physically up to doing my own maintenance and have relied on the marina to do so till now - but this one has them stuck.
Any chance you could share some info on what they should look for (on Diacom) so that I can point them in the right direction please? All I am being told by the marina is that there are no error codes......
Thank you

The marina mechanic(s) should be looking for more than just codes. They should be using the Diacom (computer) in real-time to see the engine parameters as it is exhibiting the problem.

It would be interesting to see once the engine is warmed to the point that it will not start at an idle setting, what the injector pulse values are (both start & run), and if the ECU knows if it is in starting mode while cranking. Also would be interesting to know what the injector pulses go to when you advance the throttle to WOT as your remedy to start when warm. I would imagine that is why we are seeing some over-speed codes in the history.

anyway, if the Marina mechanics are not going to really look at the data from the computer, then they are not going to help you. Maybe they can at least record the data when the engine is exhibiting the no start problem and the remedy that you have to start it. Then send you the recording file and you could then share it with some of the folks here on iBoats.

FWIW - I fixed the TPS on my starboard engine last year. The engine seemed to have a mind of its own and was revving all by itself up from idle speed, which made docking problematic. However, NO CODES . . . It was not until I had the Diacom software connected and I could see that the TPS value while the engine was running was fluctuating a few %, which made the ECU think I was bumping the throttle. I changed out the TPS, problem solved . . . all the while . . . NO CODES, because the TPS was giving a value that the ECU did not consider an error.

 

[mat5751]

Many thanks all.
Ted and Scott - thanks for your comments - I suspect you are right and that a I have a bad sensor and/or bad ground somewhere.
I'll go back to the mechanic and try again and if need be try to get some data recorded.

 

[achris]

First thing I would do is disconnect the interlink between the engine and see if the fault remains on both engines. Has a fuel rail pressure test been done? Specifically, warm the engines and check not only that the pressure is correct, but also that it doesn't bleed down after the engine is shut off. There's a spec for it in the manual...

 

[QBhoy]

Hi there.
Firstly, I think you have the wrong plugs for that mpi.
Can I also ask if this happens all the time or was it just once after sitting stopped for a while ?
just thinking about it, it’s unlikely to be anything like sensors or similar. Not on both engines. It’s more likely to be something common to each engine.
My initial thought were around low batteries, if your batteries are common and also used for leisure like the stereo and fridge etc etc. Having been thrown a curve ball with my own mpi regarding battery condition.
The other year, mine wouldn’t start but would turn over then kick a little when releasing the key. Turned out it was the battery. You’d think because it’s tuening over it isn’t the battery. Wrong. These mpi rely so much on juice from the battery they can refuse to start when the starter is taking that much away from the actual power needed to fire the engine. Real heavy dependency on power needed with these things to fire. That’s why when I let go of the key it would kickk. It was getting more power once the starter was released.
Long story short....battery needs to be tip top at all times with these and get the right plugs.
All the best

 

[achris]

...Firstly, I think you have the wrong plugs for that mpi....

Nice catch. Definitely the wrong plugs. Those plugs (BP6EFS) are for the carb vortec head engines, not the MPI. MPI uses AC 41-932, or NGK ITR4A15.

Chris.......