MEFI4 troubleshooting hard start

[Turbo_Bimmer]

Hello all.
Anyone with experience with this ECM?

My boat engine is a Volvo, but basically it is a GM 5.7L MPI Vortec from 2003.

Intermitently, it is hard to start. Long cranking time and then it starts. Other times, it starts right away.
Opening the throttle seems to help reduce the starting time.
Also, the problem seems to be related to temperature or atmospheric pressure, but can't pinpoint for now what. Early in spring and late in fall, the problem is not there as much.

I replaced over the last years many parts.
ignition coil, distributor cap and rotor, spark plug wires, spark plugs
PCV valve, IAC, removed injector to check spray pattern and it was good and cleaned them anyways.
cleaned all grounds lugs and 12V lugs, checked fuel pressure low and high (2 pumps) and they are ok.
I cleaned the throttle body, replaced the ignition relay, tested the coolant temperature sensor in boiling water and the resistance is ok according to my manual.
The MAP had some oil on it, so I cleaned it but it didn't help.

When the problem happens, once it is started, if I shut it down and restart, it will start right away, so that rules out the CPS or any corrosion in plugs...

I start to be out of ideas.

I made 2 videos filming my gateway showing some data on my phone. One good start, one bad start.
The only thing I can see strange, is the bad start, the engine seems to crank fast, but the RPMs seen by the ECU are low, like 150, but that reading is coming from the crank sensor?
The boat, once started, runs like a champ, full power, nice idle etc.

Here is a video showing a good start:

https://youtube.com/shorts/uFWD0gaGGGo

then, a video showing the hard start:

https://youtube.com/shorts/WpLmQOmbjP4

Thanks.

[1project2many]

Hello,

I would generally follow a series of tests to diagnose this. Realizing that it's an intermittent issue on a boat, I would still attempt to find a way to confirm the plugs are sparking and that the fuel rail is at the required pressure.

If you'd like to look at potential causes, there are a few hunches that may or may not be right. First, if one or more of your injectors is leaking slightly after the engine is shut off this can saturate the intake with fuel. Cranking the engine will cause a hard start as the injectors deliver more fuel. Opening the throttle can allow more air into the intake to draw away some of the excess fuel. Opening the throttle to 100% typically triggers "clear flood" programming which causes the computer to decrease the amount of fuel delivered. I am unable to see the injector on-time in your hard start video but I would suggest watching this value to see if it decreases substantially shortly before the engine starts.

Another potential cause could be lack of fuel at the rail. A fuel pressure gauge is invaluable during hard-start diagnosis and can quickly tell you if your pump is working as expected.

Forcing the engine not to start by disabling the coil would allow you to watch the scantool displayed rpm during cranking. This could help you determine if the low value seen during the hard start is a normal value. We've had crank sensors cause intermittent issues in the past. It's rare but it can happen.

Finally, fuel is a mixture of chemicals and the mixture changes through a season. Additionally, the ingredients added can change through the years. In some cases programming that worked 20 years ago is not optimum today. This can be tougher to figure out. For some folks the answer to this problem is to find a routine that allows the engine to start as quickly as possible and just follow it when there's a problem.

[tayto]

i would hook a timing light up and see if it is flashing when the hard cranking occurs and/or see if the tach is moving during crank. this can help narrow the problem down to ignition or fuel delivery. I would also check fuel pressure at WOT under load. Have seen pumps make spec pressure at idle no load, but once under load they drop below spec...

not familiar with MEFI4, but is there not some sort of base timing that can be checked/set?

[Turbo_Bimmer]

I will reply in blue


Hello,

I would generally follow a series of tests to diagnose this. Realizing that it's an intermittent issue on a boat, I would still attempt to find a way to confirm the plugs are sparking and that the fuel rail is at the required pressure.
I attached a fuel pressure gauge to the engine. Pressure is good. The only thing I see not normal, is that the pressure in the line drops from 50 to 10 in 9 minutes when the engine is turned off. But that doesn't seem to affect subsequent starts. Look at the video below. For spark, I bought a spark tester and increased the gap and the spark seemed good to me

If you'd like to look at potential causes, there are a few hunches that may or may not be right. First, if one or more of your injectors is leaking slightly after the engine is shut off this can saturate the intake with fuel. Cranking the engine will cause a hard start as the injectors deliver more fuel. Opening the throttle can allow more air into the intake to draw away some of the excess fuel. Opening the throttle to 100% typically triggers "clear flood" programming which causes the computer to decrease the amount of fuel delivered. I am unable to see the injector on-time in your hard start video but I would suggest watching this value to see if it decreases substantially shortly before the engine starts.
I will check the injector time ON during cranking. When I had the injectors removed, I connected them one by one to a fixture I made with a steel pipe, filled the pipe with fuel, and applied 45 psi or air, and waited 1 minute each and none of them were leaking.

Another potential cause could be lack of fuel at the rail. A fuel pressure gauge is invaluable during hard-start diagnosis and can quickly tell you if your pump is working as expected.
see above

Forcing the engine not to start by disabling the coil would allow you to watch the scantool displayed rpm during cranking. This could help you determine if the low value seen during the hard start is a normal value. We've had crank sensors cause intermittent issues in the past. It's rare but it can happen.
ok I will do that (disable the coil). One thing I will try to do, is to borrow a tachmeter from my work and see if the engine is cranking 125 RPM like reported by the gateway. If yes, it is too slow to start, NO? I think a ECU goes from cranking mode to running mode around 400 RPM, is it right? Maybe my problem is there; bad CKP or engine not cranking fast enough. But from the sound of it, it seems to turn fast enough...

Finally, fuel is a mixture of chemicals and the mixture changes through a season. Additionally, the ingredients added can change through the years. In some cases programming that worked 20 years ago is not optimum today. This can be tougher to figure out. For some folks the answer to this problem is to find a routine that allows the engine to start as quickly as possible and just follow it when there's a problem.
Yes, harder to figure. But other people with the same engine on boat forums don't have that problem, so I suppose the programming is not too bad???

Here is a video of a cold hard start, followed by 2 cold easy starts. The fuel pressure gauge shows that there is enough pressure, even if it drops when the engine is stopped.

https://youtu.be/pIFu_HpI8is

[Turbo_Bimmer]

I will reply in blue


i would hook a timing light up and see if it is flashing when the hard cranking occurs and/or see if the tach is moving during crank. this can help narrow the problem down to ignition or fuel delivery.
Good idea for the timing light. I will have to find one since my plastic one got destroyed falling on the concrete floor! I would also check fuel pressure at WOT under load. Have seen pumps make spec pressure at idle no load, but once under load they drop below spec...Yes I understand, but while cranking, the demand of fuel is relatively low? even a worn, low flow pump should be sufficient? But it's not a concern, since I can run hours at 3800 RPM fully loaded and not a miss.

not familiar with MEFI4, but is there not some sort of base timing that can be checked/set?
not really. Indexing the distributor cap is about the only thing that can be done to these engines... as far as I know.....

[1project2many]

One thing I will try to do, is to borrow a tachmeter from my work and see if the engine is cranking 125 RPM like reported by the gateway. If yes, it is too slow to start, NO? I think a ECU goes from cranking mode to running mode around 400 RPM, is it right?

I do not believe 125 rpm is too slow for cranking or starting. I have "pop" started engines that were turning at much slower speeds.

In the most recent video it sounds as if there is a vacuum leak or as if the IAC is fully opened. It may be the sound of the air cleaner doing it but it just doesn't sound as I'd expect. Also, it doesn't start well even on a "good" start. Nor does it idle well compared to what I'm used to. This may be normal for this craft but it's something I'd note.

Despite this engine having a new IAC, startup in the latest video does sound as if IAC is not working properly. You could use a pair of LEDs to check whether or not the ecm is commanding IAC operation properly. Some tests include watching for IAC activity after engine shutdown, and watching for continuous IAC activity during idle which would indicate the ecm attempting to change the engine's idle speed. LED's should be connected across two wires of the same color and can be connected on the backside of the IAC connector to allow the IAC to remain in the circuit. Connecting the old IAC but leaving it uninstalled so you can view the pintle, then running the engine for a short time before shutting it off would also allow you to see if the ECM is successful when it attempts to operate IAC.

Also, can you open the throttle while the engine is off before starting to check for the smell of gasoline? Have you ever noticed strong fuel smell immediately after startup which could indicate excess fuel during cranking? Those tests may be better indicators of a problem with the injector showing up after the engine is shut off. I'm not sure I'd trust that the test you designed as the best way to capture this intermittent issue.


Yes, harder to figure. But other people with the same engine on boat forums don't have that problem, so I suppose the programming is not too bad???

I would not say that. I would put "incorrect calibration" on a list of potential causes and wait to see what continued testing reveals. Is it possible you have a calibration that should have been updated due to an issue? Is it possible a couple of errant values in your calibration are causing an issue under a specific set of circumstances that are not happening to other folks?

Is there any history with this engine that may be relevant? Have you owned it since new? If you wanted the engine to start hard for a test, what would you do the try and make that happen?

[tayto]

yea something doesn't sound right... check to make sure the injectors are plugged into the correct cylinder. not sure if this is a bank/batch fire but I think sequential system so this will matter. have you adjusted the valves? i know they are hydrualic but pretty easy to do it while the engine is running. Before this I think it's time to do a basic compression test and report back.

[In-Tech]

Hiya, please check and reply with the actual service number of the Mefi4. I can't remember the last 3 suffix of the Mefi 4, but 4A ends in 479 and 4B ends in 052. Here are some screen shots of a Volvo Penta Vortec 350 4A file. I noticed on the hard start that the IAT is higher than the ECT and this is not normal for a boat. I also noticed your cranking pulse width is quite a bit lower due to the IAT or some other factor. ECT and IAT sensors are known to 'glitch'. If one or the other 'glitches' during crank or run, you will have intermittent problems in run or crank, and they are HARD to find unless you actually happen to catch it on the scanner.

One suggestion I will make here and also for other projects is to always buy Delphi sensors. We all used to call Ac-Delco the go-to, somewhere along the line they had a fight with Delphi and now it is a 50-50 or worse chance you will get China crap in an Ac-Delco box. I have switched to only true Delphi stuff anymore.

One other thing to note is if this is a spider-web vortec 350, those poppet valve injectors will NOT pulse if the fuel pressure drops below 50psi. You mentioned having some low fuel pressure issues, so, don't forget about a pump/regulator problem.

Let us know :)

[tayto]

the vortec marine intake doesn't use the poppet style injectors AFAIK.

[Turbo_Bimmer]

reply below:


Hiya, please check and reply with the actual service number of the Mefi4. I can't remember the last 3 suffix of the Mefi 4, but 4A ends in 479 and 4B ends in 052. Here are some screen shots of a Volvo Penta Vortec 350 4A file. I noticed on the hard start that the IAT is higher than the ECT and this is not normal for a boat. I also noticed your cranking pulse width is quite a bit lower due to the IAT or some other factor. ECT and IAT sensors are known to 'glitch'. If one or the other 'glitches' during crank or run, you will have intermittent problems in run or crank, and they are HARD to find unless you actually happen to catch it on the scanner.
my ECU is MEFI 4B Number on ECU: 12584052.
I am not sure there is a IAT sensor. On the throttle body, there is a flame arrester only, no air ducting, no sensor. I only see 2 temperature sensors, one is for the ECU at the thermostat housing, the other is on the coolant (water) drain valve on the side of the engine and is used for the dash gauge. May be the IAT number is a calculated value for a boat engine??

What I am thinking my problem might be from my recent readings, is it could be that the crank sensor is defective or weak. That the pulse doesn't have enough amplitude and the ECU misses some pulses... or the starter doesn't turn fast enough to create a good amplitude pulse from the crank sensor. I don't think that engine has a opto sensor, I think it is a reluctance / pickup coil (1995 - 2005 era)... it is hard to see but I think the difference is 3 vs 4 wires if I remember well. Could the problem come from there?

One suggestion I will make here and also for other projects is to always buy Delphi sensors. We all used to call Ac-Delco the go-to, somewhere along the line they had a fight with Delphi and now it is a 50-50 or worse chance you will get China crap in an Ac-Delco box. I have switched to only true Delphi stuff anymore
does a GM dealer sell Delphi or Delco? I'm afraid to buy sensors at places other than my Volvo dealer since there is many calibrations depending of years of engine...

One other thing to note is if this is a spider-web vortec 350, those poppet valve injectors will NOT pulse if the fuel pressure drops below 50psi. You mentioned having some low fuel pressure issues, so, don't forget about a pump/regulator problem.

Let us know :)

[Turbo_Bimmer]

yea something doesn't sound right... check to make sure the injectors are plugged into the correct cylinder. not sure if this is a bank/batch fire but I think sequential system so this will matter. have you adjusted the valves? i know they are hydrualic but pretty easy to do it while the engine is running. Before this I think it's time to do a basic compression test and report back.

it is a batch system. injectors are in the good place because the fuel rail is rigid and the wire loom doesn't have enough play to connect to the next injector. I didn't adjust the valves since there is no ticking sound. I didn't do a compression test because the engine pulls strong at cruising speed at 4000RPM, so I think the compression must be good...

[Turbo_Bimmer]

I do not believe 125 rpm is too slow for cranking or starting. I have "pop" started engines that were turning at much slower speeds.
isn't there a minimum RPM to create a good amplitude signal from the crank sensor? this is my new troubleshooting idea now LOL

In the most recent video it sounds as if there is a vacuum leak or as if the IAC is fully opened. It may be the sound of the air cleaner doing it but it just doesn't sound as I'd expect. Also, it doesn't start well even on a "good" start. Nor does it idle well compared to what I'm used to. This may be normal for this craft but it's something I'd note.

Despite this engine having a new IAC, startup in the latest video does sound as if IAC is not working properly. You could use a pair of LEDs to check whether or not the ecm is commanding IAC operation properly. Some tests include watching for IAC activity after engine shutdown, and watching for continuous IAC activity during idle which would indicate the ecm attempting to change the engine's idle speed. LED's should be connected across two wires of the same color and can be connected on the backside of the IAC connector to allow the IAC to remain in the circuit. Connecting the old IAC but leaving it uninstalled so you can view the pintle, then running the engine for a short time before shutting it off would also allow you to see if the ECM is successful when it attempts to operate IAC. I think I sprayed some carb cleaner in the past around the throttle body and around the intake but didn't see any change while idling...

Also, can you open the throttle while the engine is off before starting to check for the smell of gasoline? Have you ever noticed strong fuel smell immediately after startup which could indicate excess fuel during cranking? Those tests may be better indicators of a problem with the injector showing up after the engine is shut off. I'm not sure I'd trust that the test you designed as the best way to capture this intermittent issue.
yes, there is a gasoline smell after a bad start[/COLOR][/COLOR]



I would not say that. I would put "incorrect calibration" on a list of potential causes and wait to see what continued testing reveals. Is it possible you have a calibration that should have been updated due to an issue? Is it possible a couple of errant values in your calibration are causing an issue under a specific set of circumstances that are not happening to other folks?

Is there any history with this engine that may be relevant? Have you owned it since new? If you wanted the engine to start hard for a test, what would you do the try and make that happen?
I bought the boat used in 2018 and it was doing that when I bought it. Previous owner just said, open the throttle while cranking when the problem occurs. Yes it helps. I put some time and money finding this problem once in a while, since the boat is still functionnal. It is just embarassing having a shinny boat cranking for 10 seconds and having a bad fuel smell when people are watching LOL

[Fast355]

So I looked at a datalog I took of my Express van at startup I took working on the startup airflow in the tuning. It cranks at 115 rpm. It has the same L31 crank sensor, the same coil, same module and HVS distributor as your marine engine. The crank sensor could trigger at 1 rpm. The PMGR vortec starters are gear reduction units not direct drive units, so they crank slower but with more torque. No problem getting my 11:1 383 cranked over even in 20°F weather. The low idle right at startup is what I was datalogging and adjusting to fix.

https://youtu.be/2Ydppg2TtQU

I would check your fuel pressure regulator for leakage. Even though my Marine intake had a returnless system that had the return capped off on the rail, it still had a diaphram and vacuum line connected to the intake manifold. When I pressurized the rail and tested the injectors for leakage, fuel came trickling out of the regulators vacuum port. Most of the MEFI4 marine engines have a fuel rail pressure sensor. I would look at it and see what the pressure is doing. A cold start with a leaking regulator will not have fuel in the manifold to flood the engine out, while a restart will.

[Turbo_Bimmer]

So I looked at a datalog I took of my Express van at startup I took working on the startup airflow in the tuning. It cranks at 115 rpm. It has the same L31 crank sensor, the same coil, same module and HVS distributor as your marine engine. The crank sensor could trigger at 1 rpm. The PMGR vortec starters are gear reduction units not direct drive units, so they crank slower but with more torque. No problem getting my 11:1 383 cranked over even in 20°F weather. The low idle right at startup is what I was datalogging and adjusting to fix.

https://youtu.be/2Ydppg2TtQU

I would check your fuel pressure regulator for leakage. Even though my Marine intake had a returnless system that had the return capped off on the rail, it still had a diaphram and vacuum line connected to the intake manifold. When I pressurized the rail and tested the injectors for leakage, fuel came trickling out of the regulators vacuum port. Most of the MEFI4 marine engines have a fuel rail pressure sensor. I would look at it and see what the pressure is doing. A cold start with a leaking regulator will not have fuel in the manifold to flood the engine out, while a restart will.

Wow! your engine starts really fast!
Now I remember, at the end of the season here in Canada, with colder temperature, the engine starts quite well. I did check the FPR with a hand held vacuum pump, and it was holding it well. But I will check like you suggest by pressurizing the fuel rail and look if the vacuum hose is leaking gas.
I also opened the FPR to look for paint particles (Volvo painted the inside of their fuel cells in those years), and the screen was nice and clean.

I didn't see any sensors on my fuel rail, unless it is somewhere else.

[Turbo_Bimmer]

does the MAP play a role during cranking, or it is just after the engine is running? From my gateway data, the reading seems OK if I compare to the weather forecast...

[Fast355]

does the MAP play a role during cranking, or it is just after the engine is running? From my gateway data, the reading seems OK if I compare to the weather forecast...
Maybe Carl can shed some light because I have not played with a MEFI controller in a decade. My 0411 has a Cranking VE table that has RPM and MAP as a reference. A bad MAP on alot of GM stuff can cause hard starts but usually shows up as more of a rough or rich running immediately after startup.

I bought a genuine Bosch T-Map for my marine intake for about $30. Mercruiser and Volvo Penta mark them up about 500%. I will see if I can find the Bosch part number.

[Fast355]

My guess is it is the same one I had. 0-261-230-075 but you should be able to confirm the Bosch number and search it up for a replacement if it is not.

https://www.ebay.com/itm/324135895105

[Turbo_Bimmer]

My guess is it is the same one I had. 0-261-230-075 but you should be able to confirm the Bosch number and search it up for a replacement if it is not.

https://www.ebay.com/itm/324135895105

I looked today, and it is painted red like the rest of the engine, so no number to refer to...

I took the time to unbolt the crank sensor. There was some clean oil on it. I'm not sure if it matters though. Cam the oil sneak inside and affect it?

18130

[Fast355]

I looked today, and it is painted red like the rest of the engine, so no number to refer to...

I took the time to unbolt the crank sensor. There was some clean oil on it. I'm not sure if it matters though. Cam the oil sneak inside and affect it?

18130

Oil can leak into the crank sensor if it is damaged or cracked, but the oil on the metal nose is nothing to worry about. The reluctor wheel sits right in front of the timing gear. Lots of oil in there when it is running. Usually if the crank sensor is getting oil into it, it will drip out of the wiring connection end.q

[In-Tech]

Good call Fast. The cal does show it to be the Bosch MAP sensor that has built it IAT. Personally I would wipe the metal fuzz off the crank sensor, get a new map/iat and also coolant temp. Have a look at the cam sensor distributor and clean the rust out, a delphi cam sensor if need be. :thumbsup:

[Turbo_Bimmer]

Good call Fast. The cal does show it to be the Bosch MAP sensor that has built it IAT. Personally I would wipe the metal fuzz off the crank sensor, get a new map/iat and also coolant temp. Have a look at the cam sensor distributor and clean the rust out, a delphi cam sensor if need be. :thumbsup:

If the MAP/IAT sensor is bad, wouldn't it make the same hard start conditon every time I crank the engine?
Because if I have a hard start, then I turn the engine off, I can restart it in a quarter turn, for 10 times if I want.
It is hard to start after 20 minutes of resting or more, exept when it is a colder temperature like at the end of the summer.
Like if there was extra gas inside the intake manifold.
But If the boat is in my driveway for a week, I suppose the gas would evaporate if it's the problem, but the engine is still hard to start, for the first start.

I don't think I have a cam sensor. You mean the trigger module inside the distributor? with the 8 points disk?

[Turbo_Bimmer]

Oil can leak into the crank sensor if it is damaged or cracked, but the oil on the metal nose is nothing to worry about. The reluctor wheel sits right in front of the timing gear. Lots of oil in there when it is running. Usually if the crank sensor is getting oil into it, it will drip out of the wiring connection end.q

No oil to be seen on the connector. Is the crank sensor inside the timing cover (with oil)?
I tought it was outside of the timing cover, so, if inside, it is normal to have oil on it, my bad.