Need help diagnosing an engine problem

[Hyperlite]

Bayliner 2855 with Mercruiser 454 7.4 MPI Gen VI 1998.

When I crank it cold it will not start unless I give it a bit of gas. Last year it would start but rev up and down for a minute then stabilize and good to go. This season it stalls. However, if i keep it running for about a minute, then the revs stabilize and it runs fine. Even if I turn it off, 30 minutes later it will restart perfectly fine. No other issues. I checked the fuel pressure and it's 40psi. IAC valve have been replace 3 years ago, sparkplugs and all should be good as well.

I was thinking perhaps taking off the plenum and pulling the injectors for cleaning, plus putting new plenum and throttle body gaskets and inspecting the fuel pressure regulator for contaminants. May be some hoses for a potential vacuum leak. Not sure. Any help would be appreciated

[Nauti_Mike]

Sure sounds like the IAC but could be the coolant sensor used by the ECM.

[vr5200]

I don't know that powerplant very well........can it be hooked up to a scan tool?

I would look at the data before doing much tearing apart/replacing things.

Even if it costs some money to have someone come over with a scan tool, it may be good money spent.

MAP sensor and/or MAF may also be impacting things.

If any of the data is too far off....(ECT/IAT/MAP/MAF) the ECM will try to infer the best strategy......and it appears it figures things out once the engine gets a little bit warm

.......and yes, a 3 year old IAC could also be bad or "sick".

Have you verified all the small passages are pretty clean?

............and don't forget the IAC muffler (a small filter to quiet things down) needs to be clean in order for idle air to flow properly.

[Hyperlite]

Thank you, yeah i checked the throttle body, everything is super clean, i did clean the IAC muffler last year
Interestingly, i was thinking ECT sensor could be the culprit.
Honestly, i think cleaning the injectors is always a good idea, but agree, its probably best to try figure out before taking everything apart
i will see if i can get someone with the scan tool as well, but i thought Merc should beep if there are error codes that need to be scanned

[vr5200]

If an OEM ECT sensor is cheap and easy to replace.....that would be worth a try......it may be skewed just enough in certain temp ranges to cause an issue during a cold soak start.

[optimist_prime]

I have the same motor, and had the same problem. My MEFI-3 gateway told me the Engine Coolant Temp Sensor was reading too cold (high resistance) during the first few crank attempts. This was apparently confusing the ECM because it uses the reading for fuel injection mixture and it was shutting down. I found the sensor (not the one that feeds the dashboard gauge) on the front of the engine mounted on the thermostat. It has TWO wires unlike the gauge sensor. I pulled its connector, cleaned the contacts and re-seated it. The problem pretty much went away; now, I do get one stall when cranking, but no ECT alarm. I attribute the stall to a slow IAC valve, which I replaced last season. I may replace it again and see.

[cfoss]

I agree it sounds like a IAC. Or a vacuum leak.

One way to check for vacuum leak is to get the engine running, close the throttle, then stick a finger over the iac inlet, which is typically in the throttle body. If it stalls immediately then you don't have a vacuum leak, and the engine was running on the IAC, so at least it's open a bit. If you do have a MAF this won't work of course, just if you're MAP based.

Does it have a MAF or just a MAP?

My experience with MAFs is that they get dirty, and at low rates of air flow they send out poor signals (IE:Idle). When the flow rate gets higher they pick up enough airflow to flow to allow enough fuel to be injected that they run. IE when you crack the throttle the airflow increases, this is sensed and the injectors fire.

I've had good success of various models of cars with removing the MAF, and cleaning using carb/maf cleaner, then spraying with compressed air.

They went from rough idle/stalling to a perfect idle.

If it's just MAP, then it's usually tune up time. It could be an injector or sensor, but these are really rare events unless your fuel is fouled. I've checked the fuel spray pattern and quantity using a corona bottle (Clear) and a 9 volt battery before. It works but be VERY careful and take all precautions etc.

If you have been using ethanol based fuel, it could be water in fuel. I'd take a sample for analysis if this is the case.

Chay

[Hyperlite]

I have the same motor, and had the same problem. My MEFI-3 gateway told me the Engine Coolant Temp Sensor was reading too cold (high resistance) during the first few crank attempts. This was apparently confusing the ECM because it uses the reading for fuel injection mixture and it was shutting down. I found the sensor (not the one that feeds the dashboard gauge) on the front of the engine mounted on the thermostat. It has TWO wires unlike the gauge sensor. I pulled its connector, cleaned the contacts and re-seated it. The problem pretty much went away; now, I do get one stall when cranking, but no ECT alarm. I attribute the stall to a slow IAC valve, which I replaced last season. I may replace it again and see.

Yeah, that is my plan exactly, thank you, i will pull the sensor , test the resistance ( it has all the values in the Merc service manuals) then clean the wires and give it a try. This makes a lot of sense....

[Hyperlite]

I agree it sounds like a IAC. Or a vacuum leak.

One way to check for vacuum leak is to get the engine running, close the throttle, then stick a finger over the iac inlet, which is typically in the throttle body. If it stalls immediately then you don't have a vacuum leak, and the engine was running on the IAC, so at least it's open a bit. If you do have a MAF this won't work of course, just if you're MAP based.

Does it have a MAF or just a MAP?

My experience with MAFs is that they get dirty, and at low rates of air flow they send out poor signals (IE:Idle). When the flow rate gets higher they pick up enough airflow to flow to allow enough fuel to be injected that they run. IE when you crack the throttle the airflow increases, this is sensed and the injectors fire.

I've had good success of various models of cars with removing the MAF, and cleaning using carb/maf cleaner, then spraying with compressed air.

They went from rough idle/stalling to a perfect idle.

If it's just MAP, then it's usually tune up time. It could be an injector or sensor, but these are really rare events unless your fuel is fouled. I've checked the fuel spray pattern and quantity using a corona bottle (Clear) and a 9 volt battery before. It works but be VERY careful and take all precautions etc.

If you have been using ethanol based fuel, it could be water in fuel. I'd take a sample for analysis if this is the case.

Chay

Makes sense, but my problem is only reproduced during the cold start, once the engine warms up, even if i start it in 30 min, it will crank without a proble, no rpm fluctuations or anything, its only when its cold

[cfoss]

Gotcha. It could be ECT, it's not unheard of. If possible I'd check it at the ecm plug to make sure it's getting right to the ecm. I'd still pull and recheck the IAC. They are a high failure item on GM EFI. There is a different need for air when the engine is cold vs warm and if it's stuck that could be the issue. High idle when cold and all of that. Of course, that is controlled by the ecm based on the ECT signal, so there you go!

If you haven't done one in a while, I'd do a leakdown test just to verify the engine's integrity. Then it's pretty much gotta be a fuel/spark issue. And not likely spark from the symptoms, but it's not a terrible idea to check gaps/plug colour etc. while you are there.

Chay

[Hyperlite]

Pulled out the ECT sensor, contacts look nice a dry, hooked it up to the multimeter and resistance looks like it correlating with the temperature well, and with the numbers in the service manual, so looks like i need to keep digging

[vr5200]

Pulled out the ECT sensor, contacts look nice a dry, hooked it up to the multimeter and resistance looks like it correlating with the temperature well, and with the numbers in the service manual, so looks like i need to keep digging

Aw man!

I was really hoping that would be it.

[vr5200]

I have an idea for a little experiment just to see if this changes the cold soak behavior.

While cold soaked, use a hair dryer (any heat source that works for you) to warm up the ECT and the surrounding metal so that it is mildly warm to the touch.....we are trying to fool the engine a little .....it may not run quite as rich, for example....and see if the cold start behavior changes.

This is a free and easy thing to do.

Note- I do not know if the ECM uses a rationality check like some cars where it compares IAT to ECT upon a cold soak start....but it may throw a code during this experiment.

If this does not change behavior, I wonder if the IAT readings may be skewed.....that can also result in cold soak start issues.

As mentioned before, a scan tool with live data would help a lot but I understand you are trying to do all that you can personally do.

You could also try the same approach with the IAT.....but only mess with one sensor at a time

[optimist_prime]

Hyperlite, one of the best investments I've ever made for my engine is installing the MEFI-3 gateway made by Fox Marine. it's $300, and easy to install (hooks to the diagnostic connector). A phone app (free) is all that's needed to view about 30 live data engine parameters via Bluetooth. It also (if you have one) connects to a NMEA 2000 backbone to populate several engine parameters on my Garmin GPSMAP 741xs Chart Plotter, including fuel flow rate. I t even reads the total engine hours kept by the ECM.

[Hyperlite]

Hyperlite, one of the best investments I've ever made for my engine is installing the MEFI-3 gateway made by Fox Marine. it's $300, and easy to install (hooks to the diagnostic connector). A phone app (free) is all that's needed to view about 30 live data engine parameters via Bluetooth. It also (if you have one) connects to a NMEA 2000 backbone to populate several engine parameters on my Garmin GPSMAP 741xs Chart Plotter, including fuel flow rate. I t even reads the total engine hours kept by the ECM.

You know , I found something interesting, turns out you can scan Mefi 1-4 without the tool, there are multiple sources of people doing that, but here is the link

Make a Marine EFI code tool for less than $1.00 - Offshoreonly.com