1999 Mercruiser 4.3L EFI "ping" above 2,500 RPM

[rowndman]

Have a 1999 4.4 liter EFI Mercruiser I/O in a 20' deep-V - bought the engine & I/O new and personally installed in this boat. Engine runs beautifully up to around 2,500 - 3,200 rpm, then begins to "ping", and will progress to misfire, backfire, and loss of rpm if throttle is advanced further. Accompanying this symptom is erratic readings on the tachometer (a Quicksilver tach purchased new with the engine) that coincide with the backfire/misfire of the engine. Tach reading will swing from 3,500 rpm to 6,000 rpm (or more) when the engine is sputtering. Bring the throttle back to around 2,800 - 3,000 rpm, and the engine purrs. This symptom has evolved over about a 3-year period; at first, it only pinged at wide open throttle, then the pinging moved over time down the rpm range to where it is now. Along with this, the backfire/misfire developed at high rpm (over 4,200), then it also moved down the range to where it is now.

Have replaced spark plugs, wires, cap, rotor, fuel filter, run on alternative fuel sources, all with no improvement. FINALLY was able to get a local Mercruiser service tech (its unbelievable how hard it is to get anything done down here - south Louisiana - post Katrina) to put a scan tool on it & it showed code 33 - MAP Sensor. Ordered a new one & installed it yesterday; all symptoms still there. Checked base timing, it was dead on 8 degrees just like it came from the factory. Tech messed-around with a few other things, then threw his hands up in the air & said "I don't know". He left; I stayed & tried a few other things: disconnected the tach on the theory that maybe somehow the tach was affecting the ignition system - no change. Disconnected the fuel line, and connected a tank with brand-new fuel to the suction - no change. I then performed a test routine from Mercruiser's Service Manual on the Knock Sensor, and it seems to indicate the sensor is bad (theres infinite resistance between the sensor wiring terminal & ground; manual states that it should be between 3.3K ohms & 4.5K ohms). Now trying to locate a new Knock Sensor to install & try.

Has anyone else had this problem? Its getting pretty expensive & time consuming changing these high priced electronic doo-dads one by one, and not getting any relief.

 

[rodbolt]

Re: 1999 Mercruiser 4.3L EFI

Re: 1999 Mercruiser 4.3L EFI

can you tell me what the MAP sensor does if you can Ill tell you why you got that code. but most likly in explaining to me what it does it will dawn on you why you got it.
the old map was most likly fine.
then we will get onto your issue.

 

[rodbolt]

Re: 1999 Mercruiser 4.3L EFI

Re: 1999 Mercruiser 4.3L EFI

a hint
google: speed density EFI theory

 

[rowndman]

Re: 1999 Mercruiser 4.3L EFI

Re: 1999 Mercruiser 4.3L EFI

Rodbolt:

Thanx your reply. I understand what the MAP sensor does, and I absolutely agree with you that there was probably nothing wrong with the original MAP sensor; the code was probably entered because of the out of parameters pressure sensed in the intake during one, or many, of the backfire/missfire episodes. I wanted to keep the old one so I could check it with a vacuum source, but the Tech tossed it before I could capture it.

You have my attention now; please continue!

 

[rodbolt]

Re: 1999 Mercruiser 4.3L EFI

Re: 1999 Mercruiser 4.3L EFI

check yor fuel rail pressure, if its withen specs clean/replace the injector filters.
the high pressure pump and the pressure regulator outlet both have filters. is this an MPI motor or a TBI ?
another source of rattle and back fire is water sprying into the exhaust valves from a leaking riser.
and yes a backfire can set a code for a map sensor because the ECU isnt very smart.
but you should be getting the power reduction or knock retard on the scan tool.
did ya do the google search ?

 

[rowndman]

Re: 1999 Mercruiser 4.3L EFI

Re: 1999 Mercruiser 4.3L EFI

Been considering the fuel pressure angle as well, and did a few tests using outboard tanks directly connected to the fuel filter suction, with no change in symptoms. Realize that doesn't tell me anything about what's happening on the pressure side of the pump, but have been mostly pursuing the ignition side of the system because of the crazy behavior of the tachometer. Above 3,000 - 3,200 rpm, the tach needle will jump 500 - 1,000 rpm faster than actual engine speed, then the the engine will sputter, sometimes backfire, and misfire. Often the tach will return back to a true reading, then the symptoms will go away momentarily, then the whole process repeats itself - more frequently as throttle position is increased.

The induction system is TBI - General Motors issue.

This engine is equipped with Mercruiser one-piece exhaust manifolds, so there is no riser to leak. Of course, the possibility does exist that an internal hole in one of the manifolds could conceivably be squirting water on an exhaust valve, or valves, but I think this is extremely unlikely due to the the way things happen - as if a switch is turned at 3,000 rpm, but no symptoms what so ever below that speed.

As you say, the ECM isn't very smart, which kinda supports the finding I made yesterday on the knock sensor. The ECM will only register a code if it sees something it shouldn't. With a knock sensor that is sending no signal to retard advance, the ECM would think its doing a wonderful job of setting the timing because nothing is telling it anything any different, and a code wouldn't be recorded.

I have 3 years worth of Mercruiser Service Manuals (official Mercury Marine issue obtained from a former Dealer in the family - who didn't do service work on these engines) for these engines (1996, 1999, & 2000) that I'm becoming intimately familiar with. Also have 4 Mercruiser training videos from the same source that have been watched several times. I did the Google thing again (had done it about 6 months ago), but don't find much more there than I can get from these manuals. Do you have a specific site in mind?

 

[rodbolt]

Re: 1999 Mercruiser 4.3L EFI

Re: 1999 Mercruiser 4.3L EFI

the code for knock sensor wont store,its an active only code.
while running it under normal load track the ign system 12V with an analog meter. test the fuel rail pressure,it should be about 11-13 PSI,if memory serves me, on the TBI set up. while its stumbling at the problem RPM observe the spray patten from both injectors.
should be a nice large cone with no heavy droplets.
and them frigging one piece batwing POS's have trashed more motors than I care to discuss.
most of them are NLA now anyway.
do you know which ign system your engine has ?
the EST used a dry coil and the thunderbolt system used a wet coil, the wet coils liked to fail and arc from the primary to seconday towers and do crazy things.

 

[newport dave]

Re: 1999 Mercruiser 4.3L EFI

Re: 1999 Mercruiser 4.3L EFI

Is the engine serial number above or below 0L619000?

There were different igntion systems (Delco EST vs Tbolt V), different ECM generations (MEFI1 vs MEFI3) and different troubleshooting procedures used on '98 vs '99 models.

Also, if it's an Alpha model, check the shift interrupt switch.

Dave

 

[rowndman]

Re: 1999 Mercruiser 4.3L EFI

Re: 1999 Mercruiser 4.3L EFI

Rodbolt & Newport Dave:

Ignition system is EST, in Mercruiser-speak, its a MEFI 1 system, engine serial no. OL082695, Alpha One, Gen II outdrive. Shift interruptor is not the culprit.

Hope to get my hands on a new (or used-but-good) knock sensor this week and try some more this weekend. Will definitely check the ign system voltage and the spray pattern on the injectors - unless the problem is fixed, in which case I'll pop many a top.

Rodbolt, the rumor is true; the Navy does have ships! I've worked on them in a shipyard for 37 years.

 

[rodbolt]

Re: 1999 Mercruiser 4.3L EFI

Re: 1999 Mercruiser 4.3L EFI

yep
I actually got to ride some. sea duty was fun, mostly monotonous but fun.
the EST rarely give problems. the one piece manifolds did. carefully inspect the spark plugs for any signs or water burning on the electrodes.
why do you suspect a knock sensor?
you obviously did not do the google search.
I dont think MEFI 1 will go but about 8-10 degreess retard on a knock sensor detect signal and backfire would NOT be a symptom.
however as it detects knock 2 things occur, the pulse width on the injector gets longer,which will make an already lean problem worse, and spark gets retarded.
if your in saltwater look carefully for green spots on the electrodes,fresh water usually turns them a dark kida rusty brown in spots.
depends on the minerals in your water.
if you still have your old plugs they will tell the story on what is happening in the cylinder.
I dont see many MEFI 1 systems anymore so I cant remember all the fault and ECU responses.

 

[rowndman]

Re: 1999 Mercruiser 4.3L EFI

Re: 1999 Mercruiser 4.3L EFI

Suspect the knock sensor because it flunked a test routine from the service manual: resistance between the KS terminal & ground should be between 3.3k ohms & 4.5k ohms according to the manual; reading I get on this one is infinity. Checked the wire between the ks module & the ks - it's good.

My theory is that since the ECM is getting no info from the ks, it's advancing the timing as far as it can & thinks it is doing a fine job. Soooooo . . . if we install a ks that can talk to the ECM, it'll tell the ECM that it needs to retard a little bit to make things right.

Plugs removed were beautiful, with no sign of water impingment; I certainly did save them - they looked too good to trash.

Man, you really don't like those manifolds, do you? Must've had some real bad experiences with them.

 

[rodbolt]

Re: 1999 Mercruiser 4.3L EFI

Re: 1999 Mercruiser 4.3L EFI

those manifolds did not live long in my area, non really do.
if your reading was infinity it should set a code on the scan tool.
the ECU can recognize an open circuit.
the knock sensor has a 5VDC signal and the ECU is looking for an AC signal that indicates knock.
most the time we can test them by rapidly tapping the block,with engine running, with a brass bar by the sensor.
its real easy to check timing advance at WOT if you have a trusted helmsman that is not on your life insurance policy.
but most all the issues I see like you describe go back to lack of fuel in the intake runners or water in the exhaust.

 

[newport dave]

Re: 1999 Mercruiser 4.3L EFI

Re: 1999 Mercruiser 4.3L EFI

I agree with rodbolt, an open knock sensor should have set a code 43. But if the sensor tests open, replace it.

Just to clarify, on a MEFI1 system the knock sensor is
connected to the Knock Module, which provides a 8 to 10 volt signal to the ECM when there is NO knock present. When knock is present, the Knock Module removes the voltage and the ECM retards the timing. You should be able to check for the 8-10 volt signal and that will tell you if the ECM is being directed to retard the timing or not.

Also, check for plug wires touching the Knock Sensor or Ignition Control wires and check the connections where the magnetic pick-up connects to the Ignition Module.

Dave

 

[rodbolt]

Re: 1999 Mercruiser 4.3L EFI

Re: 1999 Mercruiser 4.3L EFI

also make sure the knock sensor wires are not laying on a plug wire. trust me that one will toss ya,
most the MEFI 1 systems here died some years back. average lifespan for a raw water cooled engine here is about 6 years.
some at 3, a few slightly longer.
almost all die from manifold/riser issues
in this area if its raw water cooled and over 4 years old I wont pull the valve covers.

 

[rowndman]

Re: 1999 Mercruiser 4.3L EFI

Re: 1999 Mercruiser 4.3L EFI

Thanx to both of y'all for your good advice. Ordered a new KS from a local dealer this morning; promised delivery by Thursday (out of Wisconsin). Since this thing read bad, gotta assume its hosed & needs replacement.

This engine runs in brackish water, mostly sweet, and is fresh water cooled, so I'm not anticipating any serious internal corrosion issues - like rodbolt experiences where the Wright Brothers made history. Pretty part of the country, at least it was when I visited there back in the early 70's. Stayed in a brand new Ramada Inn on the beach; desk clerk told my wife & I that we were the first customers to sleep in our room.

Fuel starvation has been lurking in the background in my head as well; plan to measure pump (rail) pressure, supply line vacuum (if any), etc. this weekend after the KS is swapped. I did find a plug wire in contact with the KS to KS Module lead and rerouted it to clear, but haven't had the opportunity to run it since then to see if that might have been the culprit.

Thanx again for your interest & advice.

 

[rowndman]

Re: 1999 Mercruiser 4.3L EFI

Re: 1999 Mercruiser 4.3L EFI

Update after long weekend:

Local dealer screwed-up on the parts order; new KS wasn't shipped so there's no verdict on whether it's the culprit. Engine runs the same with the existing KS connected or not connected, which lends a little support to it being the problem. Adapter I made to monitor the fuel rail pressure wouldn't fit, so wasn't able to check that, but there seems to be no problem on the suction side - zero vacuum on the supply line at WOT. Re-arranging plug wire routing made no change. Removed coil & cleaned/sanded its base and the mounting bracket to assure good ground (neither were dirty) - no change.

New KS now promised Thursday (again); will check it for correct resistance to ground, then install & see what happens. If that's no cure, friend of a friend has recommended someone in the area that has a reputation for being a diagnostician, not a parts-changer like the last crew. Expect I'll have to give him a try.

 

[rowndman]

Re: 1999 Mercruiser 4.3L EFI

Re: 1999 Mercruiser 4.3L EFI

Update to the continuing saga:
Got the new knock sensor, checked resistance on it before installing & it was lovely. Removed old knock sensor, installed new, torqued to 15 ft/lbs per the Service Manual (12 - 16 ft/lbs), connected the circuit. Just for hoots, checked the old one for resistance - it was fine! Checked that mutha 4 times before removing it. Anyhow, of course the new KS made no difference.
Tach still goes crazy right before the engine does, so started chasing everything I could find test procedures for that don't require a scan tool. IC module ground is good - removed, cleaned, & re-installed it anyway. Dropped the pin for the tach lead out of the connector, & ran the boat (theory being maybe there's something in the harness) - no change. Put pin back in & tach works same as ever.
While trying something, don't remember what, the engine started acting like it was about to run out of fuel. Had been doing a lot of running on the port tank (boat has 2) & knew it was getting low, so switched tanks, then checked level in port tank - still had 3" of fuel, plenty to run. On the full stbd tank, engine was still starving. Figured I got some symptoms now I can understand. Rigged-up a test gage to the TBI fuel inlet fully expecting to find low fuel pressure as Rodbolt has been hinting. WRONG! Solid, steady 32 psi at any speed. Book looks for 28 - 32, so I'm good. Opened the TBI & exposed the injectors - Eureka - both chambers full of gook that looked like black paint chips, and inlet filters to injectors pretty clogged. Cleaned all that mess up & went shopping for new gaskets / o-rings / injector filters. Surprise! - you can't buy that stuff, at least not around here (Rochester TBI unit). So, bought some gasket maker goop, and re-assembled everything in poor-boy fashion on old o-rings. Let 'er sit for an hour to give the goop time to cure, then put fuel pressure to it - no leaks. Cranked-up & headed out to test, figuring the problem is finally solved !!! - would'a bet $500 on it.
Engine ran the same as ever, except didn't have the stumble like it did on the last try.
Based on what the tach is doing, I'm just about convinced that my problem is either in the component that reads the RPM & tells it to the ECM, or in the component that sends the signal. System thinks that the engine is turning 5,000 rpm instead of 3000 & sends plenty signals to the coil to fire. Some of these signals are getting thru the cap to the plugs & the thing is firing while intake valves are open - making the backfire & ping.

Anyone had any experience with an IC Module or pole piece going south like this? Newport Dave , Rodbolt, anyone have theories?

 

[Don S]

Re: 1999 Mercruiser 4.3L EFI

Re: 1999 Mercruiser 4.3L EFI

If you have an Alpha outdrive, disconnect the shift interrupt switch and try it. You will have to hook it back up to get the drive out of gear, so look at where you are at before you try it.

 

[rodbolt]

Re: 1999 Mercruiser 4.3L EFI

Re: 1999 Mercruiser 4.3L EFI

on your fuel pressure test. was the test run at the problem RPM?
if the injectors are clogged its just like the high pressure pump being clogged or the regulator being stuck, the ECU does not monitor rail pressure or actual flow so it cannot compensate for over or under pressure.
if the ign interrupter has some internal corrosion Its possible its trying to shut the ECU off.
like DonS says dissconect and retest. just beaware that shifting from any in gear position to neutral will be almost impossible without switching the engine off or reccocting the shift switch.
the causes of pinging are few, to much compression for the fuel being used, over advanced ign timing for the RPM used, Overloading the motor for the RPM used and a lean airfuel mix in the cylinders. , about the only thing you really cannot check other than by observation and experience is water intrusion from a bad manifold.
some of the early systems used the coil primary - terminal for a tach send, if you suspect the tach circuit dissconect the gray sender wire from the coil ,if used.
also watch the voltage on the purple wire at the problem rpm.

 

[rowndman]

Re: 1999 Mercruiser 4.3L EFI

Re: 1999 Mercruiser 4.3L EFI

Don S. & Rodbolt:
Thanx your replies. It is an Alpha; the interruptor switch is easy enough to check & will do so at next opportunity.
Fuel pressure was rock-solid at all throttle positions- didn't change over 1/2 psi anywhere; I'm pretty convinced that fuel is not the problem.
During my trials this weekend, I actually removed the gray wire from the 5-wire connector between the EFI harness & the starting/charging harness, plugged it back together & test ran it with no change in symptoms, other than of course the tach didn't work. That tells me that there's no problem from that connector all the way to the tachometer, whatever is giving the false readings to the tach is happening between the distributor & the connector. The purple wire is the signal to the voltmeter on the dash, among other things, and the voltmeter remains steady at all speeds, so I kinda doubt anything is going on there either, but I'm sure gonna look into it further.
We 'preciate your bidness; Keep those cards n letters comin'.