Mercruiser 5.7L stumbles backfires after 40 mins

[...bill]

This is followup to a topic I started about a year ago:
You can read the old post here
http://forums.iboats.com/forum/engi...568-mercruiser-twin-inboards-backfire-stumble

A quick overview...
Twin Mercruiser 5.7l's of 1997 vintage that have adjacent serial numbers, 0Kxxxxxx. 4bbl Weber carbs with thunderbolt V knock sensor ignition with full fresh water cooling system. There's about 350 hours on these engines and have been maintained with Mercruiser parts.
After 40 minutes up on plane around 3200 rpm, port engine will pop/backfire once out of the blue. Usually a single event, but once this begins, it will become more frequent in about a minute or two to the point of rough running forcing us to run at 1200-1500 rpm to keep the stumbling under control. Shutting the engine down for 10 minutes allows it to run perfectly again for another 20-ish minutes, then same cycle. Starboard engine will begin to "run crappy" too at about the same time, however, not so much backfire.
This occurs enough to not trust the boat for any kind of extended trip.

Per previous post, I replaced pretty much everything in the ignition system. Wisdom from the forum is it had to be fuel related, which made sense.
Real quick:
Good compression all cylinders.
New ignition wires, plugs, cap, rotor, sensor thing in distributor.
Not related to this problem, also new risers, exhaust manifolds, fresh water heat exchangers.
Rebuilt the carbs. Found junk in carbs, so had tanks pumped and cleaned. New fuel hoses, new fuel filters, new gas. This had to be it.

And, the problem persists. Completely baffled, and dejected.

The ignition coils are the only things not replaced. How do ignition coils fail? Can they fail only when hot?
Other ideas welcome.

 

[alldodge]

In previous thread (before the tank cleaning) the filter would get clogged with crud and some water. Some how the new filter allowed crud to make it thru the filter and into the carb (don't know how that could happen). So the tank has been cleaned, and what do the filters look like now?

 

[...bill]

Last year, after the filters were cut open, we installed new 9 micron Racor filter assemblies with the metal drainable water separator.
This year, before the tanks were cleaned and with the Racor filters, I installed a vacuum gauge at the output of the filter (input of pump) to see if filters clogged while this goofy behavior occurred. Never recorded any vacuum at all and physically looked at the gauge while the backfires/stumbles happened. I know the filters weren't clogged, plus there was less than 3 hours on them. Even in the past, the filters never got clogged to the point of limiting fuel flow.
After tank cleaning, before startup, drained the Racor water separator housings and collected the fuel. Some fine particulates, but no water. Then, drained the new fuel hoses, replaced the Racor elements, removed carbs and emptied the fuel bowls (was still clean after recent rebuild). Cut a half inch off the bottom of the pick-up tubes. New fuel.
Same problem.

Another fact making me think it's the coil(s):
Last year, I had a timing light on one engine to see what happens to timing when this bad behavior occurs. Timing light would blank out synchronized to the engine stumbles. Folks here convinced me I had a lean sneeze condition and the knock sensor / TB V was changing timing/ignition to save the engine.
I'm now doubting this. Seems to me, nothing but a bad coil or bad TB V module would kill the spark. Oh, did I mention we replaced the TB V and knock modules....

As these engines are truly twins, a failure on one typically occurs on the other soon after. This happened with the water temp sensors, the oil pressure sensors. One failed, month later, other failed same way. Intellectually, it's a bit hard to swallow both coils are bad. But, it is possible. Swapped the coils last year port/starboard and no behavior change. But both engines have bad behavior.


Thots/comments welcome.
?bill

 

[alldodge]

Is the coil hot?
What is the voltage at the coil + side when this happens?
Check if the battery terminals and cables are hot

The replaced module in the distributor, is it one of the new ones with the additional ground wire?

 

[...bill]

Coil is mounted to engine, so everything is hot. Can't tell if coil is hotter than the engine. This happens only when everything gets hot.
Coil voltage is battery voltage. Checked with voltmeter while underway. Usually around 14.1V or thereabouts.
Yes, distributor module has that extra ground wire. Attached ground wire to bell housing bolt. Recently read that wire should go to engine ground stud. Can try moving that, but seems like a longshot.

 

[alldodge]

Coil should not be so hot you cannot touch it. Was wondering if there is a voltage issue. If the alternator starts putting out to high a voltage the coil will start getting hot, hence the reason for the questions).

This is also why I was wondering if the battery terminal or cables were warm or hot.

The ground is fine where it is, just needs to be a clean ground. Assuming you have bravo drives so not thinking about shift interrupter. If your loosing spark then it can only be one a few items with a TBV.

Have you tried swapping the module? Maybe even hot wire the engine so you can take the key and wiring harness out of the equation

 

[...bill]

Power and grounds at batteries are good.
I'll check voltages again while underway, but last check all was good. I believe it's OK to see as high as about 14.8V while charging.
Distributor module ground is certainly clean and secure. Was diligent about that connection.
This is an inboard configuration with Velvet Drives. So, indeed, there is no ignition interrupter. That (white) wire on the TB-V module(s) is capped off.
We swapped TB-V modules port to starboard. Same behavior. Even swapped in brand new TB-V module and a new knock control module. No difference. Re-crimped all connections to the TB-V harness as due diligence.

Did all this, hence, suspect coils. Mostly, because there's nothing left that hasn't been replaced.
I'd much rather figure out exactly what's wrong than to shot-gun the coils. Due to the elusive nature of this problem (goes away after a few minutes shut down), I'll swap out the coils as the fastest path to a possible solution.

Haven't hot wired the ignition switch. I'll run a bypass with a switch and will try that. As doing a sea trial is a huge hassle, I'll rig the bypass (on/off) with the coil replacement and try that.

Any experience with Sierra vs. Mercruiser coils? I can get Sierra parts quicker.

Thanks,
...bill

 

[alldodge]

The hot wire I was thinking about is done at the motor, this way you get rid of everything to the key and back. Take a 12V off the electric choke and connect the + side of coil.

So far as parts, many folks have had good luck with Sierra, but you can get a coil from the auto parts store, same thing.

 

[...bill]

Yes, I'm on the same page. Grab live 12V at source (choke or start solenoid) and jump to + side of coil. I intend to put a simple switch in the jumper so I can turn this hack off and on at will. Ordered Sierra coils. Sea trial arranged for 3 days from now. I'll report back. Welcome any other ideas.
If this doesn't work, may inquire about nautical exorcisms.

Thanks,
...bill

 

[Fun Times]

Does your boat have a engine synchronizer gauge for the two engines? If so the gauge may be causing an issue itself or providing a shorted electrical path to both engines should one of your tachometers be shorting out internally interrupting the coil(s).

Do your tachometers seem to start fluctuating more like a electrical issue (happens fast) vs a fuel related issue (slightly slower responses)?

Bypassing the helm electrical harness at the engine would be a good idea to help know what you should be concentrating on.

For the fuel system...Does your fuel tank have one fuel outlet fitting for both engines or each engine has it's own fitting?

The fuel tank vent system...Are there one or two vent ports? Maybe the vent system is partially clogged or compromised slowly blocking off that big fuel tank. What size fuel tank you have in there?.. With the fuel tank filler-cap off, Would you be able to try blowing and sucking air in the vent ports on your boat listening for air movement with little restriction?

 

[...bill]

This is an old restored Chris Craft wooden boat.
No synchronizers, and the tachs are mechanical! They're 10% off, but they work fine.
Bottom line, nothing connected to the "gray" tach wires in the helm. I indeed disconnected the gray wires anyhow at the coil early in the troubleshooting.

The tach does begin to fluctuate a little bit before the backfire. We can see it, but we can't feel it until the backfire. Very subtle.

Bypassing the entire Cannon connector may be beneficial. On this boat, not much in the helm: the key switch, the ignition fuses, Mercruiser warning buzzer, and vintage water temp and ammeter gauges. Ammeter is disconnected.
I'm going to try bypassing the ignition circuit as stated above if the coil replacement doesn't work.
I've confirmed many times that the coil + voltage is rock solid when the misfires happen.

There are 2 90 gal. fuel tanks. Each engine has it's own fuel source. There is a way to crossover fuel flow, but we keep that function switched off. All new fuel valves and hoses by the way, so no air leaks.

There is one vent per tank. However, Ye Olde Chis Craft fuel fill cap is also vented. We are sure the vents are free and clear as they're 1/2" copper pipes and easy to poke. As mentioned, I hooked up a vacuum gauge at the fuel filter exit (at pump inlet). I would have seen vacuum rise if the filter clogged or if the tank vent clogged.

Thanks for the ideas,
...bill

 

[Fun Times]

Interesting both engines tend to want to start to run bad near the same time-frame yet the engines seem apart from each other component wise and yet you'd think it's very rare but as you mentioned above it's not all that rare for component to fail around the same time.

Ignition coils do sometimes act up in different ways and by that I mean not just work or don't (< Spark)....So maybe plus maybe why you see something on the tach gauge.

Some of the TB-5 ignition control modules of that era did have a module "case" ground wire so double check both of yours if they are there or look to have been there at one time.

Though this next theory is doubtful as it seem you've been running with the engine hatch cover open for testing, Does the engine compartment seem to have adequate ventilation and spacing as to where it's not getting to hot under there or eventually running out of air space possibly from being over sealed up... AKA starving for air....I haven't read the other topic so was wondering how long these Merc 5.7 engines have been in there?

Any major changes done to the boat since they were installed?

The engine room temp shouldn't be getting past 176 degrees, https://www.perfprotech.com/blog/art...ser-vapor-lock

Maybe the engines were set up to run a different elevation then what you're running in now? Weber 4 bbl Carburetor

Flame arrestors aren't dirty or different style than OEM.

Obviously I'm just throwing some rarer type theories out there for you to consider.

 

[...bill]

I've seen ignition coils on cars do wacky things (back when there was one ignition coil). Tach response is just early indication that things are beginning to break down. Usually, there's no warning. Running great, then bam, backfire.

Interesting about the TB-V modules. We have the current incarnation of what Mercruiser sells as the TB-V module, including the connector harness that connects to the engine harness. There are indeed 2 ground wires. One has the big eyelet and longer wire that's connected to the block. The other is a short black wire with a small eyelet that's connected to one of the TB-V module mounting screws.
The TB-V module is mounted to a bracket and sandwiched with the knock controller. I noticed the mounting bracket is well painted as well as the block where the bracket mounts to the engine. Point is, doubtful that bracket electrically connects to the engine. I'm "assuming" that ground wire on the mounting screw grounds the mounting bracket. Due to paranoia about this issue, I scraped paint off the mounting bracket and the block, so that's all electrically connected mechanically. Made no difference...

Engine hatch cover can be open or closed and this still happens. I've been in there too many times, so it's not too hot and there is oxygen.

The engine history: Boat was re-engined around 2005. Restorer (not us), found these 2 engines with low hours. Supposedly, < 50 hours, but we have no other history. They are Mercruiser 5.7L, 1997 with adjacent serial numbers. 0K189098, 0K189099. They are completely stock, and maintained with Mercruiser parts. These engines now have ~340 hours on them.
I rebuilt the carbs last fall, and it was pretty clear I was the only one that's been in there since built. I did not record the primary jet sizes, so they could be jetted wrong. It runs great up 'till this happens, so skeptical it's jetted wrong. Good idea!

Flame arrestors are stock Mercruiser and they are very clean.

There are a couple of goofy things about this boat that may be notable.
First, it?s under propped. Until next haul out, we?re OK doing 17 knots at 3500RPM. Don?t believe that affects current issue, but will be a good discussion for later.

Main goofie-ness is the way the charging system is wired. There are 2 start batteries, a house bank and an inverter. This was wired before ACR?s (automatic battery combiners) came to be. It uses diode combiners instead. I traced every wire of this scheme and confirmed it?s working OK. I don?t like it, but it works. Each alternator connects directly to its dedicated start battery. + from each start battery goes through a diode and then to the house bank. So, either or both alternators charge the house bank.
I?ve entertained all sorts of cases of diode failure modes, and can?t come up with anything that could cause the engines to run badly. Worst cases: diodes open and house bank doesn?t charge. Or, diode shorts and batteries simply combine. I?ve monitored charging voltages and currents and it appears to work ok. Downside is house bank never gets to 100% while underway, but that?s ok as we?re always on shore power at end of day.
Would like to rewire with ACR?s, but not until we get the engines reliable.

Sorry to be long winded. Any clues welcome,
?bill

 

[Fun Times]

It's under propped. Until next haul out, we're OK doing 17 knots at 3500RPM. Don't believe that affects current issue, but will be a good discussion for later.

What top RPM are you able to currently achieve? Your range is marked as 4200 - 4600 RPM on your 250 HP engines. http://boatinfo.no/lib/mercruiser/ma...ser17.html#/26

Let's hope it's just the coils.:encouragement: Good luck.

 

[...bill]

We could redline the engines, but we don't go there. No problem getting to max rpm's, however we've never tried to bounce off the rev limiter.. Typically, we keep it under 4000 and cruise 3200-3500. Due to under-prop, carb secondaries never kick unless accelerating. We'll size props correctly after it runs properly (some day).

 

[...bill]

Update...
Went for a sea trial in Puget Sound this evening. Both engines never skipped a beat with over 2 hours of hard running. We're cautiously optimistic and had a beer to celebrate.

Replaced both ignition coils with new Sierra parts. Parts removed were original 1997 coils. While I was at it, moved the distributor sensor ground wire from the bell housing to the block ground stud.
Loosened and checked all ground and power connections for due diligence. They were all good.
Next event will be this weekend when we'll attempt a 100 mile 2 day round trip in Puget Sound. If engines make it through that test, we may deem this issue fixed.
I'll report back, after that event.

I'll perform some forensic testing on the old coils after our return.

Thanks for the support here,
...bill

 

[Fun Times]

Hope this is the end of a long diagnostics test for you.....At least you now know all the other components are up to specifications as of right now for your next big trip out :encouragement:....Have fun, be safe and good luck out there.

 

[...bill]

Epilogue.
Update from the weekend cruise. In a word, fantastic.
45 non-stop nautical miles Saturday in Puget Sound running hard for 2.5 hours. Never skipped a beat. Ride home Sunday was another 45 nautical miles and 2.5 hours, also uneventful.

Root cause of the problem: both ignition coils were faulty. As the coils heated, they would begin to short/leak internally causing bizarre behavior. These coils are over 20 years old, and clearly, their time was up.
There is no easy way to test this failure mode as it?s a combination of temperature sensitivity with high voltage leakage effects. I simply swapped out the coils, mainly because there was nothing left to replace. In this case, the shotgun method worked. However, this problem persisted for 3 summers.

For those reading this trying to discover if you too can blame your problems on ignition coils, my answer is "maybe".
The real issue in this endeavor was the failure to use effective problem solving to discover the ROOT CAUSE sooner.
I discovered many real issues on this boat that allowed jumping to the conclusion that "this has to be it".
So, I caution anyone trying to diagnose a difficult problem to take a step back and unemotionally, record what you know, and what you don't know.
The first step in effective problem solving is to identify the problem. Sounds stupidly simple, but as humans, we usually skip this step and jump directly into problem solving without defining exactly what the problem is. We immediately begin looking under every rock to discover something wrong. If you find anything wrong with anything, it?s easy to conclude that has to be it! We don?t fathom that unrelated things can be simultaneously broken that have absolutely nothing to do with the problem you?re trying to solve.

The problem: Twin Mercruiser 5.7L engines stumble/backfire after 40 min of hard running. 10 minutes of cool down returns normal operation.

I clearly understood the bad behavior, but should have written down what can cause this to happen and devise an experiment to prove or disprove each.
First mistake: I jumped directly into fixit mode. I asked owner/buddy when was last sparkplug change? He said, ? I dunno, maybe never.?
So, changed plugs/wires/cap and did a compression test. It was good practice to replace all of those parts, but it did not affect the problem. It was not a waste of money, but it was a waste of time.
Second mistake: Jumped deeper into looking under rocks. Swapped the carbs, swapped the coils, installed new distributor ignition modules. None of this proved anything because I had not planned how to demonstrate any of those parts caused the ?problem?.
Third mistake: This is about when I noticed a timing light would stop flashing coincident with engine stumbling. This could have led me to the root cause, but others suggested looking at the fuel system as fuel is common to both engines. Makes sense to look at fuel and this led to a wild ride of discovering assorted time consuming fuel issues. Carbs turned out to be filthy with junk in the fuel bowls and clogged secondary circuits. Tanks contained a lot of crud which led to flushing/cleaning. Fuel hoses were long past due for replacement.
Again, not a waste of money, but a waste of time. These issues had nothing to do with the root cause.
I was simply looking under rocks and finding crabs.

Shoulda/coulda. What would have been better diagnostic practice?

1. State the obvious: it?s a mechanical issue, a fuel issue, or an electrical issue.
2. Devise a plan to prove/disprove each in order of effort:
   1. Mechanical integrity: Do a compression test, maybe a leakdown test too. In hindsight, compression test was sufficient to rule out major mechanical issues. I indeed did this early, but only because of shot-gunning spark plugs.
   2. Fuel: Run engines from external tanks. Although there were real fuel issues to resolve, running on external tanks would have quickly proven those fuel issues had nothing to do with the current problem.
   3. Electrical: Swap parts port/starboard. After that, swap in known good parts beginning with the least expensive first. Zoom in on why timing light dropped out when problem occurred. I would have still spent more time in this area than any other, but could have zoomed in quicker having moved past the mechanical and fuel systems.

I pride myself at fixing difficult problems. In this case, I was successful, but didn?t follow much of my own preaching. Still, it is remarkable both coils failed at the same time.

Here's Sophia Marie parked at our destination in beautiful Poulsbo Washington.


Thanks again for the help and wisdom in this forum,
?bill

 

[alldodge]

:thumb: