Cold start rough idle and sometimes 303 code

[ShadowsPapa]

Spell check ugh, I need the OBD plug in which still hasn’t shipped.
As to valve cover, I’m about to the point of pulling it to confirm. I may bum a borescope from parts store first to check for coolant since it sat all night.

Right now, all things are on the table since there's no other information to narrow it down.
I would ask - has the coolant level dropped at all? If there's coolant getting into a cylinder(s), then it has to come from somewhere, meaning coolant going out, the level MUST drop the same amount.
Normally, these only drop a little bit even over the course of a year or three. I've only ever needed to add coolant once in my 2020 - 8 ounces, and it never dropped again. Likely it had an air bubble from the factory. The others have held totally steady on the coolant level. I wait until it's about a month old then use a sharpie and put a dot at the level when the thing is cool in the morning and temps are in the 70s or so. Then I can tell at a glance where things are. It will normally vary from that dot a little - maybe higher, maybe lower, depending on the temperature outside, because coolant does expand with heat, but not by much from that dot.

Sponsored

 

[Josh00333]

I took my Jeep to a shaman, he blessed it with the blood of a roster no codes so far.

I suggest finding shaman, now no blood of a hen, you need a rooster, only thing that cures ghost codes.







What? not help full? Sorry

 

[Hoist]

More to follow, now where did I leave that shaman………

 

[Hoist]

took a picture of the cylinder. Nothing obvious however, is that a barcode on the piston? How could that survive on a piston?

 

[ShadowsPapa]

took a picture of the cylinder. Nothing obvious however, is that a barcode on the piston? How could that survive on a piston?

Likely etched in.

Not sure but perhaps grade 1 or 2 is a size difference like they have historically done with parts.

 

[Hoist]

Interesting. Etched would make sense.
been putting off pulling valve cover given how much crap is in the way But that may be next step. I haven’t been able to use Jscan since I don’t have the plug in yet. Asked the company in Phoenix why no ship and they were looking into it. Yup.

 

[Old Dogger]

Possible vacuum leak. My Cadilac Escalade did exactually what you are describing. The cause was an Intake manifold gasket. It leaked while it was cold, and then it swelled enough when hot to seal.

 

[Lost1wing]

If another cylinder is carboned up, a coolant leak could be the reason why you can see the barcode on nbr 3. If another cylinder is as clean as nbr 3, move on with removing the valve cover. Oh, after the tool shows up and you do a misfire count.

 

[Lost1wing]

.

 

[Lost1wing]

Darn phone.

 

[Hoist]

I have to disconnect the lines from the valve cover either way so I’ll check another cylinder so I have something to compare. I may swap coils with another cylinder to see if it follows.
Havent had the code since I cleared it last but still get the intermittent roughness and sometimes A light. Since I haven’t had the code, I can’t say it’s nbr 3 for sure, but that’s the only code I’ve had.
In all the reading I’ve done, nbr 3 seems to have most of the problems. Has anyone identified the “why” and can you really make this engine reliable?

 

[ShadowsPapa]

Possible vacuum leak. My Cadilac Escalade did exactually what you are describing. The cause was an Intake manifold gasket. It leaked while it was cold, and then it swelled enough when hot to seal.

Very different engines and construction. The timing of the loss of misfire is also off to allow for that.
Of course a simple vacuum gauge would tell.

 

[Lost1wing]

Two things you will find with a misfire count.

First would be that the misfire is only on nbr 3. You have swapped the plug and the coil. The fuel injector can not be replaced without further disassembly. I would pull the valve cover and look at the cam. If nothing found. Swap fuel injectors. ( New orings).

Second possibility will be multiple misfires but the computer is flagging the worst one (nbr 3). Now this could be an egr problem, vacuum leak or even just a timing issue, especially if the misfires are 1,3 and 5.

It would be nice to know the issue before pulling the valve cover, but I think you are going to be just about there in any situation.

Are you able to reach the nbr 3 fuel injector connector? Can you remove and reinstall it?

 

[ShadowsPapa]

Two things you will find with a misfire count.

First would be that the misfire is only on nbr 3. You have swapped the plug and the coil. The fuel injector can not be replaced without further disassembly. I would pull the valve cover and look at the cam. If nothing found. Swap fuel injectors. ( New orings).

Second possibility will be multiple misfires but the computer is flagging the worst one (nbr 3). Now this could be an egr problem, vacuum leak or even just a timing issue, especially if the misfires are 1,3 and 5.

It would be nice to know the issue before pulling the valve cover, but I think you are going to be just about there in any situation.

Are you able to reach the nbr 3 fuel injector connector? Can you remove and reinstall it?

Depending on the misfire count, it would be very unusual for it to not show a multiple misfire code with others, even at a much lower count, misfiring.
I had huge differences in misfire numbers - one cylinder was in the hundreds, while others only maybe a dozen, 2 dozen, it always still shows a multiple cylinder misfire code. I think the highest reached for one cylinder was about 1,000, the others way down in numbers - still multiple cylinder misfires.
So I'm betting on a single cylinder misfire.

A vacuum leak will usually also trigger other codes - the air going in isn't accounted for, the cylinder runs lean and you have the PCM trying to account for the lean O2 sensor readings, triggering codes. P0171 for example, lean, bank 1. You'll find that sort of thing with JSCAN if you know what to look for, but it should trigger a MIL and set a code.
There are also codes for intake air system leaks, varying with what's detected.
You can't have an intake leak without a lean condition. These detect pressure, flow, intake air temperature, and so on. An intake leak also causes pulses in the air going in.

Again, we are sort of flying blind - tossing stuff at the wall to see what sticks, until or unless some more information comes in.

If he could scope the thing or even just connect a vacuum gauge, it would be settled.
JSCAN and some diagnostics would help.

 

[Lost1wing]

Depending on the misfire count, it would be very unusual for it to not show a multiple misfire code with others, even at a much lower count, misfiring.
I had huge differences in misfire numbers - one cylinder was in the hundreds, while others only maybe a dozen, 2 dozen, it always still shows a multiple cylinder misfire code. I think the highest reached for one cylinder was about 1,000, the others way down in numbers - still multiple cylinder misfires.
So I'm betting on a single cylinder misfire.

A vacuum leak will usually also trigger other codes - the air going in isn't accounted for, the cylinder runs lean and you have the PCM trying to account for the lean O2 sensor readings, triggering codes. P0171 for example, lean, bank 1. You'll find that sort of thing with JSCAN if you know what to look for, but it should trigger a MIL and set a code.
There are also codes for intake air system leaks, varying with what's detected.
You can't have an intake leak without a lean condition. These detect pressure, flow, intake air temperature, and so on. An intake leak also causes pulses in the air going in.

Again, we are sort of flying blind - tossing stuff at the wall to see what sticks, until or unless some more information comes in.

If he could scope the thing or even just connect a vacuum gauge, it would be settled.
JSCAN and some diagnostics would help.

I believe he has an Innova 5100 series scan tool. He code see if it is running lean. But yes, it should throw a code for that. I have seen a stuck valve cause misfires on other cylinders, but I saw a p0300 with the bad cylinder fault. His may not be that bad yet to give a p0300 fault. Like you said, just a guess or what we used to call a WAG.

Sponsored