2010 Traverse 3.6 - Misfire codes & flashing MIL

Hey all, My name is Lee, I've been a tech for over 30 years and my hair is beginning to fall out over this traverse. This post will be long, cause I've done a lot so far. Car has only 20k miles and is super clean.
I see random misfire counts at idle (GM says some are normal), all 6 cylinders and no 1 cylinder more than others. On medium acceleration (sometimes light, sometimes heavy), mil flashes & sets P0300, 301, 302, 303, 304, 305 & 306. I don't feel misfiring and the car has plenty of power. Those 7 codes set together every time and no other codes.
we've checked and cleaned carbon from valves (was bad), compression 165 - 175, leak down < 4 - 5%, new ac delco plugs (.043 gap) & coils, scoped cam & crank, checked trigger wheel & replaced crank sensor anyway (ac delco only $20), checked timing marks, valve springs & cam lock to actuators (tight & cam variance 0.5 degrees @idle, averages +/- 1 or 2 driving & never spikes higher then 4 degrees). exhaust back pressure <1@ idle & 2500, <2.5 on hard snap on both banks. I updated the ecm to latest & with the flash for cold start misfires. Total fuel trim bank 1 averages 0 to +10%, bank 2 averages -2 to +7. Injector drop 425 - 434 psi (anyone know real world good numbers for 3.6L injector drop?). Ran a can of 3M inj cleaner to the hpfp @60 psi until empty. First fuel tank of gas was run till empty and refilled... twice. battery tests good, AC voltage less than .25 volts, alternator ripple and ground voltage SUCKS on this car! however, ive spent hours and hours checking & cleaning all the ground contacts, they are good. and the car still sets misfire codes with the alternator disconnected anyway. i believe the crappy ripple and grounds is due to the injector firing voltage, but i'm not sure. ive driven with maf unplugged, front oxys unplugged, and then rear oxys unplugged. ssdd. ignition coil & fuel injector fault pids never show fault. fuel pressure reads around 1700/1750 psi under acceleration when mil begins to flash. im sure we checked other stuff too but cant remember right now. what am i missing? gotta be missing something. oh, ive also pulled off the belt, still minor misfires at idle, but didn't drive it with belt off. pulleys & tensioner movement feel good too. mil flashes with torque converter commanded on and off, doesnt matter, and tcc slip is less than 100 at 40% on (or 40 psi... I forgot what that tcc pid displayed). Thanks for reading and Im open to ideas. - Lee

Have you done a crank variation relearn?

Might just need a new set of injectors sounds like it does a ton of sitting

Your description gives the impression you're not sure/in doubt wether misfires actually are happening when they MIL flashes. (It's strange that all of the specific cylinder codes are registered at the same time.....is the FFD identical?)

If this is true as a first step I'd try to confirm/disregard potential misfires, using a scope. To do so if you had a pico you could graph the rpm as math channel based on the CKP sensor and a sync signal. With any other scope you can scope the alternator's voltage output (AC- coupled for better resolution) and again a sync signal to identify cylinders.

Both ways will identify misfires if present and next steps would depend on the result of this test.

This vehicle does not have a crankshaft learn procedure.

It is possible. I'm hoping someone would know what the actual injector balance results would show. I can say that after running the 3M cleaner through the injectors, I saw no difference at all.

I am not sure if it's actually misfiring. It does not feel like it. Im thinking if all cylinders began misfiring enough to flash the mil, I would feel something. Some shaking or slight rpm drop. But, I feel nothing.
FFD?
I use the old snap on MODIS for a scope. I've scoped the crank & cams at idle, pattern is good. I didn't scope while driving. Ya lost me at the rpm math/alternator thing to confirm misfires. Can you explain that a bit more?

ProAuto wrote:

It is possible. I'm hoping someone would know what the actual injector balance results would show. I can say that after running the 3M cleaner through the injectors, I saw no difference at all.

This is the injector balance test results from a 2011 GMC Acadia with a new set of ACDelco injectors:



www.scanshare.io/share/S-CTmsafCkWDHDfeHZwnFQ#0,1,2,3,4,5

ScanShare and Shopstream Connect like to drop decimal points. :silly: The actual results are 430-440 PSI, not 4300.

That said, I have never liked the automated injector balance test, and I wouldn't place ANY weight on it if I were you. I've had multiple vehicles pass the automated test over and over, only to get fixed with a set of injectors.

Do you have a data capture of the Misfire data list when it starts flashing the MIL? I believe Juergen meant Freeze Frame Data when he said FFD.

Tyler is right with his FFD comment, I meant freeze frame data. Sorry for being cryptic.

On the rpm/ alternator thing:

The crankshaft speed at a given rpm is not constant, each time a cylinder fires it (slightly) accelerates and then decelarates. These deviations from a constant, lineal speed are supposed to be even between cylinders. While misfiring the deviations caused by an offending cylinder will be more pronuonced.

The alternator´s voltage output allows for detecting these deviations of crankshaft speed with a scope; the voltage output depends (among other things ) on crankshaft speed and will reflect exactly the little changes in crankshaft speed. To make them visible with the modis try to AC-couple in peak detect mode. Don´t put to much time on the screen as you want to see detail, yet enough to see various complete crank turns, so it´s a combination of screentime and buffer .Use a second channel sync to a particular cylinder and apply the firnig order to identify all of them and their contributions to crankshaft speed.

I hope this clears things up a little bit.

juergen.scholl wrote:

To make them visible with the modis try to AC-couple in peak detect mode.

Maybe this needs a processed "AC" voltage to be useful? Tried this yesterday on a 4 banger with an unplugged injector and couldn't see any change in AC voltage with the 1008..

Matt T wrote:

juergen.scholl wrote:

To make them visible with the modis try to AC-couple in peak detect mode.

couldn't see any change in AC voltage with the 1008..

Not sure about hantek 1008 soft- and hardware capabilities..... I take it your's modified to show AC voltage, isn't it? I'll look for capture later and upload it.

Matt, the attached pictures show the technique at work, there is a heavy filter, a good measure of horizontal and vertical zoom applied. This powerful scope will allow to detect the AC-voltage riding on top of the DC portion without further ado

I suspect the hantek simply is not up to that task, the modis is another league and capable.

Pictures attached.

Picture attachment is giving me a HARD TIME, can´t get the relevants through?


....

That's helpful. Im only about 2 percent less then those numbers, so I wouldn't think that's a problem. Also, my fuel trim isn't going high and o2s aren't going lean before the codes set.
I do have recordings and freeze frame data stored. The freeze frame data is after the alleged misfires begin so I'm not seeing anything helpful there. One thing that sucks about this car is I can't see misfire counts and O2 data or other helpful pids simultaneously.

I'm gonna try it just to see. I am curious, but not optimistic as my alternator output is very hashy and I can barely even make out ac ripple.

Got it, thank you. I'm gonna try it. But seeing as the ecm is picking up microsecond differences between crank pulses, and my alternator voltage is hashy, Im thinking I'm not gonna see anything there.
What do you think about these crank signals? The last one is during high misfire counts. The hash is always there on this car and I have no idea where it's coming from, or if it's causing a problem. Alternator disconnected, hash and misfire counts are still there.

I have to lower the resolution on my pics to post here. Ill take them tomorrow. Sorry bout that.

ProAuto wrote:

One thing that sucks about this car is I can't see misfire counts and O2 data or other helpful pids simultaneously.

I know the feeling. Definitely frustrating.

Now probably isn't the time to go learning a new scan tool, but HPTuners VCM Suite has become my go-to on GM products like this because you can view any combination of PIDs you want. O2's, miss counts, cam timing, fuel trims, all at the same time. No other aftermarket scan tool that I know of can do the same.

www.scannerdanner.com/forum/post-your-re...t.html?start=0#50122

This may or may not be worthwhile for you, depending on how many GM's you see at your shop.

That is good to know. Thank you. If that tool can do it, then that means the ecm is capable of pushing misfire and other pids simultaneously. I wonder if I can read data from 2 scan tools at the same time via an obd tee. Hmmmm.