Cylinder 6 Misfire Mystery

[threefiveleven]

Ok folks this isn't a short story, so I'll thank you up front for taking time to view my post. I'm at a loss.

2007 TBSS. Completely stock. About a year ago I worked through the fuse box issue. I repaired mine using instructions from tbssowners.com. Helpful people there too.

While it was down anyway, I did other repairs that I'd been putting off. All 4 corners of the exhaust manifold bolts were broken off. Front/rear of both sides. I was able to weld nuts onto the broken studs, and though it was a pain in the butt, it was an uneventful repair. New bolts and gaskets. Battery was disconnected during the repairs. Also replaced lower control arms which had worn out bushings. New plugs and wires. Just under 130k on the clock.

Point being, I did all of these repairs while it was down, and prior to these repairs I had no problems with the vehicle. It was setting the codes that are generally attributed to the fuse box issue. After all of these repairs, the very first test drive gave me hard shifts (P-R, R-1, and 1-2 shifts hard when cold, but then become normal within less than a mile. May or may not shift hard again on the next drive.) as well as a Cylinder 6 misfire code.

Eventually, one at a time, I moved #6 plug, wire, coil, and fuel injector to #4. Made no difference. Still sets #6 code. Again, this has been more than a year ago. The code comes and goes, but it never runs badly or gives any indication that a cylinder is actually misfiring. Sometimes the code will disappear and it will shift normally for a week or so, but it always comes back. Compression test looks perfect on all 8 cylinders. I have visually inspected for vacuum leaks as well as tried spraying brake cleaner around the intake to check for vacuum leaks in the gaskets. If there is a vacuum leak, I cannot find it and/or it is miniscule.

Thinking that my fuse box repair was somehow at fault, I ordered a new GM fuse block. Made no difference whatsoever.

Resorting to frustratedly throwing money at it, I also replaced the MAF, throttle position sensor (on the throttle body itself) and cam position sensor. No change at all.

I can't disregard that this happened immediately after the repairs I did last year. Not just sorta after, but right out of the driveway that day. I kept thinking whatever is causing it would get worse and eventually set another code or otherwise become obvious what the problem is. I've been thinking that for a year. I have to drive it, and it won't pass inspection. Plates are 4 months expired as a result.

Wild guesses are welcome, or if anyone is skilled at reading the entrails of a chicken, hey, I'll take it. Traditional troubleshooting is failing me because it isn't actually running badly. It's as if the computer is getting bad input that leads it to believe #6 is misfiring even though it isn't. I can't even guess at the shifting problems, but they are closely related to the misfire code in their timing.

Thanks again for reading.

 

[TollKeeper]

After checking to make sure that the COP (coil on plug) is actually firing...

I think I would start with a leak down test. When you run live data, does it give counts of misfire, and how fast the misfire is counting.

Usually a cam is not a huge concern on the non-DOD/AFM trucks, but that doesnt mean it doesnt happen. Pulling the valve cover off, and rotating the engine manually, watching for lift on the valves, using a micrometer to compare to the other valves, would be my next suggestion. This in and of itself doesnt mean the cam is wiped, but could mean a pushrod is bent, has fallen out of its journal, or has a broken end.

 

[threefiveleven]

After checking to make sure that the COP (coil on plug) is actually firing...

I would start with a leak down test. When you run live data, does it give counts of misfire, and how fast the misfire is counting.

Usually a cam is not a huge concern on the non-DOD/AFM trucks, but that doesnt mean it doesnt happen. Pulling the valve cover off, and rotating the engine manually, watching for lift on the valves, using a micrometer to compare to the other valves, would be my next suggestion. This in and of itself doesnt mean the cam is wiped, but could mean a pushrod is bent, has fallen out of its journal, or has a broken end.

I don't have a good scan tool. I use Torque Pro for Android, and if it has an ability to see live data I don't know how. If you have a better suggestion for seeing live data, I'm all ears. I am disenchanted with this vehicle and would like to get it running perfectly again (it is relatively flawless and ultra clean) so that I can sell it and buy an older truck. So I'd rather not spend money on a good scan tool. My vehicular needs are simple and a solid old square body pickup will do quite nicely. Perhaps pay someone talented with a scan tool to look at it for me.

When you described measuring the valves, are you saying to measure the amount of compression on each valve spring? Is that the ideal way to measure to compare to others?

 

[Beacon]

I think you were on the right track with the idea that the computer is getting a bad signal. Misfire is determined by the crankshaft position sensor. I would start there. You've already swapped out with number 4 and it's still showing number 6. I would check the wiring to the CSP sensor, maybe it's pinched/ partial broken. Also I've read that having a case relearn done, has cleared up misfire problems in the 6.0

 

[Mooseman]

I had an intermittent misfire on my 4.2 like you without the hard shifting. Not the same engine but the troubleshooting steps are the same:

Very Slight Miss at Idle

Yesterday on the TB, at 117k km, saw the CEL come on and code came up for P0304. Fired up Torque and it was missing on cylinder #4 but only at idle or stopped at a light. It's not a big miss since I can't feel it at all and it's just a couple of misses per refresh cycle. While driving...

gmtnation.com

For the hard shifting, I had that when I accidentally left the MAF disconnected on the Saab. Shifted very hard but it did set a code for the MAF.

How's the gas in the tank? Being stored for a year, maybe there's a bit of water in it?

 

[threefiveleven]

After checking to make sure that the COP (coil on plug) is actually firing...

I would start with a leak down test. When you run live data, does it give counts of misfire, and how fast the misfire is counting.

Usually a cam is not a huge concern on the non-DOD/AFM trucks, but that doesnt mean it doesnt happen. Pulling the valve cover off, and rotating the engine manually, watching for lift on the valves, using a micrometer to compare to the other valves, would be my next suggestion. This in and of itself doesnt mean the cam is wiped, but could mean a pushrod is bent, has fallen out of its journal, or has a broken end.

Thanks for your help btw.

I missed your first statement, making sure the COP is firing. Do you mean that switching it from #4 to #6 isn't adequate to be sure it's firing properly? I can do the good old 'fire it up' and check for spark at the plug wire, but that may or may not tell me that it's firing adequately to prevent a misfire? Is that what you mean?

 

[threefiveleven]

I think you were on the right track with the idea that the computer is getting a bad signal. Misfire is determined by the crankshaft position sensor. I would start there. You've already swapped out with number 4 and it's still showing number 6. I would check the wiring to the CSP sensor, maybe it's pinched/ partial broken. Also I've read that having a case relearn done, has cleared up misfire problems in the 6.0

I can do the CSP sensor too. It's not terribly expensive. In fact I got one, and installed it, only to find out that it looked identical but I'd gotten the wrong part number. So I put the old one back in and never got around to re-replacing it. I read about the CASE relearn but I don't have the tool to do it.

 

[threefiveleven]

I had an intermittent misfire on my 4.2 like you without the hard shifting. Not the same engine but the troubleshooting steps are the same:

Very Slight Miss at Idle

Yesterday on the TB, at 117k km, saw the CEL come on and code came up for P0304. Fired up Torque and it was missing on cylinder #4 but only at idle or stopped at a light. It's not a big miss since I can't feel it at all and it's just a couple of misses per refresh cycle. While driving...

gmtnation.com

For the hard shifting, I had that when I accidentally left the MAF disconnected on the Saab. Shifted very hard but it did set a code for the MAF.

How's the gas in the tank? Being stored for a year, maybe there's a bit of water in it?

Oh no it only sat for about a month. I've been driving it the rest of the past year since the repairs. Countless fresh tanks with no change in behavior or code.

 

[TollKeeper]

Taking the wire off, and verifying the color of that spark. Switching the coil to see if the misfire follows the misfire is very helpful. But that doesnt show if the coil itself if firing, or even getting power, or reference signal to fire.

As for the Torque Pro app, I know very little about that app. So I cant help you there. I think its @Mooseman ? that is the go to guy there.

 

[Mooseman]

Yes, you can use Torque Pro to view live data. It's been a while since I've used it but you go to a blank page and add a gauge. Then you just select the PID you want to monitor. Scroll down until you find Engine Misfire Cyl.1 and add it. Use either a numerical digital type of gauge.

@TJBaker57 is actually our SME on these things.

 

[budwich]

torque app will do what you need. You go into the settings menu and you will find "manage extra pid". In there, you can scroll down / thru the various pids. Eventually, you will get to cylinder misfires (history and current) for each cylinder. You can create displays using the pid to monitor any of them just like any other display (ie. temp, rpm, etc)... use a small / tiny "digital". You would need to create page of displays, 8 on a page for current and 8 for history. Hopefully, you are already doing some live data monitoring. If not, you can set that up from the main menu "wheel" of choices. Tjbaker57 knows alot about the various apps that can be used. There are probably easier ones but torque can do what you need in a "pinch". Note: misfire counts aren't "cumulative", they are "period" based so you shouldn't see them continually climbing unless you have serious misfires. In addition, to get a cylinder code, the misfires for that period have to cross a boundary for counts, 1 or 2 etc won't cause a code and probably won't even "flicker" the "check eng light" (which will flicker for a level of misfires).

Do you do any monitoring of data currently: fuel trims, timing, temp, vacuum, etc?

 

[TollKeeper]

@TJBaker57 is actually our SME on these things.

I knew if it wasnt you, you knew who it was.

Hopefully my above responses are understandable. At work, and have a monitor over my shoulder, so having to check this as time permits.

 

[TJBaker57]

I can do the CSP sensor too. It's not terribly expensive. In fact I got one, and installed it, only to find out that it looked identical but I'd gotten the wrong part number. So I put the old one back in and never got around to re-replacing it. I read about the CASE relearn but I don't have the tool to do it.

My thoughts are that you need a CASE relearn to begin with.

Any repair that involves disturbing the rotating components like the harmonic balancer or even the bolt that holds it, the flywheel (flexplate), or the like, or replacing or otherwise disturbing the crankshaft position sensor requires the relearn. Without that the PCM will not be able to accurately detect misfires, or so I have read in service literatures.

Depending on your location you may be able to find someone who owns a Tech 2 or equivalent to do the relearn.

 

[Reprise]

A couple of thoughts from my end...

Since you replaced the crank position sensor (or will, again), you'll want to do the crank position learn sequence again, just for grins. It's an easy enough thing with a high-end scan tool -- enter the mode, hold a particular engine speed range for a few seconds in park, and done.

When I replaced mine during a top-end rebuild on my LQ4, I didn't do the relearn -- for months -- as I was a bit ignorant about it. But -- I had no problems with misfires, etc., either, during all that time. Admittedly, GM was very conservative with the timing (and everything else) on that engine, as it went into the 3/4-ton & up light-duty trucks & vans. BTW... I guarantee you have more timing in your engine, by default.

It wasn't frequently driven when I first got it back together, but I was monitoring live data with it almost every time I drove it. If I bumped up timing enough (in real-time), I could introduce misfires, but lowering it back down removed them immediately.

After I finally did the crank learn sequence, I didn't encounter any additional misfires, either. By rights, I should've (?) seen some misfires before doing the crank relearn, as I manipulated the crank a ton during the buildout.

Another thing that I've run across other people reporting (not here) is 'false misfires'. The ECM reports the misfires, but they're not really occurring. I don't recall the conditions, but they're enough to trigger the CEL, so it's above 30 or 60, which is the threshold, IIRC (and as @budwich mentioned, the threshold has to be crossed, or you'll never know it w/o a scan tool)

I don't know if the 'false misfires' issue is only on Gen III engines (yours is considered an early Gen IV), or if it's across all LS-based motors. If you want more info on this, send reply, and I'll track it down for you.

Oh, and I recall as I write this, that part of the issue was attributable to the knock sensors (in the lifter valley on Gen III; on the lower side(s) of the block on the Gen IV). The Gen III sensors are really touchy, it appears -- people recommend replacing them if they're removed from the engine (no matter how new), and they don't tolerate being dropped at all.

The Gen IV sensors use a different method of detection, so they're probably not as flaky as the Gen III (and probably why GM switched to that method). Again, I can share my experience, which was breaking the connector on one of the sensors during an intake gasket replacement a few months before the top-end rebuild. I replaced both with new Delphi sensors, and reinstalled them during the rebuild a few months later -- and ran into no issues afterward.

These two things might be worth the proverbial grain of salt, but you asked for feedback. Hopefully, something mentioned in the replies helps you get the issue resolved, and lets you get legal again. 🤞

On edit, as I recalled some more info: IIRC, certain cylinders on the LS are more prone to report misfire than others -- and I think one of the problem children (more than one) is #6. Now that I've brought that up, I'll see what I can find out about these various misfires and report back (with links). Check back in 24hrs; I should have something posted by then.

 

[NJTB]

If I remember correctly, the valve springs on these have an outer and inner spring. Triple check the inner spring isn't broken.

 

[threefiveleven]

I knew if it wasnt you, you knew who it was.

Hopefully my above responses are understandable. At work, and have a monitor over my shoulder, so having to check this as time permits.

Yes absolutely, I follow you. Don't get yourself in trouble on my account! lol. It'll be tomorrow before I can get to it anyway. I'm working 2 jobs this week and have another vehicle in pieces so the Blazer is next on my list but it's getting me around for now.

 

[threefiveleven]

My thoughts are that you need a CASE relearn to begin with.

Any repair that involves disturbing the rotating components like the harmonic balancer or even the bolt that holds it, the flywheel (flexplate), or the like, or replacing or otherwise disturbing the crankshaft position sensor requires the relearn. Without that the PCM will not be able to accurately detect misfires, or so I have read in service literatures.

Depending on your location you may be able to find someone who owns a Tech 2 or equivalent to do the relearn.

I'm thinking I'll get a new CSP Sensor and then find someone near me to do the CASE relearn for me. From there start inspecting valvetrain. I just feel like it must be related to the repairs I did because this happened literally on the way out of the drive after those repairs. But regardless, I'll have to just keep digging til I find the answer.

 

[threefiveleven]

torque app will do what you need. You go into the settings menu and you will find "manage extra pid". In there, you can scroll down / thru the various pids. Eventually, you will get to cylinder misfires (history and current) for each cylinder. You can create displays using the pid to monitor any of them just like any other display (ie. temp, rpm, etc)... use a small / tiny "digital". You would need to create page of displays, 8 on a page for current and 8 for history. Hopefully, you are already doing some live data monitoring. If not, you can set that up from the main menu "wheel" of choices. Tjbaker57 knows alot about the various apps that can be used. There are probably easier ones but torque can do what you need in a "pinch". Note: misfire counts aren't "cumulative", they are "period" based so you shouldn't see them continually climbing unless you have serious misfires. In addition, to get a cylinder code, the misfires for that period have to cross a boundary for counts, 1 or 2 etc won't cause a code and probably won't even "flicker" the "check eng light" (which will flicker for a level of misfires).

Do you do any monitoring of data currently: fuel trims, timing, temp, vacuum, etc?

I had no idea it could do all of that. I've mostly just used it to pull codes. I'll have time tomorrow to go sit with it and drive it a little with some live monitoring. I'm not currently doing any live data monitoring.

 

[threefiveleven]

A couple of thoughts from my end...

Since you replaced the crank position sensor (or will, again), you'll want to do the crank position learn sequence again, just for grins. It's an easy enough thing with a high-end scan tool -- enter the mode, hold a particular engine speed range for a few seconds in park, and done.

When I replaced mine during a top-end rebuild on my LQ4, I didn't do the relearn -- for months -- as I was a bit ignorant about it. But -- I had no problems with misfires, etc., either, during all that time. Admittedly, GM was very conservative with the timing (and everything else) on that engine, as it went into the 3/4-ton & up light-duty trucks & vans. BTW... I guarantee you have more timing in your engine, by default.

It wasn't frequently driven when I first got it back together, but I was monitoring live data with it almost every time I drove it. If I bumped up timing enough (in real-time), I could introduce misfires, but lowering it back down removed them immediately.

After I finally did the crank learn sequence, I didn't encounter any additional misfires, either. By rights, I should've (?) seen some misfires before doing the crank relearn, as I manipulated the crank a ton during the buildout.

Another thing that I've run across other people reporting (not here) is 'false misfires'. The ECM reports the misfires, but they're not really occurring. I don't recall the conditions, but they're enough to trigger the CEL, so it's above 30 or 60, which is the threshold, IIRC (and as @budwich mentioned, the threshold has to be crossed, or you'll never know it w/o a scan tool)

I don't know if the 'false misfires' issue is only on Gen III engines (yours is considered an early Gen IV), or if it's across all LS-based motors. If you want more info on this, send reply, and I'll track it down for you.

Oh, and I recall as I write this, that part of the issue was attributable to the knock sensors (in the lifter valley on Gen III; on the lower side(s) of the block on the Gen IV). The Gen III sensors are really touchy, it appears -- people recommend replacing them if they're removed from the engine (no matter how new), and they don't tolerate being dropped at all.

The Gen IV sensors use a different method of detection, so they're probably not as flaky as the Gen III (and probably why GM switched to that method). Again, I can share my experience, which was breaking the connector on one of the sensors during an intake gasket replacement a few months before the top-end rebuild. I replaced both with new Delphi sensors, and reinstalled them during the rebuild a few months later -- and ran into no issues afterward.

These two things might be worth the proverbial grain of salt, but you asked for feedback. Hopefully, something mentioned in the replies helps you get the issue resolved, and lets you get legal again. 🤞

On edit, as I recalled some more info: IIRC, certain cylinders on the LS are more prone to report misfire than others -- and I think one of the problem children (more than one) is #6. Now that I've brought that up, I'll see what I can find out about these various misfires and report back (with links). Check back in 24hrs; I should have something posted by then.

Dude that's a lot of good info, thanks! I'm wondering, if I followed you correctly, if my knock sensors are on the lower part of the block, if I may have whacked one or something while wrestling the exhaust manifolds out to replace the gaskets. I'll do a visual inspection but I also wondered if the welding might have fried any electronic components. Welding nuts on the broken exhaust bolts.

 

[TollKeeper]

I'm thinking I'll get a new CSP Sensor and then find someone near me to do the CASE relearn for me. From there start inspecting valvetrain. I just feel like it must be related to the repairs I did because this happened literally on the way out of the drive after those repairs. But regardless, I'll have to just keep digging til I find the answer.

Could always check the Tech2 forum, and see if anyone lives near you with a Tech2. I think? this is all that is required for the CASE. You profile doesnt show where you live, so I cant offer anything.

 

[Reprise]

Dude that's a lot of good info, thanks! I'm wondering, if I followed you correctly, if my knock sensors are on the lower part of the block, if I may have whacked one or something while wrestling the exhaust manifolds out to replace the gaskets. I'll do a visual inspection but I also wondered if the welding might have fried any electronic components. Welding nuts on the broken exhaust bolts.

Appreciate the kind words, but... info is only good if it's *accurate*. And I know I was wrong on at least one thing. What I was thinking about was "false knock", not "false misfires". If knock is (falsely) detected, knock retard (KR) is introduced by the ECU, because its assumption is that the knock was caused by detonation / misfire. We'll move to the Gen IV sensors on your engine next.

Here's a picture of one of the sensors on the passenger side of an LS2 engine (Vette). As an aside, they look a LOT like the Gen 3 sensors (located in the valley cover, under the intake). The sensor is just left of center, above the oil level sensor in the pan, with a white plastic protuberance in the center of it. As you can see, it's quite low in the block, so I doubt you really did any damage to it. There's another one in the same place on the other side; each 'listens' to one bank. BTW, the crank sensor is also in this pic, all the way at the back, just in front of those starter bolts. Your starter likely covers the knock sensor, on this side, and the subharness that connects to the starter probably has the connector plug to the KS, as well.




On the Gen 3, besides the sensors, the harness plugs and the plastic in the sensor itself deteriorates with age (it gets quite brittle, and if you're removing an old one, there's an excellent chance you'll break the plug or the sensor off. Thus, it's recommended to replace harness along with sensors, if one or the other has to be R&R'd. An additional issue with the Gen 3 location was that water can leak into the recesses of the valley cover where they sit, and corrode them (especially the rear one). Anyway... that's your sensor, above.

Sensors can be reduced in sensitivity, or tuned out of the ECU entirely (via HPTuners or EFILive), although I wouldn't do that, myself, especially with a street engine. But you can effectively disable them in software (and your emissions testing facility would be none the wiser). Sensors can also be tested on-vehicle, with a high-end scan tool.

Again, that just deals with knock retard, not misfires themselves.

Back to misfires...
I looked for awhile, and didn't find much on "LS cylinder x or y susceptible to misfire, vs. others". I know I saw it previously, but if I can't find it, it's not helpful.
Something I did come across was info that the LS motors run a bit rough (and can misfire) when cold. I can attest to this (and if you look at the ECU tables that control idle RPM, you can see they're set to a higher speed the colder the engine is (ambient vs. ECT sensor).

Now one thing I *did* find was some generic info on misfires (not LS- / GM-specific, but does cover OBD-II programming. A link is HERE, if you want to take a look. Note that it's from a collision repair site (!), but I thought it was useful, if only at a generic level.

Finally, one last thing I saw in a few places... if you've already swapped plug / coil / wire... one of the next things you'll want to look at is the injector, as flow might be less than optimal on that one cylinder, causing a lean condition. Also (and this is me spitballing)... the injectors are held in to the intake with o-rings. I think they're a synthetic material (viton -?) -- but even if they are, if the material degrades, it could be admitting more air at startup, until things heat up and expand. This is similar to intake manifold gaskets degrading and doing the same thing.

So... what would I do? As a quick test, you could swap another cylinder's injector with #6, and see if the problem moves with it. It may take a couple of days to show itself. When you pop them out, take a look at the bottom of #6 especially, to see if there's a lot of crud on it. However, I would *not* clean it right off -- put it in the other location, as-is. That way, if it sets the misfire on that cylinder, you can remove / clean / retest. It would be great if all this was, was a dirty injector or two, right? Next best would be if the #6 injector proved to be marginal, flow-wise, relative to the others, as replacing one injector isn't too hard on the wallet.

If you do remove (all) the injectors, it wouldn't be a bad idea to put new o-rings on. There are two per injector; using a small pick tool, it's easy enough to swap them out, and replacements are cheap (I used Fel-Pro on mine; if you need the part # info, let me know. They're sold in sets of 8 (to cover one bank of 4 injectors).
Note that the o-rings may be different sizes top / bottom. Make sure you put them in correctly!

Finally, the intake manifold gasket...
These are near legendary for degrading over time on LS engines. However, when they do, they admit too much air at cold startup, and will set a 'Lean Bank x' code -- which you haven't described. If it gets bad enough, the engine will run really rough when cold, until things heat up and the gasket expands to seal the leak -- until the next cold start.
Again, I doubt this is the (main) issue... but you can test with some carb cleaner sprayed at the junction of intake / head, while the engine is cold. If it immediately runs smoother, it's a pretty good sign you have a gasket that has seen better days. Fel-Pro makes improved versions (the cheaper one is fine, unless you want to get spendy.)

I think that covers things I brought up (or, at least, that's enough typing, for now).
Keep us posted!

 

[threefiveleven]

Hey everyone my week has exploded and it may be the weekend before I am able to read through all of these adequately to understand and then do more troubleshooting.

I at least wanted to write to express my gratitude for everyone's time. I really appreciate it and I'll be glad to have this mystery behind me. I'll report back soon!

 

[threefiveleven]

Appreciate the kind words, but... info is only good if it's *accurate*. And I know I was wrong on at least one thing. What I was thinking about was "false knock", not "false misfires". If knock is (falsely) detected, knock retard (KR) is introduced by the ECU, because its assumption is that the knock was caused by detonation / misfire. We'll move to the Gen IV sensors on your engine next.

Here's a picture of one of the sensors on the passenger side of an LS2 engine (Vette). As an aside, they look a LOT like the Gen 3 sensors (located in the valley cover, under the intake). The sensor is just left of center, above the oil level sensor in the pan, with a white plastic protuberance in the center of it. As you can see, it's quite low in the block, so I doubt you really did any damage to it. There's another one in the same place on the other side; each 'listens' to one bank. BTW, the crank sensor is also in this pic, all the way at the back, just in front of those starter bolts. Your starter likely covers the knock sensor, on this side, and the subharness that connects to the starter probably has the connector plug to the KS, as well.


View attachment 100230

On the Gen 3, besides the sensors, the harness plugs and the plastic in the sensor itself deteriorates with age (it gets quite brittle, and if you're removing an old one, there's an excellent chance you'll break the plug or the sensor off. Thus, it's recommended to replace harness along with sensors, if one or the other has to be R&R'd. An additional issue with the Gen 3 location was that water can leak into the recesses of the valley cover where they sit, and corrode them (especially the rear one). Anyway... that's your sensor, above.

Sensors can be reduced in sensitivity, or tuned out of the ECU entirely (via HPTuners or EFILive), although I wouldn't do that, myself, especially with a street engine. But you can effectively disable them in software (and your emissions testing facility would be none the wiser). Sensors can also be tested on-vehicle, with a high-end scan tool.

Again, that just deals with knock retard, not misfires themselves.

Back to misfires...
I looked for awhile, and didn't find much on "LS cylinder x or y susceptible to misfire, vs. others". I know I saw it previously, but if I can't find it, it's not helpful.
Something I did come across was info that the LS motors run a bit rough (and can misfire) when cold. I can attest to this (and if you look at the ECU tables that control idle RPM, you can see they're set to a higher speed the colder the engine is (ambient vs. ECT sensor).

Now one thing I *did* find was some generic info on misfires (not LS- / GM-specific, but does cover OBD-II programming. A link is HERE, if you want to take a look. Note that it's from a collision repair site (!), but I thought it was useful, if only at a generic level.

Finally, one last thing I saw in a few places... if you've already swapped plug / coil / wire... one of the next things you'll want to look at is the injector, as flow might be less than optimal on that one cylinder, causing a lean condition. Also (and this is me spitballing)... the injectors are held in to the intake with o-rings. I think they're a synthetic material (viton -?) -- but even if they are, if the material degrades, it could be admitting more air at startup, until things heat up and expand. This is similar to intake manifold gaskets degrading and doing the same thing.

So... what would I do? As a quick test, you could swap another cylinder's injector with #6, and see if the problem moves with it. It may take a couple of days to show itself. When you pop them out, take a look at the bottom of #6 especially, to see if there's a lot of crud on it. However, I would *not* clean it right off -- put it in the other location, as-is. That way, if it sets the misfire on that cylinder, you can remove / clean / retest. It would be great if all this was, was a dirty injector or two, right? Next best would be if the #6 injector proved to be marginal, flow-wise, relative to the others, as replacing one injector isn't too hard on the wallet.

If you do remove (all) the injectors, it wouldn't be a bad idea to put new o-rings on. There are two per injector; using a small pick tool, it's easy enough to swap them out, and replacements are cheap (I used Fel-Pro on mine; if you need the part # info, let me know. They're sold in sets of 8 (to cover one bank of 4 injectors).
Note that the o-rings may be different sizes top / bottom. Make sure you put them in correctly!

Finally, the intake manifold gasket...
These are near legendary for degrading over time on LS engines. However, when they do, they admit too much air at cold startup, and will set a 'Lean Bank x' code -- which you haven't described. If it gets bad enough, the engine will run really rough when cold, until things heat up and the gasket expands to seal the leak -- until the next cold start.
Again, I doubt this is the (main) issue... but you can test with some carb cleaner sprayed at the junction of intake / head, while the engine is cold. If it immediately runs smoother, it's a pretty good sign you have a gasket that has seen better days. Fel-Pro makes improved versions (the cheaper one is fine, unless you want to get spendy.)

I think that covers things I brought up (or, at least, that's enough typing, for now).
Keep us posted!

Ok I'm with you that it seems unlikely that I damaged a knock sensor with my repairs, so I'll reserve that as a last resort but consider it good info!

I'm going to order a CPS today and assume that's a good place to start.

I've swapped the #4 and #6 fuel injectors, with the same results... still sets #6 misfire code. I didn't inspect the injectors closely as you've described here though. Still, it seems unlikely that two injectors would behave the same after being swapped.

I was suspicious of the intake gaskets too, as I've heard they are a common problem. But like you've mentioned, it doesn't set any lean codes. However, I think it's time well spent to check with some carb cleaner like you've said. I did this before, but I don't recall if I did it with the engine warm or cold. I'll do both just to be sure.

I have some more responses but I'll post them separately to kind of summarize where I think I'm headed from here.

 

[threefiveleven]

Ok I had some time today to go back through everyone's info, and again, thanks for all the input!

I'm ordering a Crank Sensor today and will install probably next weekend.

I tried to add a dial in Torque Pro to give me a visual indication of when the misfire trips, but I couldn't find it in the list. I'll research this a little more and if I can't figure it out I'll reply back to the person who gave the instructions there.

I feel like that's the next best step- watching when the misfire code sets. I had no idea that was even possible. But, fore example, if it's an intake gasket leaking at cold startup, I'll see that by watching the misfire set. If it's a valve spring or other valve component, it seems like that'll set the misfire code more steadily. If it's setting the code only at certain rpm's or hard throttle etc, that's other good troubleshooting info. Since the code does come and go, it seems like this is the best way to decide what to try moving forward.

I'll probably have a little time to try that this evening before next week's chaos starts. My son has the Blazer today because we have his truck apart so we're sharing a vehicle.

Again thanks for everyone's help, I was overwhelmed by all the knowledgeable responses!

 

[highlander]

Did you ever figure this out?