pepehogge

I recently purchased a 1996 Camaro SS at a salvage auction. I originally bought it to harvest the T-56 transmission for another project and then sell what was left of the Camaro to make a few bucks back. However, when I got it home and inspected the car, it was actually in great shape (Arizona car). It was totaled by the insurance company because of an engine fire on the driver's side of the engine bay (hence the Burning Mystery...). The fire ruined the driver's side plug wires, some vacuum lines, all of the fuel lines, several portions of the engine wiring harness, and the driver's side boot on the rack-and-pinion steering. No hood damage or paint damage due to the fire (hood must have been open), and the fire seems to have been extinguished pretty quickly (lots of fire extinguisher residue). After charging the battery (yep, it held a charge!), I was able to turn the engine over to see if there was any sort of major mechanical issue. Sounded just fine. It was in such good condition that I just couldn't bring myself to part it out, so I decided to repair it and put it back on the road. Long story short, I've got it running and it has passed the AZ state inspection and now has a Rebuilt brand on the title rather than Salvage. However, ever since I got it running, I've had an issue with misfires (particularly bad on Cylinder 2), and with the Bank 2 Long Term Fuel Trim being extremely high (20-25%). It is obviously running very rich. I've scoured this site as well as Shoebox's site for useful information (both of which were amazingly helpful as I was repairing the car, thank you all!), but I haven't been able to figure out what is causing the misfire or the high fuel trims. I think they are related, but I can't figure out if one is causing the other or vice versa. I'm hoping I can pull on all your combined experience to help me get these last two issues resolved.

A few more details on the car: It is a 1996 Camaro SS with the LT1 and T-56 6-speed manual. I also has a Lingenfelter performance package which included engine work to make it a 383 stroker. It also has Edelbrock shorty headers. Right now, it has just over 64,000 miles on the odometer. Documents I got from Lingenfelter say it has a 74211 cam, billet connecting rods, and the 383 crankshaft. When I got the car, it had an MSD 6AL2 ignition controller on it. I've removed that and am using all stock ignition components (see below for details). When I got it, it also had the Bank 2 Catalytic Converter removed and replaced with a straight pipe in its place. The driver's side cat was still in place (although I'm suspicious that it may have been hollowed out due to the Bank 1 downstream O2 readings, see below). It had no fuel tank and all of the nylon lines at the front and back of the car were burned to a crisp.

Ok, here is a list of what I have replaced with all new parts. All parts are ACDelco, original GM, or as close as I could get at Jegs, Summit, Rock Auto, and Amazon. I didn't use any cheap imported garbage.

1. All new nylon fuel lines
2. fuel tank (oh yeah, the entire fuel tank was missing. and that was really fun to replace without a lift...)
3. fuel pump/sending unit
4. fuel filter
5. Evap lines (I did not change or replace the evap canister. It did not appear to be damaged)
6. Plug wires (changeg them all, though only the drivers side were burned)
7. Plugs (NGK TR55GP, gapped to .050. I changed all 8 of them as well)
8. steering rack (replaced the whole thing)
9. Clutch actuation assembly (master cylinder, hydraulic line, and slave cylinder all came as a pre-bled assembly. The fire caused a leak in the line between the master and slave)
10. new Summit brand Optispark
11. Ignition Control Module
12. Ignition Coil
13. Temperature Sensor (the one in the drivers side head that goes to the gauge cluster. It's the single wire type, and the wire had been burned off).
14. Bank 2 Catalytic Converter (welded in)
15. Masterflow exhaust (clamps, for now)
16. Bank 2 Sensor 1 (upstream) oxygen sensor - The voltage readings from the original sensor seemed ok, but I was hoping this was the source of the fuel trim issue. After I installed this new sensor, there was no change in the fuel trim behavior or the misfires. Bummer.

I'm using Torque Pro to log data from the ECM. I can usually drive the car around for anywhere between 30 minutes to an hour before it will throw a P0302 code for misfires on Cylinder 2. The misfire history shows that cylinder 2 will rack up around 3000 misfires in that time. There are misfires on other cylinders as well, but nothing close to what I am seeing on number 2. I've also noted that the misfire count tends to increase the most at low engine load (cruising/coasting downhill). In addition, anytime I let off the throttle the Bank 2 Long Term Fuel Trim will jump up to the max value of +25%. While under load, it is closer to the Bank 1 LTFT at around 12%, but on average Bank 2 is always higher than Bank 1.

Here are a few of the things that I have done to try and find out what is going on with the misfire:

1. Removed and cleaned all of the fuel injectors. No leaking injectors, and all seem to flow the same. (DIY cleaning and flow test)
2. Swapped injectors around between cylinders. Had no effect, cylinder 2 still had the misfire issue before and after swapping the injectors around.
3. Inspected all vacuum lines and searched for any vacuum leaks. Didn't find any.
4. Checked for exhaust leaks. No audible exhaust leaks or tell-tale black smudges indicating any leaks
5. With the engine idling (but cold, I don't need another engine fire...), I sprayed brake clean sequentially around the cylinder heads and the intake manifold. No noticeable increase in RPM's to indicate an intake or vacuum leak.
6. Disconnected the evap line at the throttle body and plugged the evap inlet on the throttle body, to see if anything was happening with the evap system that could be causing these issues. No change in behavior.
7. Reset the IAC position. This didn't change the misfire or fuel trims but did cause a really hard start and rough idle. I removed the throttle body and cleaned out the idle air passages (very dirty), as well as the IAC itself, and that seems to have fixed the hard start and rough idle that happened when I reset the IAC position.

I have attached my most recent log file, as well as a few pics of the spark plugs that I pulled to inspect after only about 100 miles of driving. If there is anyone willing to take a look at the log file, I'd appreciate it. I'm not new to technical data, as I am an aerospace engineer, but I am new to diving this deep into cars and OBD2 data.

Apologies for the length of this post, just wanted to get down everything that I have already done to save on follow-up questions. I really appreciate any help you all can provide!

Injuneer

Late at night so I just skimmed thru the long post. Just curious - have you checked the fuel pressure, and the fuel pressure regulator? The FPR could have been affected by a driver side fire.

You indicated you thought the driver side cat had been gutted, based on Bank 1/Sensor 2 data, but never covered that later. Or maybe I just missed it.

I believe the TR55GP's are gapped at 0.060” (1.5mm) from NGK. Did you regap to 0.050”?

pepehogge

Thank you for the quick reply!

I have not yet checked the fuel pressure regulator with a pressure tester. I did replace the vacuum line that goes to the FPR from the intake manifold. The original vacuum line was damaged in the fire. I've been holding off on buying the fuel pressure tester to save a couple $$, but I will get one and test it out.

Yes! I neglected to add the additional data for the O2 sensors. I'll put that here in volume 2 of this post... I should say that I am speculating that the drivers side cat was gutted. I looked back at the O2 data (plots below), it actually looks like Bank 2 Sensor 2 may be the one that has an issue, and Bank 1 Sensor 2 actually looks ok. And that would indicate that the driver's side cat may actually be intact and functioning. I can't really wrap my brain around why someone would cut out only one cat though. That just seems wrong! Anyway, I collected some data on one of my other vehicles (that is running correctly!), and the Bank 1 Sensor 2 data from that vehicle is similar to what I am seeing on the Camaro. So my initial speculation seems to be wrong, and I think I may need to replace Bank 2 Sensor 2. It is my understanding that the downstream sensor are not used for fuel management, so while I will likely do that it seems unlikely to help with my misfires or high fuel trims. Also, if I am way off base with my understanding of how the O2 sensors work or what this data really tells me, don't be afraid to let me know! Looking to learn.

To answer your third question, yes I did gap the TR55GP's down to .050". Honestly, I just searched for the stock gap setting for a '96 Camaro with an LT1 and it came back as 0.050", so that's what I used when I installed the new plugs. Seems like that could be a potential issue, and I have not tried swapping plugs around to see if the misfire follows the plugs. Might try that one night this week to see what happens.


Bank 1 to Bank 2 O2 data comparison

Sensor 1 to Sensor 2 O2 data comparison

Injuneer

You don't need anything fancy to check the fuel pressure. A simple gauge with a fairly long hose (if taping to windshield to observe pressure while driving), and the correct -4AN connector with a pin to depress the valve core is all you need. Pull the vacuum hose off the intake manifold nipple and check for wet fuel or a very strong fuel odor. That would indicate a leaking pressure regulator diaphragm.

To check the fuel pressure attach the gauge to the Schrader valve, turn the key to "RUN" (not START) and the pump will prime for 2 seconds. Observe the pressure when the pump shuts down, and observe how quickly the pressure drops. Should reach at least 40 PSI, and drop slowly. Start engine, remove vacuum line from regulator (hold finger over end of line to avoid a vacuum leak). Pressure spec is 43.5 PSI (3 bar). GM accepts anything in the range of 41-47 PSI. Reconnect vacuum line to regulator, pressure should drop by 6-8 PSI with that cam.

0.050" is correct for the plugs. Just wanted to make sure you didn't leave them at 0.060". #8 is very clean compared to the others.

Correct, after-cat O2 sensors are primarily there to monitor the condition of the cat. The after-cat sensor voltage should just be a dampened version of the pre-cat sensor readings, with a slight time offset. I don't see that with the data in the log, but the scan rate (1 frame per second) is so slow the inlet vs. outlet comparison is masked. I don't see this issue of importance with regard to the misfires, and high fuel trims.

Is the tune in the PCM the one that Lingenfelter provided? Typically, you should get diagnostic codes for a missing cat, based on after-cat O2 sensor readings. Some people reduce the probability of getting the codes when the cat is deleted by using spark plug spacers (de-fouler) to move the tip of the O2 sensor out of the exhaust stream. Then there were "O2 sensor simulators" used in place of the sensors, but the EPA had tries hard to put the manufacturers out of business. You might want to check for those. But I don't think the cats or lack thereof are a contributor to the fuel trim issue.

Is the exhaust system actually a Magnaflow cat-back?

I'll send you a private message regarding the data log.

pepehogge

Wow, thanks again! Great info here on how to check the fuel pressure. I've got a pressure gauge on the way and will update when I get results later this week.

When I get the car up on jacks to pull the plugs (makes it much easier to get to number 7) and swap them around, I'll inspect all three of the O2 sensors that I haven't changed. I had no idea that O2 simulators existed, so that is interesting. I'll also be looking again for any exhaust leaks, just to double check.

I came across the scan rate setting in Torque the other day and it is set to 1 sample per second, so I'll change that to something closer to 10 Hz and see if my adapter can handle it. If not, I'll get a USB adapter instead of the bluetooth I currently have. Being an engineer, more data is always better!

As far as the tune, unfortunately I don't know if it has been changed since the Lingenfelter work was done. I don't have tuning software. Did some quick research, and it looks like it would set me back at least $400 to get into Jet DST or LT1Edit, so I may see if I can find a local tuner to just read what is currently in there for a tune and check on what the heck is going on with the O2 sensors. Or I might justify getting JetDST because I have two other projects waiting in the wings which would also be compatible with it. Decisions...

I also need to fire my editor (oh yeah, that's me...). The exhaust I put on the car is a FlowMaster (transposed that in my original post). It's the 17133 cat-back exhaust, and I got it from Jegs. Thanks for getting me to clarify that one, I had to go back and look at my order history to sort it out!

Injuneer

The O2 sims are only for the after-cat positions. And as noted they are illegal. Early versions were offered by Casper Electronics, but the EPA fined them $70,000 and stopped production. I suspect the market void has been filled by a Chinese source. But not important at this stage of the investigation.

Upping the scan rate would definitely make the log more useful. Don't worry about the "tune" at this point. The critical issue would be whether the programmed injector flow rate and offsets match the injectors that are actually installed.

No problem with the exhaust info.... I'm just a nit-picky engineer I sent you a private message.

GaryDoug

Torque Pro add-on is great, but only for users dedicated enough to learn how to use it effectively. Obviously you are one of those gifted people. Good job.
I use it on my 98TA to monitor AC pressure so I know when to recharge (slow leak).

pepehogge

Thanks GaryDoug. Being an engineer who plays with software, algorithms, and telemetry data all day long, I couldn't help myself when it came to trying to get all the possible data I could out of Torque.

On the original topic here, my fuel pressure gauge arrived today. Weekdays are pretty busy, so I probably won't get to swapping plugs around and testing the fuel pressure until this weekend, but I'll post an update when I get it done.

GaryDoug

Another side talk: Since you are using the Torque Pro add-on, you may benefit from knowing the data for setting up other PID's for your engine. Let me know if that interests you. Description, range, PID.

Injuneer

@GaryDoug

Gary:

Is "Fuel Trim Cell" available in OBD-2? It's in OBD-1, but not in your OBD-2 program. The factory service manual shows it as data available to the GM scan tool.

Ditto with "Loop Status".

Torque Pro also has a column labeled "H2OS Sensor (mV)". Is this a typo, or is there a sensor measuring water content of something? There are four columns for the pre-cat and post cat heated O2 sensors, so not sure what "H2OS Sensor is. Not mentioned in the 1996 FSM that I could find.

Thanks,
Fred

GaryDoug

See the 96 file I attached above. It is from a commercial OBD2 LT1 scan app and it does have this:
"Fuel Trim Cell (Cell) [0 - 255] 1190"
where 1190 is the PID.
My app only has the mandated parameters, not any extended ones except ones you can add in the last tab.

I don't find the H20S sensor PID in Torque. Where do you see it? Mine may be an old copy.

Injuneer

If you open the file from post #1 with Excel, it is the header for column “F”. Sure looks like O2 sensor data, but for what? There are four columns for pre-cat and post-cat HO2 sensors.

GaryDoug

OK, I see it. That parameter is in the set of GM parameters that comes with the TorqueScan add-on for Torque. It is an extended PID, not a mandated PID unique to GM. I don't know anything about that PID's function but that list is for all GM vehicles, not just the LT1. It does seem to be returning some data though.

Injuneer

“A” appears to be the value read from an 8-bit binary field (0-255), In the case of the IAC for example, it’s clear there is no conversion (scalar?) required for the scanner to display the value. The range of the IAC counts for example is 0-255. Other 8-bit values require a more complex conversion, like the ignition voltage. The range is 0-20 V, so the binary value is divided by 10 to get the display value in range. A couple of the others don’t work out that clearly.

The H2OS (assuming it’s actually on O2 sensor) range is 0-1000 mV, so that Formula allows it to display a range of 0-1106 mV. Now we just have to figure out what that sensor represents. As usual I still don’t have a clue.

GaryDoug

Right. A is the 8 bit value read. Longer returned messages have subsequent letters B ,C ,D for each byte. You can see that fuel tank pressure returns 2 bytes. FYI, the digit 22 (for extended mode) is added to the front of the string with the PID (1145) following making it "221145". Torque takes care of the rest of the required prefix if you set the prefix to "Auto" in the PID definition in Torque.