tylerwerrin

Yeah, I swapped the fuel rails with the 96 fuel rails & 96 fuel injectors. From what I recall, the 93 intake was really different from the 96 one - I don't think the 93 fuel rail crossover pipe would even fit on the 96 intake.

tylerwerrin

I wonder - perhaps it has something to do with Speed (MPH) vs. Engine RPM - Since my rear end on the Jag is 2.88, and the Camaro had 3.23's, could that be causing the PCM to make it's RPM calculations/TPS position calculations based on in congruent information?

I know I had to change the rear end information for the Speedometer in TunerCats... but I don't recall changing any other values on the tune. Just throwing it out there as an idea.

Injuneer

The PCM gets RPM from the optical module in the Opti. At startup, the PCM reads the closed throttle TPS voltage and sets that as 0% TPP. It interpolates between that 0 point volts and a point ~4 volts higher as 100%. Neither of these values has anything to do with the rear axle ratio. That only affects the speedometer input.

I did find some unusual data at one point in your latest log. The throttle % was increasing, the MPH was increasing, but engine RPM was dropping. There was no indication of a lifting throttle gear change. The PCM remained in the same cell. At the same time, both left and right O2 sensors were reading extremely lean (less than.100 mV), but the LTFT's were dropping rapidly from 156 to the 123’s. Exactly opposite of the expected response. The STFT’s started increasing from 133 to the 150’s, as they should, but accomplished nothing, because LTFT's were dropping. So the injector pulse widths remained relatively unchanged, and the O2’s remained lean. Makes no sense at all. Nothing I have ever seen before.

tylerwerrin

What position # did you find that bit of data? I could probably tell you what I was doing at that time (driving-wise) - I drive the same loop in the country, usually a period with cruise control on, which has a slight hill - I may have disengaged cruise, quickly shifted from 4th to 5th and re-engaged cruise control going back up the hill.

Could a faulty PCV cause this problem? Like a PCV that doesn't fully close back up (intermittently stuck)?

I have a couple spare PCM's - I could try a different one and see if anything changes.

Injuneer

I don’t have access to the data log right now. I'll give you the range of line numbers in the morning. Everything I posted above was from memory when I was looking at it this afternoon. I do remember the MPH was not that high.... maybe 10 to 20 MPH. I think RPM dropped from 1,300 to 950.

I've been thinking about whether the PCV valve could be part of the problem.

Injuneer

Starting at line 3650, both O2 readings go lean, and stay there, through line 3676, a period of 3.24-seconds. It goes into Cell 5 at line 3655, and stays there until line 3672. Cell 5 LTFT's drop like a rock, from 156 to 132, while the STFT's climb from 133 to 163. At mid-point, the LTFT's reverse and climb slightly. EVAP DC is increasing from 2 tp 18% and as is logical with the TB hose nipple blocked off, O2 readings are unaffected.

In general, there are just as many cells w/ 160 LTFT's, and high STFT's. And there are some cases where the O2's go lean and STFT's in Cell 6 seem to increase along with EVAP duty cycle. See lines 3949 through 3974. I can't explain it.

ACE1252

I'm an idiot. Didn't read enough to know this wasn't in a F-body.

I just looked over the fuel trim reset log. Seems to me that every time he tips into the throttle, there is always a lean spike from both o2s.....couldn't this be a fueling issue? Clogged fuel filter, fuel pressure regulator, or fuel pump issues? MAF and MAP are tracking with the throttle inputs. I know he mentioned it's smelling rich, but looks to me like the engine is hurting for fuel delivery......fuel leak somewhere? Have you verified that the fuel pressure is good under load(when driving)? Is the car pulling strong at WOT?

Compared to the logs of my Z28, I do not see any kind of abrupt lean response from the o2s when I tip in from 0% throttle. Granted, I also don't have the wonderful data log resolution of an OBD1 system either.

Injuneer

The puzzling part is that the Cell 16 LTFT's are dead on..... 128/128 in one log, 130/129 in another. A drop in fuel pressure at increasing load, due do to a dirty filter might explain that.

The “tip-in” lean out also seemed to be accompanied by in increasing EVAP purge solenoid duty cycle. That’s why we tried to rule that out as causing the increasing of STFT's with increased EVAP DC.. It's something that is affecting both banks equally. Maybe there's a crack in the inlet track that only opens up during increased air flow? But as soon as the throttle blades open, doesn't matter whether it high load, low load, tip-in, steady throttle, whatever, he has 160/160 LTFT's.

While tip-in could lean things out, and bump the STFT's up, it doesn’t seem.to explain why most of the open throttle cell LTFT's are 160/160 under virtually all driving conditions.

It's always good to have another set of eyes looking at the data.... sometimes I feel like an army of one. Any ideas are appreciated.

tylerwerrin

Some additional information on the fuel system:

I'm running an external "in line" fuel pump, a Walbro 255. I have a small pre-pump filter and a larger post-pump filter. I swapped in a new pre-pump filter during the previous tests a few weeks ago. There were no changes. I have a new post-pump filter I can swap in and see if that makes a difference. I can also pick up another cheap ebay external fuel pump - they are only like 20-30 bucks, and it might be worth it to see if there are any changes.

The fuel pressure regulator is relatively new - a couple thousand miles at most. I tested the regulator with a hand held vacuum pump and it holds vacuum fine so I think it's OK, but perhaps it is failing under increased load.

I could try testing the fuel pressure under load, while driving around; idle fuel pressure is fine (steady 35~), but haven't tried taping the gauge to the windshield and cruising around to see if it drops.

Any more thoughts the the PCV becoming stuck?

ACE1252

Can you log a few WOT pulls? Don't have to do anything stupid crazy with high speed. Say from 0-65mph......at least 1st to redline, 2nd to redline, maybe a hint into 3rd if the speed is not too crazy.

Injuneer

Here's a chart I use for analyzing some of these logs:

tylerwerrin

I'll do some WOT pulls once I get back out to the desert next weekend - it sucks having the car so far away, but if I had it here in LA, there's no way i'd be able to do a WOT pull with traffic!

Thanks for this chart - it really helps to visualize just how frequent 160 ltft's are - basically under every "under load" cell, I'm at 160~.

Odd that Cells 2 & 3 are not pegged at 160 as they indicate the highest "loads"recorded, RPM wise, which puts a dent in the low fuel pressure theory.

As for position #3650:
I was just driving straight on an uninterrupted quarter mile stretch of road, got it up to 4th gear, made a right turn onto a 2 mile stretch of uninterrupted road and set it to cruise control, stopped, turned around and drove back home, with cruise control on, first in 4th and then dropped it down to 5th at Position#5682.

Injuneer

When I bought my Formula new in 1994, I lived in Santa Clarita. It was easy to find an industrial park that had little if any traffic at night. A pull in 1st gear can be just as revealing as one in a higher gear.


My thought with regard to the chart is that if the engine was running out of fuel at high loads, the distribution would range from close to 128 in the lower right corner and much higher numbers in the upper left corner - a diagonal line. That's not what you have. I have a chart form an engine that was running out of fuel - see attachment.

With regard to fuel use, it's a combination of MAP + RPM.. High MAP increases the density of the air in the cylinder. RPM increases the number of cylinders of air being used.


MAP is much more indicative of engine load than RPM. That's why the spark table, for example uses MAP as the measure of load, and RPM as the measure of how much you need to advance the spark. So..... low MAP (under 30 kPa) is very low load. You will see 35 kPa at idle. So those cells are mostly used in deceleration, at low throttle openings - no load at all.


I'm not sure what to make of that block of data. I just don't understand why the PCM was dropping the LTFT's 5 counts at a time, in a fraction of a second.

tylerwerrin

I didn't have a chance to do any WOT pulls - but I did get a chance to check a couple other things - two of which I believe is pointing to the culprit: fuel starvation.

The fuel line which goes from the tank to the pump had a kink in it; more accurately, there are 2 rubber hoses which connect to an intermediary metal line, and the rubber hose going from the tank was the one which was kinked (out of view, obscured by an access panel). I always noticed that the pump ran loud and "lopey", but I assumed it was just a reality of external fuel pumps with no rubber isolator. Now, I believe that the pump was being starved for fuel.

I checked the fuel pressure at idle, which held at a steady level of around 40psi... while revving, it would drop to around 30. Then as another test, I adjusted the kink in the fuel line, and noticed that the fuel pressure would go down, and then recover back up to 39. It would get down to 20psi if I played with it enough, and I even got it to run down to 0psi a couple times, stalling the engine out. Perhaps the hard pulls were causing the kink to bind up more, reducing fuel pressure.

As a precautionary measure, I checked the secondary post-pump filter and noticed that on the inlet side, when the gas dribbled out, it was dark and sludgey - the outlet side was clear, so the filter appeared to be doing its job, but had become dirty. I replaced this filter with a new one.

To make sure I don't clog the filter again, I'm going to pull the metal line which runs from the outlet of the pump to the filter and clean it out - perhaps some varnish or other sludge has built up in the line.

I had to head back to LA for work today, so I won't get to run any additional tests until I have more time - I know this runs contrary to what the data has been indicating, but at the very least this is one problem that has been identified and confirmed.

tylerwerrin

I finally have had a chance to go through the fuel system nearly completely - the only thing I haven't replaced (yet) is another fuel pump.

As I noted in the previous post, the pump was acting lopey and I discovered a kink in the line (hidden from view). I replaced the line completely which now runs from Tank > Prepump filter > Pump > post pump filter > high pressure line > fuel regulator (previously there was a piece of steel line which bridged the gap between the tank and the wall of the trunk). I had noticed that the pre-pump filter had little bits/chunks of black materials stuck at the inlet baffle on the filter (it's a Wix 33299), so I ran the car several times until the black stuff went away. I suspect it was the old kinked fuel hose which had decayed and backwashed back into the tank and had hoped that might have been the end of it.

The fuel injection hose is all new, the filters are all new.

Yet... the weird fuel trim numbers remain. Cell 6 and 11 are the biggest offenders. If any one has any additional thoughts or avenues of pursuit, please help!