Jeanius's data log - need some help

[Injuneer]

If you go to this thread, you will see Jeanius is having a problem with a dramatic drop in gas mileage.

http://www.f-body.com/forum/showthre...5&page=2&pp=15

He sent me a data log, and it was a very good scan.... 16-18 frames per seconds. 6 minutes long, with idle, part load driving and a couple WOT blasts. Almost everything was "by the book"..... very "normal".

Things that were good:

MAP - high steady vacuum at idle (31-32kPa), very little pressure loss at WOT (99-101kPa on a BAR of 103)

TPS - 0.59V at idle, 4.55V at 100% TPP

IAC - 31-38 counts at idle, increased as high as 82 at cruise

VOLTS - 12.7-13.3V, but an interesting 0.1V jump every time knock counts jumped up, triggering up to 10deg retard

MAF - 7.4gps at idle, 235gps at WOT/5100rpm. There was a weird value in record #2167 indicating 248gps at only 3775rpm, as the throttle was reopening after a shift. Why?????

BLM's - Balanced left to right (but see below)

Pulse Widths - Balanced left to right

Things that were suspect:

BLM's - Pulling fuel in all of the well developed Cells. At idle they were 112/112, indicating its having to pull 12.5% excess fuel to keep it at 14.7:1. This causes problems at WOT, since the PCM ignores negative fuel corrections and holds the BLM"s at 128. Something is causing it to run rich, and the BLM's are having to pull out a lot of fuel to keep it in control.

O2's - At WOT, the passenger side is always a bit richer than the driver's side, even though the BLM's are balanced and the pulse widths are balanced. This may be problem of mismatched O2 sensors, or the sensors not operating at the same temperature.

Coolant Temp - Generally high, at 210-217degF. May be contributing to knock retard

IAT - Generally on the high side at 98-103degF. May be heat soak of the sensor.

Knock Retard - A significant problem. Every time he opens the throttle, there is a small amount of knock retard.... generally less than 1deg. But at WOT, it will start by pulling a degree or two, and suddenly the knock count jumps up, and it pulls anywhere from 5deg-10deg of timing. When the knock count jumps up, the system volts consistantly jump 0.1V. Any thoughts on this one? Not every WOT blast is accompanied by this sudden jump in knock counts/retard.

Offhand, I don't see anything that is causing it to experience a sudden drop in fuel mileage. Most signs are healthy. The things that might need to be looked at:

Why do the BLM's have to pull so much fuel?
Is the coolant temp contributing to knock?
Why the rogue reading on the MAF?
Why is the passenger side richer than the driver's side at WOT?
Why the constant small amounts of knock retard?

Just curious.... is your EGR system still installed and working?

Leaking injectors can cause a couple of the symptoms..... obvioulsy it would hurt gas mileage. And leaking injectors can cause the "less than 128" BLM's because the excess fuel in the exhaust consumes some of the residual O2 in the exhaust, causing the O2 sensors to report a "false" rich condition.

Something I also need help on from a conceptual point of view: What generates "knock counts"? How are knock counts related to knock retard? I have reviewed hundreds data logs, and I have seen increasing knock counts that do not cause knock retard, knock retard when knock counts are not increasing, and just about every other possible combination.

How does the PCM arrive at knock counts? Is it calculated from conditions that are known to promote knock, like certain combinations of MAP, coolant temp, throttle postion, RPM, etc? It does not seem to be related to specific knock sensor activity.

Does anyone know how "knock counts" are determined?

[TraceZ]

Fuel pressure regulator?

If the diaphram had a pinhole, causing a leak into the vacuum line that would allow extra fuel in the motor.

Or, if the fuel pressure is to high, that would also cause a rich condition.


The amout of fuel entering the motor is determined by injector size, fuel pressure and pulsewidth. If the pulsewidth and injector size are ok, then it must be the pressure that is off, or a leak in the system somewhere.

[Kevin - Blown 95 TA]

The only thing I noticed, and it might be my imagination is that the Right O2 sensor seems slower than the Left one - just curious if the one he replaced is the one on the left? They are both working though. Also with knock counts, the PCM can pull timing based on sudden MAF changes and not increase the knock count and the knock count can increase but not pull timing if it doesn't meet a certain threshold criteria. The short scan doesn't show me much that's wrong, but it's could even be something like plugs & wires misfiring and causing unburned fuel to trick the computer and triggering knock counts.

[Joe 1320]

Hmmm..... jst a few thoughts. Knock can be caused by several things, one of which is fuel. Both lean and rich. Just a gut feeling, but I suspect something in the ignition end. That little voltage fluctuation could be a function of ignition breakup. I had the driver module harness connection get loose on me. When the contacts were intermittant, it behaved just like this.

It's worth a look see into the ignition system as a whole.

Also..... assume there is no mechanical condition like leaky head gasket, valve or ring issues?

[Al 96 Ram Air T/A]

What about a clogged catalytic converter? Not sure how it would manifest itself in a scan, but it sure would require the engine to work harder (i.e. burn more fuel) if it were bad...wouldn't it?

Just a thought...I'm really not as proficient on the OBD2 software as I want to be.

[fastTa]

Fred, if you could, send me the data log, and I will take a look at it.

kgrubbs@satx.rr.com

[Jeanius]

Fred- Once again thanks for your help, My EGR system is still installed and as far as I know, working

Kevin Blown 95 - I did replace the drivers o2 and not the passenger


This started happening soon after I did the Intake Manifold Leak fix, When I put everything back together there was some white on the injectors showing where they slip into the manifold , I thought it might just be because there was dirt covering the white before but the rail is screwed in all the way and I pushed each injector down to make sure and would go in any further. Maybe this is causing them to leak?



The car was throwing the SES light 3 months ago and I had the codes pulled and I got the left o2 bad, and a misfire in cyl 2 and 6, I replaced the o2 hoping that was the problem and havent gotten the SES light since, but the black smoke at WOT and horrible gas milage continues.


Here is the scan if anyone wants to look at it
Scan

[fastTa]

Jeanius(not sure what your first name is), can you send me the data log directly to my email addy as an attachment:

kgrubbs@satx.rr.com

I cannot get your link to work. Common problem with Yahoo.

[Jeanius]

First name is Andy Sorry about the link, anyone else unable to view it? Youve got mail

[Craig 94 TA GT]

What is the duty cycle of your injectors at WOT?

Have you played around with the MAF before?

At worst, do you think he should get a wideband sensor to see the actual A/F ratio?

[Injuneer]

Kevin.... I'll send you the log in .xls format, with the problem areas highlighted.

If you had air leaking into the intake runners via the injector bosses, it would show up as a lean condition and the BLM's would be elevated, having to add fuel. Does look odd the way there is so much white..... although its been 7 years since I replaced my stockers, so I can't remember exactly what they looked like.

[fastTa]

This is how I understand the knock sensors themselves. They are called Piezoelectric sensors. The Piezoelectric effect is essentially an example of how energy is converted from one form to another. Mechanical to electrical in this case. A knock sensor is made up of a polarized stal encased in a sturdy conductive media. When a polarized stal is compressed, it generates a specific voltage by means of a specific mechanical deformation. The greater the magnitude of the knock, the more the polarized stal is compressed and therefore the greater the electrical current generated. The higher the voltage generated, the more significant measures the computer takes to correct the situation, more specifically the attack vs. RPM table within the PCM. The conductive media directs the A/C current first to a rectifier in order to convert the AC voltage to DC. The the DC signal is directed to the knock module. The knock module is a signal conditioner(filter) that attempts to eliminate false knock......"attempts"

The knock counts themselves and their affect and relationship to the act of retarding timing are still a bit of a mystery to me as well. Obviously knock counts are cumulative. I have seen hundreds of data logs that show significant increases in knock counts but do not retard timing at all. Consequently, I have seen many data logs where the PCM is retarding timing significantly without any appreciation of knock counts.

What the PCM and/or knock module determine to be a measurement that is commensurable in going forth and appreciating the knock count I believe has something to do with the ability of the knock sensor to not only "sense" the periodicity and and frequency signature, but the magnitude of a particular "sensed" knock event. I can only assume that a "valid" knock event worthy of appreciating the knock count must be very close to 6400Hz and suffice the algorithms that are reportedly responsible for ultimately deciphering knock. However, I suspect for the PCM to decide to retard timing based directly upon an appreciation and specific number of knock count, there must be a knock event of a minimum magnitude to base the retardation of timing directly from a valid knock count recording.

An interesting note is that within LT1 Edit, you can alter several knock sensing tables and/or parameters. Such as attack vs RPM, decay vs. RPM(the stock decay settings are VERY slow), minimun coolant temp to enable retard, burst knock, low octane retard, number of knock count to enable, voltage threshold. This encourages my thoughts on the PCM not just looking for the periodicity and frequency signatures, but also the specific magnitude of a specifically recorded knock event.

The first thing I noticed on Andy's data log was the long term negative fuel corrections, consistantly higher o2 voltage(richer) on bank 2 during PE, high coolant temp, and especially the lower than normal timing advance during trasition from part throttle to PE.

Next thing I noticed was that almost on cue, every time he cracked the throttle open in excess of 10%, the PCM was retarding timing. I noticed serval spots on the log where the instant the knock counts were appreciated, the amount of retard increased significantly(8.9 degrees in one case).

I saw in most cases of appreciated knock counts, the battery voltage did increase 0.1 volts, but there were also some instances where the battery voltage increased at the same time of an appreciating knock count. I would chalk this one up to coincidence.

Rogue MAF readings aren't all that common, but they aren't rare either. I chalk up rogue MAF flow readings to the following. If you were just so fortunate to happen to crack open the throttle blades at the exact time that an pulse within the plenum had just bounced off the back of the throttle blades, theoretically there could be enough off a low pressure zone right behind the TB to create am instant "draw" of air. If this were to be a reality, it would most likely occur after #3 IVC.

Enough blabble.....I think Andy is getting significant spark scatter. Alot of the tell tale signs are present to indicate spark scatter. lower than normal spark advance at low load, low RPM, and low TPS. I also saw the occasional operation in Cell 18 during low load, RPM, and TPS as well. Another indicator of possible spark scatter.

Now another reason why I think Andy is getting spark scatter is because spark scatter can be a significant contributor to detonation. I believe the retardation of timing that you are seeing is a result of real knock. Typically detonation will leave a significant amount of unburned end gas and will consequently be allowed to exit the CC into the exhaust. Typically a period of light detonation will cause all of the available oxygen in the CC to be used up. The oxygen sensors will normally report this as a rich condition since much if not all of the available oxygen has been consumed by the remaining unburned end gas.

This problem can be even further exacerbated because detonation can drastically reduce EGT's. The lower the O2 sensor is out of it's effective temperature range, the less effective it is at "sensing" oxygen. Now whether or not the EGT's are being reduced low enough or not to hinder the bank 2 primary 02 sensor in any way I do not know. An EGT measurement would be the only way to prove that.

I wondered at first if the detonation were serious enough and consequently lowering the EGT's enough, if the the PCM could be powering on the O2 heater to bring it back to above it's minimum opertating temp. I also wondered if this could be related to the adventitious 0.1 drop in battery voltage. But this seems to be a stretch at best.

Faulty EGR systems can also create and/or exacerbate detonation.

This is just my opinion and my fingers are tired..............

[Joe 1320]

....I think Andy is getting significant spark scatter. Alot of the tell tale signs are present to indicate spark scatter. lower than normal spark advance at low load, low RPM, and low TPS. I also saw the occasional operation in Cell 18 during low load, RPM, and TPS as well. Another indicator of possible spark scatter.

Now another reason why I think Andy is getting spark scatter is because spark scatter can be a significant contributor to detonation. I believe the retardation of timing that you are seeing is a result of real knock. Typically detonation will leave a significant amount of unburned end gas and will consequently be allowed to exit the CC into the exhaust. Typically a period of light detonation will cause all of the available oxygen in the CC to be used up. The oxygen sensors will normally report this as a rich condition since much if not all of the available oxygen has been consumed by the remaining unburned end gas.

This problem can be even further exacerbated because detonation can drastically reduce EGT's. The lower the O2 sensor is out of it's effective temperature range, the less effective it is at "sensing" oxygen. Now whether or not the EGT's are being reduced low enough or not to hinder the bank 2 primary 02 sensor in any way I do not know. An EGT measurement would be the only way to prove that.

Agreed...........As I suggested in my previous post, the ignition system needs to be looked at.

[fastTa]

Agreed...........As I suggested in my previous post, the ignition system needs to be looked at.

Unfortunately the only true way to prove scatter is occuring is with a very high speed dual scope that can simultaneously monitor both the hi-res and low-res signals and not many people have access to such a scope.

It is always a bit difficult to definitively pinpoint a problem simply by looking at a data log, but it is a good tool to disect certain problematic relationships if you take the data log for what it is.

I think Jeanius is having a combination of problems that are possibly inducing detonation. High coolant temp, high IAT readings, a few weird fluxuations with the BLM cells, possible CCP malfunctioning, possible EGR malfunctioning(sticking valve?), and what I see as some possible spark scatter.

The best thing to do is just start ruling culprits out and trying to fix the known problems such as high coolant temps and high IAT readings. The IAT readings may be high as a result of a warmer than normal motor.

[Kevin - Blown 95 TA]

I think a lot of problems do end up being basic. That doesn't mean easy to find, though. I once had a guy ask me to tune his car - he had already changed plugs and it was running like crap. He wouldn't let me touch the plugs until, after wasting a bunch of time trying to get it to run, I insisted on checking them, and one was completely smashed with the ground into the center electrode another was completely fouled out.