Analyzing log files

[Mr. P.]

Elaborating on everything Joecar said -

In forced-induction applications, you'll learn that ultimatelty it's all about the capability of the fuel you choose to run; you are fuel-limited. Even guys that want *insane* torque levels figure out that it's dang near impossible to make 1500-hp on a forced-induction 6.0-litre motor with pump gas (because pump fuel will not tolerate cylinder pressure i.e. dynamic compression ratio levels that high), and they upsize to 7.0-litres because the increased chamber size/displacement allows them to achieve those power levels via a lower, more manageable dynamic compression ratio, one that pump fuel will tolerate. The fuel you run dictates everything, IMO.

From EFI-101 class:

"Stoich" (pronounced 'Sto-Ick') is slang for "stoichiometric ratio" (pronounced 'Sto-kee-oh-metric'), which from high school chemistry class is defined as the exact amount of fuel, plus the exact amount of oxidizer (air) needed to have chemically perfect combustion, i.e. no leftover emissions, there is exactly present the correct amount of fuel for each molecule of oxygen.

All fuels are chemically different, and will have different stochiometric ratios; in high school autoshop I was taught that gasoline has a stochiometric ratio of 14.63:1, meaning that for every 1 "part" of gasoline you must provide 14.63 "parts" of air to achieve a chemically perfect combustion. Fuels are VERY forgiving, gasoline especially - it will burn anywhere from 10:1 all the way to over 16:1, if the proportions are anywhere near close it will ignite/explode; but at 10:1 ratio, there will be a lot of leftover unburned fuel, because not enough oxygen molecules were provided!

Fuel refineries work very hard to make sure the products they sell meet these ratios, and different fuel blends will have differing ratios - for education purposes check out VP Fuels website, they publish the stoichiometric ratios on all their custom blends and you will see that C16 not only has a different octane rating but also a different stoich, meaning that a carburator jetting change -OR- change to the PCM's "stoich ratio" setting {B3601} is required. This same adjustment is necessary if you were to change to methanol, or E85, or E10... you gotta know the fuel you are running and it's stoichiometric ratio and plug that number into the {B3601} cell in your tune.

...In other words, E10 requires a richer AFR than E0 does? ... Meaning it takes more fuel using E10 to reach the proper air/fuel mix than it does with E0? ...

YES; significantly so. There is a vocabulary lesson required, the difference between "gasoline" and "fuel" - 40-years ago what came out of the service station nozzle was one-and-the-same, but today you are not purchasing "gasoline", rather you are technically purchasing a "fuel product" of which only 90% of that product is actual gasoline and the rest is 'something else'. This is the explaination I share, it's waaayyy oversimplified but.... Imagine two pie tins, fill one with an inch of mid-1960s "pre-emissions" gasoline and the other with today's E10 pump fuel, then light them both on fire, and feel the warmth from each fire with your hands - what you will observe is that the mid-1960s gasoline gives a MUCH HOTTER fire than today's fuel! After 30-years of EPA laws, fuel refiners have been forced to blend fuels having less and less heat energy, meaning that engine power outputs go down because the chemical energy just isn't there in the fuel to begin with; not only has the fuel octane gone down (that's been very obvious to the general public for decades) but more importantly the actual potential energy IN the fuel has been weakened over time. A large part of why 80's vehicle performance sucked so bad has a lot to do with the OEM's struggling to deal with the fuels available on the market -in addition to- congressional emissions laws, CAFE requirements, etc. E10 pump fuel is only 90% gasoline, and 10% FILLER! IMO, if the EPA could have realistically gotten away with mandating 10% water in pump fuel they would have happily done that instead (ethanol is the next safest thing, in their minds)!!! What this means for us throttle junkies is: fortunately there still remains "gasoline" in E10 pump fuel, you just gotta squirt a lot more of it into the engine than you would have 40-years ago to make power, because 10% of today's fuel is just "filler" - it has no combustion benefit that I can see.

Another common misconception: that E10 is related to E85, or simply just "E85 light" - as a fuel you cannot compare E10 to E85, they're totally different animals with totally different combustion characteristics. E85 fuel has an attractively high octane, but it burns very cool (remember the pie tins?) and you have to squirt a sh!t-ton of it down the intake port to make power. Meaning: purchasing a bigger fuel pump, bigger fuel injectors, bigger fuel plumbing, bigger fuel comsumption = MORE COST. These are my opinions and impressions of E85 fuel, do your own research, I came to the conclusion that yes a guy can make big power with E85 but the costs are prohibitive (compared to achieving the same power potential with E10 pump fuel, or E0 gasoline).

... I'm presuming that this is one of the constant values you need to have correct before you even start with the tuning? Like the IFR table?

YES; table {B3601}.

... the guy who tuned my car tuned it for a commanded AFR of 14.63. So basically unless you find some non traditional source of E0 then the tune should actually be for 14.13? ...

I've seen a few different fueling strategies, and at the end of the day as long as the engine gets what it needs then I guess all is well enough; some guys calibrate like Dephi does, they keep the stoich unchanged at 14.63:1 while increasing the commanded AFR under high load conditions, and they make it work. Personally, my philosophy is to not "lie" to the PCM, or it will not know how to accurately fuel the motor! It is my suggestion that whatever value is input into {B3601} matches the fuel you are actually injecting through the rails; make the computer work FOR you, rather than adopting a mindset of "workaround tuning".

... So, again, what was E10 supposed to gain for us? Looks to me that gas mileage will take a hit. And I hear they want to go to E15 now? ...

It's a California thang - 30-years ago, vehicle makers by themselves could not make engines running clean enough for the State of California; as a result, the California Air Resources Board (CARB) figured that they could immediately get a 10% improvement in air quality by geting all the cars in the entire state to burn 10% less gasoline by simply watering that gasoline down 10% with ethanol - the 10% figure was determined as the most amount of ethanol they could get into the fuel without citizens popping motors everywhere! CARB knew that everyone's cars would run lean on this new emissions fuel blend, that was not important to them - their goal was to get citizens to burn 10% less gas, hence putting 10% less crap into the air. So laws were passed in California requiring all fuel sold to contain 10% ethanol, and the result is that the fleet of vehicles in the state burned 10% less gasoline (because 10% of the fuel wasn't gasoline!) and air quality emissions DID improve. However the costs were increased fuel prices at the pump (California fuel is unique, and expensive), loss of MPG, and loss of torque.

Because of the success of California fuels in fighting emissions, the EPA now mandates E10 at the pump in all 50-states (as of last year). The OEMs have adapted to the problem with FlexFuel systems, so running lean isn't an issue now because the fuel line tells the PCM in real-time exactly the ethanol content in the fuel and the stoich value is now always varying according to what's in the tank - regardless of whether it's E0, E10, E50, E85, or even E100. But, you still cannot get away from the physics that ethanol burns much cooler than gasoline (ie has less energy density, as a fuel) so MPG takes a hit. And torque still suffers unless one uses their right foot to increase the amount of both air & fuel going into the combustion chamber... at the end of the day, it's cylinder pressure that makes the piston go down the hole, turn the crank, and make burnout marks - and that comes from MORE HEAT.

E15 - all I can offer there is... government forcing the market to invent a need for more ethanol, to the benefit of ethanol producers? I dunno...

- Steve.

[Rich Z]

Thanks for the informative reply. I have yet to run into any REAL person who wanted their gasoline watered down with ethanol. And quite frankly, using corn, which comes directly out of OUR food chain, is about as dumb of a move as I can imagine our government doing.

But back to the tuning stuff.

So what do the oxygen sensors do with all this? Aren't they still determining the correct ratio of air to fuel based on what they detect in the exhaust? So regardless of what we TELL the PCM the AFR ratio should be, aren't the O2 sensors compensating for what is REALLY being burned in the cylinders? Meaning that as long as the car is running in closed loop, isn't everything just ducky as far as the AFR is concerned? And OK, I'm a bit hazy about this, but does the PCM go into open loop during WOT? Which would mean that this is where the AFR tuning needs to be done to compensate because the O2 sensors are no longer running the show?

So given a choice of tuning at either 14.63 or 14.13 knowing that you may have either E0 or E10 gasoline in your gas tank at any given time, which would be the BEST AFR to tune for?

I plan on using E0 whenever I can, as long as I can find it in 93 octane. But once I stray away from home, this may not be possible to find, so I would HAVE to use the E10 blend. Obviously I don't want to have to carry my laptop and V2 everywhere I go and make a tuning change at the gas station. Of course, neither am I planning to be making WOT runs up and down the streets, neither, so perhaps the oxygen sensors will handle the difference and it will be completely transparent to the operation of the engine?

Thanks for your insight with this. I feel like I just spent the night at a Holiday Inn Express now.

[Rich Z]

In your logs, MAP shows you might have an airleak, or valve timing/compression problems.

It looked very wobbly when there was no reason for it to wobble.

Joe, I scanned through the log files, and darn if I can figure out what you were seeing to figure this out. I watched these variables:

RPM, TP, MAF, and MAP and they all seemed to be playing well together. But as I have mentioned before, I really don't have a clue as to what is really NORMAL in these displays yet. I'm only able to go by what SEEMS normal to me, which might be completely WAY off base.

My engine was build by LME so I'm fairly confident it was put together correctly. That would leave the shops that worked on my car botching up something. Early on I did spray starting fluid around every possible vacuum leak area, and got no signs at all of a vacuum leak. Of course, I never though to check at the wastegates, and no telling how long that vacuum line was loose. So if there is an identifiable and repeatable method to show that I might have a vacuum leak, then I would have something to use as a target knowing that I have to fix whatever it is that is making that display anomaly.

[joecar]

I thought MAP looked too wobbly in Log_0002 the other night, but today it looks more reasonable (lol, I wasn't drinking on either days).

[joecar]

...

So what do the oxygen sensors do with all this? Aren't they still determining the correct ratio of air to fuel based on what they detect in the exhaust? So regardless of what we TELL the PCM the AFR ratio should be, aren't the O2 sensors compensating for what is REALLY being burned in the cylinders? Meaning that as long as the car is running in closed loop, isn't everything just ducky as far as the AFR is concerned? And OK, I'm a bit hazy about this, but does the PCM go into open loop during WOT? Which would mean that this is where the AFR tuning needs to be done to compensate because the O2 sensors are no longer running the show?

So given a choice of tuning at either 14.63 or 14.13 knowing that you may have either E0 or E10 gasoline in your gas tank at any given time, which would be the BEST AFR to tune for?

I plan on using E0 whenever I can, as long as I can find it in 93 octane. But once I stray away from home, this may not be possible to find, so I would HAVE to use the E10 blend. Obviously I don't want to have to carry my laptop and V2 everywhere I go and make a tuning change at the gas station. Of course, neither am I planning to be making WOT runs up and down the streets, neither, so perhaps the oxygen sensors will handle the difference and it will be completely transparent to the operation of the engine?

Thanks for your insight with this. I feel like I just spent the night at a Holiday Inn Express now.

The oxygen sensors report the difference between actual combustion and stoich, which the PCM uses to trim future combustion.

Yes, in closed loop the PCM is trimming to the fuel's stoich, but when you go WOT the PCM adds any positive trims on top of PE fueling... this is why you want to achieve CL trims close to zero... setting the correct value of stoich in the tune allows the PCM to calculate the correct initial fuel (before trimming).

When you got WOT, the PCM does not technically go OL... it still applies the last positive trims, i.e. it is in a partial CL mode where it applying the [positive] trims but is not updating the trims themselves; we did some experiments with intersecting PE and OLFA tables (and the protection modes disabled and/or given gorssly identifiable AFR's), we saw the following:
- OL: going to WOT -> the active tables was PE and OLFA, the PCM selects the richer of those (I saw the AFR jumping between PE and OLFA);
- CL: going to WOT -> the active table was PE only (i.e. OLFA never came active, I saw AFR from PE only);
[ remember that I had the PE and OLFA tables intersecting each other (i.e. to show the richer one coming thru ].


Also note that setting the correct value of stoich in the tune allows the PCM to get OL fueling correct (by allowing it to calculate the correct initial fuel).

[Mr. P.]

...So what do the oxygen sensors do with all this? Aren't they still determining the correct ratio of air to fuel based on what they detect in the exhaust? So regardless of what we TELL the PCM the AFR ratio should be, aren't the O2 sensors compensating for what is REALLY being burned in the cylinders? Meaning that as long as the car is running in closed loop, isn't everything just ducky as far as the AFR is concerned? ...

I view the O2 sensors as a Delphi "workaround", or add-on.

The O2 sensors DO NOT have any DIRECT bearing on fueling calculations; if the PCM calibration/tune is perfect, you don't need O2 sensors at all (!) in fact in my own truck I've been daily driving for YEARS without using the O2s (called "open-loop speed-density" mode) and lately my truck is getting its best MPG ever because I've been repeating the AutoVE process daily for about 4-weeks now and my main VE table is starting to get really dialed-in.

The process basically is:
1. get a current sit-rep by scanning the 'important' engine sensors (IAT, ECT, TPS, MAP, RPM, etc)
2. lookup values from the tables in the tune, and calculate how much fuel to inject into the engine
3. inject fuel, then ignite the air/fuel in the combustion chamber
4. see what the O2s have to say about what just happened

All the O2s do is provide feedback; under the right circumstances, the PCM will consider that accumulated feedback and derive from it an error percentage (called fuel trimming) which it will use the next time it must repeat Step #2. On the surface, this explaination may sound like somehow the O2s are "driving" fueling, but that isn't so - the various tables in the tune drive the fueling, and the O2s are simply there to tell the PCM after-the-fact how good or bad a job it's doing. Also remember, the O2 sensors themselves can only report 1 of 3 "answers" (1) the last combustion event was fueled correctly, (2) the last combustion event was rich, (3) the last combustion event was lean. It's like hooking up your truck to a trailer with your spouse helping you - "...you're too far left... no, now you're too far right... ok drive straight... no, now you're left again..." you really have no idea if your are off an inch or a yard, and no idea how far to countersteer. This is the same sort of hindsight feedback the PCM gets, so the PCM has no idea how FAR off it is with it's fueling, all the PCM knows is that it's last "guess" wasn't right, and to try again next time... (this is why we buy widebands!).

The O2 sensor feedback system was never intended or designed to "drive" the fueling math in the PCM; the reason O2 sensors are present is because after the brand new perfect vehicle rolls off the assembly line, parts begin to wear and the engine will mechanically not be "perfect" anymore which affects volumetric efficiency anywhere from 1-5% (or more?!) - the O2 sensor system is provided because by law the vehicle needs to still run cleanly at 100K-miles, and it's a lot cheaper to fit this workaround on a vehicle than to provide free ongoing dealer service for every car until it is 100K-miles old just so it can pass smog inspection!

O2s are a feedback loop only, the tune/calibration in the PCM still has to be reasonably accurate!!! O2s will NOT take a mediocre tune and "learn" how to cope and make it a badass, race-winning, fully-researched engine calibration.

Another thought - when in closed loop, the PCM purposely fuels the engine "rich-lean-rich-lean-rich-lean........." so it can SEE the O2's reporting this behavior a split-second later; meaning, that in my own truck when in closed loop my city MPG goes from high-14s to mid-12s! In open-loop operation, the PCM keeps fueling steady, the PCM isn't making every other squirt of the injector purposely a tiny bit rich (wasting fuel).

...So given a choice of tuning at either 14.63 or 14.13 knowing that you may have either E0 or E10 gasoline in your gas tank at any given time, which would be the BEST AFR to tune for? ... perhaps the oxygen sensors will handle the difference and it will be completely transparent to the operation of the engine?

OK. Dose of reality here: if the engine runs lean while you are only wanting 65-hp (say, cruising down a flat freeway) and it DOES ping or have some form of 'mal-combustion' there is so little energy involved that you will likely not be harming any engine internals. OTOH, if you are making 750-hp and you have a detonation event, you will likely be snapping a top ring land or driving over your own crankshaft! Add to that the fact that when you put your foot down past a certain point, the PCM will abandon the O2's completely and go on a hope & prayer that the PCM tune is right and fuel the motor completely on math; there is no O2 feedback during WOT, this by definition is going from closed-loop to open-loop. And even if the PCM were to stay in closed-loop during WOT, by the time the O2 sensors had anything important to say about what has just happened in the combustion chamber the damage will already have been done (remember, they are consulted on step #4!). Point being, the math (i.e. your tune!) needs to be as correct as possible.

Typically I suggest that people cave-in to the reality that E10 is now a fact of life, and use a stoich of 14.17; in your case however since you will be running the car 90% of the time E0 then I would go ahead and keep the stoich set at 14.63, and before taking a road trip of any kind I would prepare and change the stoich value to 14.17 the night before (then change it back again after you've returned home and refilled with E0). PITA I know, but it's better than carrying a laptop in the car. Whatcha gonna do... :/

- Steve.

[Rich Z]

I thought MAP looked too wobbly in Log_0002 the other night, but today it looks more reasonable (lol, I wasn't drinking on either days).

Well, THAT is encouraging to hear! I did have to do a lot of vacuum re-plumbing on my car, so certainly anything is possible. I had to install a manifold block so I could hook everything I needed to vaccum. Seems to be working, because I can see fuel pressure changing with manifold vacuum. But the few times I kicked it slightly into boost, I don't remember hearing the blowoff valve at all between shifts. But honestly I've been babying the car while trying to fix all the problems. Now that I've got the wideband hooked up (or will again when I put the exhaust back together on the car) I want to kick a little boost into the engine so I can see what the log files look like. Of course, I still don't have a clue how to set up the PIDs for this custom operating 2 bar MAP I've got. I thought I read somewhere that there are one or more additional tables involved. When my tuner was street tuning the car, he couldn't do much in boost because I had some older tires on the car (after being in shops for a couple of years, they just got hard and not able to hold much traction), and they just broke loose WAY too easily when boost kicked in. He told me he erred on the site of fatness, to be safe, but I would still like to see what the WB is actually seeing.

Oh well, all in due time, I guess.

And Joe, in case I need to do my tune over from SCRATCH, since the fuel pressure regulator is mapped to vaccum AND the IFR table is controlling fueling, is there some sort of a getting started tutorial to let me know how to actually set up the injector info? I do have an Excel spreadsheet that I got from the guy who worked up the injectors, but I don't even have a hint of a clue about how to turn that spreadsheet into something EFILive can digest.

And before anyone asks, no, I don't have access to a dyno. There are a couple around town that I know of, but I don't trust any of them to have my car in their shop. So I'm pretty much going to have to do this literally by the seat of my pants.

Thanks....

[joecar]

Pids: use the calc_pid.txt from the Calc.VET thread.


IFR/FPR technical: Calculating-Injector-Flow-rate


IFR spreadsheet: enter rail pressure (measured with reference removed), injector rated flowrate, injector rated pressure.

[Rich Z]

Pids: use the calc_pid.txt from the Calc.VET thread.


IFR/FPR technical: Calculating-Injector-Flow-rate


IFR spreadsheet: enter rail pressure (measured with reference removed), injector rated flowrate, injector rated pressure.

OK, thanks. I'll have to check my notes. I can't remember what I set the fuel pressure at with the vacuum hose disconnected. And I've got that info on the injectors somewhere....

[Rich Z]

Also, I think it's in the third log that you will see the 6 seconds where my IAT reading went bonkers.

I was planning on taking the car out for a drive today after fixing a bunch of minor exhaust leaks, and I had the laptop on, running EFILive, and was logging the warm up cycle to note how it goes from open loop to closed loop. I noticed immediately that the IAT signal was pegged at 284 degrees and staying there. So I apparently had a SOLID failure to track down. The main reason I was going to take the car out today was to do some logging specifically for this IAT problem. I had cleaned the contacts on the connector thoroughly, so I wanted to see if maybe that corrected the problem. Apparently not. Turns out I was barking up the wrong tree anyway.

So I shut off the car and set up the laptop where I could watch it while I was poking around under the hood with that harness. Connie was actually monitoring the screen for me, and she called out when the signal dropped from 284 to 90 degrees or so while I was jiggling a particular section of the IAT/MAF cable harness. So it looked like we had zeroed in to the problem. Hard failures are SO much easier to find than intermittent crap.

So I pulled the insulation sleeve off of the harness to take a look. And here we go:









Whoever did this piss poor soldering job left jagged ends sticking out from the soldered joints, and didn't bother to even cover the purple wire individually. Just plain laziness. So what happened was that a sharp point on the purple wire's joint had cut right through the insulation on the yellow wire and was shorting the two of them together. There were a couple of other sharp points on the other wires, so they may have been poking through the insulation as well. If not now, certainly sometime in the future.

Two things really wrong with this solder job. (1) Before putting the shrink wrap on the solder joints, smooth the sharp solder points down. This can easily be done with no more than needle nose pliers by crimping them down flat against the wire. Actually wrap the two wire ends together properly, and you rarely have sharp points sticking out every which way. (2) With multiple wires being soldered in a harness, stagger the cuts on the wires so the joints aren't right next to a neighboring joint. This not only helps to prevent this kind of problem with shorts, but also keeps you from having a big LUMP in a harness when you try to put a cover over the entire wiring. You know, looks like a snake that ate a big rat or something.

So I crimped the sharp ends in the solder joints, then rewrapped them all with silicon tape. Double checked the signal to make sure the short was gone, then rewrapped the entire harness with insulation and wire wrapped it back in place along the air bridge.

So I believe the intermittent IAT problem has been solved now as well.

Maybe tomorrow I can take the car out for a spin. I didn't hook the wideband controller up again, as I want to burn off any anti-seize compound that might be in the exhaust pipes above the sensor first.