Help understanding fuel trims

[Shdwdrgn]

I've mentioned this in other threads, but I need some guidance trying to figure out my mileage problem. This has been going on for well over a year now, however I didn't have any solid info until recently. Essentially what is happening is that my gas mileage is virtually identical regardless if I run through a tank out on the highway or in town. Each tank that I've checked has been coming in between 15.2 and 15.8mpg. Last year I went for the easy fix and replaced my front oxygen sensor, however nothing changed. I considered the custom CAI, however the original one failed and I had to go back to stock for a few weeks until I could make a new one (of slightly different design). Again my mileage did not change. Even with the new steel front bumper, the readings are always in that range.

My best guess is that something is 'stuck'... maybe a bad sensor, maybe something else. Last time I checked readings from my ODBII (sitting in the driveway and revving the engine), I remember there were two fuel trims shown - one didn't change at all, and the other one only changed within +/- 0.1.

I want to hook the ODBII back up to the computer again and get some solid readings, perhaps tomorrow after I get home from work while the engine is hot, but I would like to have some clue as to what I am looking for, what I should be seeing, and what readings are important. If anyone can help me out on this, give me an idea of what is 'normal' while sitting in the driveway, I would appreciate it.

 

[CaptainXL]

There is a pretty good article here:

Fuel trim can be a valuable diagnostic tool - Eastern Manufacturing

And a good video here:

[video=youtube;mzsOY_AhgY4]http://www.youtube.com/watch?v=mzsOY_AhgY4[/video]

Just remember that the O2 sensor is not controlling anything. It is just reporting what's happening with the air and fuel mixture. The precise term is Air to Fuel Ratio which is 14.7:1

 

[jaguarjoe]

14.7:1 gives minimum emissions
12.x:1 gives max power
16.x:1 gives maximum economy

Those values are by weight, not volume.

The upstream O2 may not control anything directly but it has a great influence on the PCM none the less.

 

[Denali n DOO]

Here's some info I read about fuel trims. I was getting a P0171 n P0174.

What are fuel trims all about?

 

[Shdwdrgn]

So it does appear that my fuel trims are working at this point. The short-term trim was cycling up and down, and the long-term only fluctuated slightly. The front O2 sensor was also cycling up and down. This all appears to be normal, although since I haven't seen a 'normal' scan myself, it's just a judgment call.

Another member messaged me about the possibility of the CPAS... This was something I had not heard of before, but it was an interesting experience today learning how to pop my serpentine belt loose and remove the power steering pump. The CPAS did not appear to be gunked up, but there was a bit of crud on the screens, so I hosed it down good and put it back in. I'll see tomorrow during my dive in the work if that makes any immediate difference.

One thing I thought of today... it used to be that when I would coast down a hill on the highway, my DIC instant-mileage would swing all the way up to 99mpg. It no longer does that, appearing to top out around 68mpg. I know that's not a very good measurement of anything, but it is a change that I've noticed. Maybe it means something to someone else?

In the meantime, any other thought on what I might check?

 

[Sparky]

If it would help you I'll try to remember to have Torque log my truck's fuel trims on my way to work and back for a comparison point.

 

[CaptainXL]

Could be the thermostat sticking open as well.I know mine is. if you drop the trans into second ...lets say to pass someone on the highway , does your coolant temp go down pretty far?. I did this today and when I came to a stop the exhaust smelled really rich. I bet if I monitored it with my Torque app I came out of closed loop. If your stuck in this vicious cycle mileage probably would be closer to city mileage.

 

[Shdwdrgn]

Hmm ok that reminded me of a couple other readings I saw... My coolant temp was reading 195, even though the dash was reading straight up (215 I think?). I thought it was odd that there was a discrepancy between the two. Also, my thermostat was replaced a couple years ago due to the old one being bad (and the dash was reading low).

Also while watching the settings, the fuel system was in closed-loop while idling, but went into open-loop wen I revved up the engine. Hopefully that is normal?

 

[Sparky]

No I don't believe that's normal. It should stay in closed loop. I think some go into open loop at WOT but not simply revving.

I could be wrong however... Now I've gotta test this on mine.

 

[CaptainXL]

Yeah agreed with Sparky on how open and closed loop works. Sounds like a stuck stat. Same boat as me.

What happens is the ACDelco tstats are imho crap. There is a rubber gasket that come loose and keeps the thermostat open during all times. My thermostat was changed 2 years ago as well with the ACDelco. Specifically speaking the ACDelco made in Mexico. Mine is now starting to act up same as you.

 

[Shdwdrgn]

No I don't believe that's normal. It should stay in closed loop. I think some go into open loop at WOT but not simply revving.

I could be wrong however... Now I've gotta test this on mine.

That's pretty close to what I was doing when it went to the open loop... I hit the throttle hard.

 

[CaptainXL]

That's pretty close to what I was doing when it went to the open loop... I hit the throttle hard.

See to me there is a difference between reving the engine up and down and full throttle. It shouldnt go into open loop just reving it a bit at idle. Full throttle would put rpm's around 5000-6000 when downshifting/passing in traffic. This is when you should see it go into open loop. Pretty much a useless test when parked cause the rpm's wont go much past 4k due to the rev limiter.

Another telltale sign that the thermostat is stuck is that it will take some time for the coolant to come back up to the 12 oclock position on the guage after passing/flooring. If you dropped down into 3rd or 2nd it might never fully recover cause the water pump is just overcooling the engine at those higher rpm's

Do this. When on the highway going 70, manually drop it down into 3rd. Run it like that and goose it a bit. If the coolant temp drops significantly and/or doesnt recover in 3rd gear but only if you put it back into drive/overdrive then I would say the tstat is bad.

 

[Shdwdrgn]

All right, I'm around 88k miles now, I figured I would go ahead and replace the spark plugs. They looked worn, but not in horrible shape (they were consistent in the brownish residue and wear on the tips). There doesn't seem to be any excessive oil blow-by, although there was some residue around the coil pack housing on #4 -- I assume from water leakage. The old plugs were AC Delco 41-981, and the new plugs are AC Delco Iridium 41-103.

While doing a quick test-drive after changing the plugs, my CEL popped on, so I hooked up the computer when I got home. Fortunately nothing to worry about, it was just my old 'friend' P0410. However since I was there, I collected the following info while sitting in the garage...

---

Intake manifold pressure: 3.6 to 4.2 PsiA at idle

Fuel trims:
short term: wanders from 0% up to 16%
long-term: wanders between 0% and 3.13%

Coolant temp: 197deg. Holding at 2000rpm, temp drops to 188deg after about 20 seconds.

O2 sensors:
#1: wanders between 0.065 and 0.935v
#2: wanders between 0.645 and 0.740v

---

So, does any of that appear to be of concern? I'm especially curious about the water temp, since that is obviously a common problem, and my mileage problems popped up about a year after the last time the thermostat was replaced.

I also just threw in a bottle of techron and refilled my tank. Not sure what else I can do at this point.

 

[CaptainXL]

From a cold start (sitting overnight) start and then idle the engine. If the upper radiator hose is warm before the needle on the gauge gets to the 12 oclock position then the thermostat is sticking open. 188 is borderline bad. The engine should heat up rapidly. If it is taking a while and you notice the temp guage lingering below the 212 mark then thats a sign as well. It takes my truck 15 - 20 minutes to warm up at idle at 32 degrees F. Way too long. Should be more like 5-7 minutes.

Use a vacuum guage to test in inches of mercury not absolute psi. Vacuum reading should not go up and down at idle. O2 readings look normal.

If you used an acdelco thermostat a year ago then I would say that is the problem.

 

[Shdwdrgn]

There is definitely no problem with warming up. From a cold start, it begins blowing hot air in about 1/2 mile. The last time I had my thermostat replaced, my temp. gauge was sitting around 200-205, so I knew there was a problem. Currently the gauge sits straight up at 210 after a few minutes of driving and never wavers.

 

[CaptainXL]

Going back to your original comment about getting the same gas mileage in the city and highway. This doesnt make any logical sense if you think about it. Time to do a compression test to see if the mechanicals are ok. If the highway mileage really is lower than it is supposed to be then I would concentrate on the fuel system. Specifically check for fuel leaks when at higher speeds or maybe an exhaust restriction. You may want to get a professional opinion or do some research.

 

[Shdwdrgn]

Ugh... I just finished off my first tank of gas since changing the plugs. It came back at 13.64mpg, a full 2mpg lower than I was getting before. Have I messed something up? I would think if one of the new plugs wasn't firing properly, I would feel it in the engine performance, but I have seen nothing different. I still have plenty of power accelerating, and it still runs smooth when both cold and hot.

And I thought I was confused before...

I'm going to run through one more tank of gas to confirm my readings, then switch back to my stock PCM and see what happens.

 

[CaptainXL]

Ugh... I just finished off my first tank of gas since changing the plugs. It came back at 13.64mpg,

Is this highway or city or a combination of both?

You really need to take care of the P0410 and then rescan. It is not uncommon for the PCM to ignore and withold testing of other systems if certain codes trump them due to the fact that they will knowingly alter or skew diagnostics testing.

I hope this makes sense.

 

[Shdwdrgn]

It makes sense, however that particular code is excessively common, and not something I've heard will cause any other problems. I just can't see throwing hundreds of dollars into repairing a component that is practically designed to continue to fail, when it serves no realistic value to the operation of my vehicle. Quite frankly, if I ever take the time to find the damn thing, I'm going to rip it out.

 

[CaptainXL]

Water getting into the system is common. But if the proper amount of air isnt making its way to the catalytic converter you could potentially be setting yourself up for some cascading costly repairs. I dont think its an expensive fix at all. Tons of articles on the trailvoy site about the issue.

 

[BRomanJr]

It makes sense, however that particular code is excessively common, and not something I've heard will cause any other problems. I just can't see throwing hundreds of dollars into repairing a component that is practically designed to continue to fail, when it serves no realistic value to the operation of my vehicle. Quite frankly, if I ever take the time to find the damn thing, I'm going to rip it out.

The P0410 code has a few causes and one of them is the solenoid valve is malfunctioning. This means it could be stuck open or partially open. This would cause a "vacuum leak" and positive fuel trims.

IMHO ALL codes should be given prompt attention. Like CaptainXL said, most emission systems affect other emission systems and the entire system must be working as designed or it can affect mileage/performance/driveability.

 

[CaptainXL]

I think bromanjr's explaination is the best thusfar. Get the p0410 fixed and go from there.

 

[Shdwdrgn]

I know everything *was* working fine on that system, but after having to pay to have water dumped out twice, it got really old. My understanding is that this is a system which runs for 30 seconds when you first start the engine cold in the morning, and then shuts off and does absolutely nothing the rest of the time. What is the point of an air system that fills up with water every 3 months? From what I had read before, the whole air pump could be completely removed without affecting anything else.

 

[BRomanJr]

I know everything *was* working fine on that system, but after having to pay to have water dumped out twice, it got really old. My understanding is that this is a system which runs for 30 seconds when you first start the engine cold in the morning, and then shuts off and does absolutely nothing the rest of the time. What is the point of an air system that fills up with water every 3 months? From what I had read before, the whole air pump could be completely removed without affecting anything else.

The air pump itself is not an active part of normal use once warmed up but the valve/solenoid must be sealed when the pump is not running and the system is dormant.

*EDIT*
After some thinking, I wondered if the valve not sealing is causing both of your issues, after the air pump runs, if the valve stays open a bit condensation may find its way into the pump lines and the pump. And after its cycle, the "extra" air being sucked in causes your economy problem. Water in the lines is not that common, there is another problem causing it.

 

[Shdwdrgn]

OK that's starting to make sense, and I re-read your earlier explanation. It's all starting to click now. I wonder if there's a way to manually close off the solenoids, or perhaps its time to investigate how to actually remove the whole system? It's odd, for all the problems people have with this particular system, I've never actually seen a diagram of where the components are or how the system is supposed to work. It's off to google...

[EDIT] Hmm if I'm looking at the right part, it seems like it would be easy to just replace the gasket on the back side of the solenoid with a piece of solid gasket material, thus effectively blocking off any airflow?

 

[Shdwdrgn]

Well this is incredibly encouraging! I pulled off the air solenoid this evening and blew some air in from both sides, and there was definitely some air leakage going through. I proceeded to take the unit apart and hose it down with TB cleaner, but I just could not get enough carbon cleaned off to get the solenoid to seat back down solid again. Perhaps its frozen in position, or perhaps the material that closes off the valve is simply destroyed... either way I gave up after about an hour of cleaning. I have a roll of heavy black gasket material on the shelf, so I cut out a solid piece to block off the airflow into the engine and bolted everything back together again (and managed to loose the dipstick nut down into the frame... )

I'll be going out for dinner soon, then have to make a trip in to Denver tomorrow, so I should get a good idea then if this had made any difference. Considering the amount of carbon buildup on the inside of the solenoid, there was definitely a leak. Thanks for the suggestion -- I'm crossing my fingers!

 

[Conner299]

I'm slowly learning about all the stuff you guys are talking about. I have a lot of info at my disposal, that I am always trying to turn it into knowledge. I'm guessing you are getting your MPG from the DIC cluster. I recommend the aCar app(Droid). If you take the time to use it, it is very handy for recording services, fill-ups, etc. I use it to record my fill-ups, and automatically calculate my MPG's. I feel they are pretty accurate, as I usually fill up at the same place. I have heard it mentioned here, but not sure if you have it. Torque Pro is invaluable. While reading this thread, mention of coolant temp, seemed to be referred to by using the factory gauges. I have heard, this is rather unreliable. My factory gauges show about 7 to 10 degrees higher than what Torque is telling me. Don't know if this helps, or if I'm just repeating what someone else has already said. But... I hope it does help. I don't trust oil life gauges, or MPG gauges, that come from factory.

 

[Shdwdrgn]

You assume I have one of them new-fangled smart phones... You definitely cannot use the DIC for accurate mpg, but it can give a rough idea of short-term changes. I know what ranges my DIC generally shows for certain types of driving, but all of the numbers I've given for mpg here were calculated by the tank. The other gauges, you just have to know how to read them. For instance, the oil gauge tells what the computer *thinks* the pressure should be, which is why it doesn't fluctuate much. The temp gauge should be roughly accurate -- a few degrees isn't going to hurt you, but it does give you enough information to know when there is an issue with the thermostat.

 

[BRomanJr]

Well this is incredibly encouraging! I pulled off the air solenoid this evening and blew some air in from both sides, and there was definitely some air leakage going through. I proceeded to take the unit apart and hose it down with TB cleaner, but I just could not get enough carbon cleaned off to get the solenoid to seat back down solid again. Perhaps its frozen in position, or perhaps the material that closes off the valve is simply destroyed... either way I gave up after about an hour of cleaning. I have a roll of heavy black gasket material on the shelf, so I cut out a solid piece to block off the airflow into the engine and bolted everything back together again (and managed to loose the dipstick nut down into the frame... )

I'll be going out for dinner soon, then have to make a trip in to Denver tomorrow, so I should get a good idea then if this had made any difference. Considering the amount of carbon buildup on the inside of the solenoid, there was definitely a leak. Thanks for the suggestion -- I'm crossing my fingers!

Sounds like a good short-term test fix. I'm skeptical that the gasket material will hold up to exhaust gasses (heat and pressure) over time.

I hope you get a result of some kind from this test, but do a search for the best price for the solenoid valve replacement while you have the time.

 

[Shdwdrgn]

I'm honestly not seeing the point of replacing the solenoid. Its definitely shot... the air pump is most likely dead, and since I was getting water in the line, the checkvalve is also probably bad. It would be FAR cheaper to strip out all the components, cut a metal plate to cover the hole where the solenoid mounts, pull the fuses, and pay $50 to have my PCM reprogrammed to ignore P0410. In fact, even if the solenoid is the only component I have to replace, this is still the cheaper solution. I just don't see any benefit to trying to fix this system?

Made my trip to Denver today, clocked about 100 miles round-trip. The MPG on the DIC did not just zip right up, however it IS increasing. Time will tell, and I figure the PCM has to relearn the engine conditions without the leak. This seems to be backed up by the fact that the MPG barely changed on the way down there, but started coming up during the trip back home.

 

[CaptainXL]

SBUBandit in this thread said PCM for Less can remove the SAIS codes. It's definitely not needed. Mine is blocked off. Any dealer or GM parts store can get the blockoff plate.

http://gmtnation.com/f23/removing-secondary-air-injection-system-3502/

 

[gmcman]

Also keep in mind winter fuel normally causes a slight hit on MPG. Also, do you let the engine idle for extended periods for a warm up? Your coolant should be at 195 deg under normal conditions, if your scan tool shows 195 then I wouldn't worry too much about it, 215 deg is not normal unless under a good load for awhile.

The temp gauge should be almost straight up, give or take a needle width. Looking at the gauge from the drivers position it (mine) appears to be straight up, lean over and look at it straight on it's 1 needle width to the left. If 210 is straight up and the temp is 195-197 deg, how can it be 210? When my temp reaches upwards of 205 then it's about straight up.

 

[Shdwdrgn]

Removing the P0410 code was one of the reasons I was planning on buying my own copy of HPTuners, however that is on hold for now because I think replacing my diff gears is going to provide a lot more benefit for the buck. If I'm understanding properly from the HPT forum, not only can you get in and change the existing code in the PCM, but there are also alternative operating systems that can be programmed. I would really love to dig into our code and rewrite my own for things like making the oil-pressure gauge respond with more accurate and real-time readings, but I guess that will have to wait for another day, unless I can find a simulator to play with.

 

[Shdwdrgn]

@gmcman - My actual coolant temps were at 197 after driving. My dash gauge reads straight up 210 degrees during operation, as it should. When my t-stat went bad before, the dash gauge dropped to 200-205, then returned to 210 after repairs. All of this was mentioned earlier in the thread, including the fact that there is an obvious discrepancy between the dash gauge and the temperatures reported by the PCM.

No I do not have to let my truck warm up, I park it in an attached garage.

Winter fuel has been accounted for - as I stated, this issue has been going on for well over a year now. We're not talking about a slight hit on my mileage, we're talking about more than a 25% change. I used to be able to get over 400 miles on a tank of gas on the highway, now I can't even reach the 300-mile mark (and my typical daily driving only gets me about 225 miles to a tank).

 

[CaptainXL]

You can't hold off on getting the P0410 fixed. If you don't get it fixed now you will never find out what other codes might be pending. You seem to have a more serious issue working behind the scenes. It is likely that other codes are being suppressed until the P0410 is dealt with. It needs to be fixed first or else you will continue to have poor mileage.

If your mileage worries you about the state of health of your engine you can do a compression test to give you piece of mind.

 

[Shdwdrgn]

I don't understand... why would other error codes be suppressed because another is showing? Shouldn't the computer be showing everything that fails testing? I can't think of any reason where that would be a desirable trait in troubleshooting a problem.

 

[CaptainXL]

why would other error codes be suppressed because another is showing? Shouldn't the computer be showing everything that fails testing? I can't think of any reason where that would be a desirable trait in troubleshooting a problem.

It is a desireable trait because it is made so that you can keep driving your vehicle.

The PCM does a lot of negotiating and troubleshooting all by itself. It is constantly looking at tables and figures to see what is normal.

There is also a heirarchy in the programming that works like a flow chart about 1000 pages long. It's called instructions/code. And it loops through this code over and over and over checking the state of your engine. The computer looks to see if certain things have been accomplished in order to turn some things on and other things off.

For instance. The vehicle doesn't go into closed loop until the PCM sees a proper temp from the coolant temp sensor. Is this all that is looks at? Absolutely NOT. It also looks at engine run time, IAT temp, rpm and also other pending codes that may interfere with the engines operation.

There are many variables inside the PCM programming that are too complex to comprehend all at once. Some codes are suppressed so that other operations of the PCM can continue.

Now let me see if i can directly answer your question the best I know how.

In the service manuals there are guides which show each P, B and C codes. Among other codes. Each of these codes corresponds to a test event which the PCM performs with logical statements such as true/false, AND, OR, NOR, XOR, etc. Each test has at least one code assigned to it and these are the codes which our scan tool can lookup in the PCM memory.

Setting a code or CEL is dependent upon certain prequalifying criteria which the PCM checks. If none of those criteria are met then the test will not be run and a code will not be stored. All prequalifying criteria need to be met in order to proceed with certain tests.

If a test that the PCM performs ever fails it will either set a code or a code and the CEL will come on as well. This depends on the seriousness of the code. Usually class A misfires will show sooner rather than later. Also, for instance if you happen to be in open loop mode or run over a pot hole that triggers the knock sensor then the misfire test will be withheld until the PCM sees the proper operating environment to do so.

Ever since Motorola made the first PCM for vehicles, manufacturers have been constantly adding to and adjusting certain criteria in the programming that allows the vehicle to run without any false positive codes being stored. Each vehicle has it's own characteristics and certain criteria need to be changed in order for the engine to work properly. For instance some engines have 185 degree thermostats and some have 200 degree thermostats. The programming is adjusted accordingly.

Simarily, your code more than likely indicates an issue which the PCM deems is serious enough that it cannot do any further testing because it would invalidate other tests that kick off additional codes. Like I said, it's like a flowchart. If the PCM see's a "NO" it will stop testing other emmissions items until that one is resolved. For instance it wouldn't make any sense to set off a code for "bank A mixture" when the PCM knows there is a problem with "Injector 1". That's kind of how it works. In your case the PCM is pretty much saying "STOP! Looks like the emissions system is not working right...better not check anything because all my other tests will be inaccurate due to this one anyway." Something along those lines.

I know alot of what I said was not filled with accurate examples or very technical, but like i said you can read about it all in the service manual.

Here is a good article about readiness testing which uses the loop and logical concepts I talked about.

http://www.autotap.com/techlibrary/obdii_and_emissions_testing.asp

 

[gmcman]

@gmcman - My actual coolant temps were at 197 after driving. My dash gauge reads straight up 210 degrees during operation, as it should. When my t-stat went bad before, the dash gauge dropped to 200-205, then returned to 210 after repairs. All of this was mentioned earlier in the thread, including the fact that there is an obvious discrepancy between the dash gauge and the temperatures reported by the PCM.

The only reason I push this fact is I don't want members to think that 210 is normal temp when 195 actually is. This gets tossed around alot and just call me overly obsessive if you want. Not bashing you, just want to make my point clear so if someone reads a scan tool and sees 195 and hears that "straight up" is normal since 210 deg is "straight up" that they don't think they need a new stat right away.

No I do not have to let my truck warm up, I park it in an attached garage. That's good.

Winter fuel has been accounted for - as I stated, this issue has been going on for well over a year now. We're not talking about a slight hit on my mileage, we're talking about more than a 25% change..

Have you changed your tires? Reason I ask, about 4 years ago I wanted to save a few $$ and went with a pair of Kumhos over the Michelins...lost 40 miles per tank. After 17K miles, I never topped 230 miles and I always had 270-280 on the small tank. Sold the Kumhos early and went back to Michelin's and regained my mileage. Low rolling resistance tires make a huge difference...just a thought.


Other than that, you need to address the P0410 code before any tune.

 

[gmcman]

Here's the point of my ramblings......as viewed fom the drivers position, the temp gauge as normally viewed.

Second pic is if you look at it straight on.....this is at 195 deg and yes I agree....none of our gauges are perfect and there are surely discrepencies.

 

[gmcman]

Sorry, no matter what order I load them it put the wrong pic first.

2nd pic, viewed straight on.