Posted by Paul Goswell on September 23, 2009, 5:33 pm
Year of Manufacture *: 2002 Engine Code or Type *: aps
im prob on wrong forum but can any body tell me of a good link or posting that explains how fuel trim works short and long as i need to understand how it works thanks paul
Re: fuel trim
Posted by Peter Cooper on September 23, 2009, 6:48 pm, in reply to "fuel trim"
Posted by paul goswell on September 23, 2009, 10:33 pm, in reply to "Re: fuel trim" Paul Goswell
thanks peter i had a skoda 2.0 turbo in today mil light on for fuel trim the engine seemed to run ok but the short and long fuel trim was at -25% i found a small vacuum pipe under the inlet had collapsed when i replaced it and deleted code the long term was at 0.0% and short trim started at 18% and stepped down to 2% after a few minutes so i presume it should now be ok
Re: fuel trim
Posted by Dave Harney on September 24, 2009, 8:36 am, in reply to "Re: fuel trim" Dave Harney
Hi Paul, Here is some info I found a while back..... I've had to break it up into two posts as its too big !!
Block Learn/Integrator,
I was introduced to the concept of fuel trim at a GM training center back in the mid-'80s. I had a bit of difficulty absorbing the concept and I think part of the problem was the non-intuitive words that GM had chosen to describe the process. "Block-learn" apparently made perfect sense to an engineer who knew that long-term fuel tables in the computer were set up in blocks that corresponded to varying rpm and engine load conditions. "Integrator" apparently seemed a good word to designate the process by which the PCM would respond to rapidly to integrate the changes in O2 sensor voltage into its base pulse-width. But to a learning tech, the words didn't speed the learning process much. On top of the odd words, "block-learn" for long-term fuel trim and "integrator" for short-term fuel trim, GM engineers chose meaningless (to us, anyway) numbers to designate the parameter. The number 128 was selected as the middle of the road number for fuel trim, i.e. at 128 the computer was running exactly at the air/fuel ratio that had been pre-calculated by the PCM, based on its major inputs. Numbers above 128 indicated that the computer had to add additional fuel via a wider pulse width based on feedback from the oxygen sensor. Conversely, numbers below 128 indicated that the computer was narrowing the pulse width to reduce fuel delivery.
Why not use pulse-width numbers?
I often get this question from my students: "Why can't we just use pulse-width numbers instead of fuel trim?" It's a fair question and I usually demonstrate the answer by using a lab scope to measure pulse width on an idling engine. It will read anywhere from 1.2 to 4 ms depending on the engine, and I'll ask them, "Is that a good number?" to which they will typically respond, "What's the spec?" That is, of course, my point. There are no published specs for base pulse width in most cases because all parameters that affect a computer's base width would have to be taken into consideration.
But even if there were specs, they wouldn't necessarily help us to see at a glance if fuel trim is out of spec by a small amount. This is because the computer adjusts pulse width very quickly and by very small amounts (amounting to microseconds), and what we see on even a lab scope amounts to an average reading. To prove this to yourself, hook up your scope or your scan tool and create a vacuum leak. Compare the change in fuel trim with the displayed change in pulse width. You'll notice that the fuel correction is much more noticeable in terms of fuel trim than it is in terms of pulse width.
Base pulse-width
So where does the computer get its base pulse width from? This depends on the type of system we're working with, either speed-density or mass airflow sensing. Speed density systems use a MAP or a VAF (Vane Air Flow) sensor as their main input. Mass airflow systems, as the name implies, actually measure the mass of the air coming into the engine. Remember that what the computer is after, once the engine warms up, is a base pulse-width that will come close to a 14.66 to 1 air/fuel ratio by mass. That is, 14.66 grams of air for every gram of fuel.
Speed-density systems use a complicated formula based on MAP, rpm, TPS, and engine volumetric efficiency to calculate base pulse width. During cold operation the computer is going to also use the coolant temp sensor, and, on some applications, the air temp sensor, to calculate a base pulse width that's intended correspond to an air/fuel ratio that will make for good cold drivability. There is no fuel trim during warm-up, wide-open throttle operation, or deceleration because during these modes the computer is not using the oxygen sensor to make adjustments to the air/fuel ratio. Knowing this fact can help us in our diagnosis. Problems that occur only when hot or only when cold can be more easily isolated if we know what hot fuel trim values are.
Volatile or non-volatile?
Just what we need, more words! But they're important for understanding fuel trim and the ability of the computer to learn and to set codes when things get out of hand. In the early days of fuel injection, the computer usually forgot everything it knew about fuel trim tables as soon as the key was shut off. This was known as "volatile" RAM (Random Access Memory). As the years progressed, manufacturers found that is was desirable to retain the memory of long-term fuel trim tables to improve overall performance, including open-loop performance, so they added non-volatile long-term fuel trim. That way, if a problem existed in fuel control and the computer had learned its way through it, on the next start-up the computer could pick up where it left off without having to relearn the problem all over again.
One important point to ponder here: What if I were to fix a problem that affected fuel trim, say, by replacing an out-of-spec mass air flow sensor, and did not disconnect the battery? On restart, the computer would continue using the old fuel trim values, which are now incorrect, and I might find my car runs worse now than it did before. Consult appropriate repair manuals for information on how to clear long-term fuel trim tables. Disconnecting the battery is an increasingly non-recommended procedure because the customer looses such valuables as radio settings, power seat memories, mirror memories, etc. Dig through the functions on your scan tool first and see if you can find a menu selection for "clear all adaptive strategies." In many cases, clearing codes on OBD II-equipped cars will also clear adaptive strategies. On a few Asian vehicles, oddball steps are required such as stepping on the brake pedal after the battery had been disconnected.
Weighted or non-weighted cells?
Huh? More non-intuitive words? Well, let's just say that you need to know why you can look at a long-term fuel trim number at idle of +25 yet have no trouble code for "system always lean" (P017? Series codes). The reason is that the idle cell is not a weighted cell and the computer is not going to use it to set a code. But this in itself is going to be a help to you in diagnosis. Drive the car (preferably on a dyno, or with someone else watching the scan tool) and record the long-term fuel trim numbers through a range of engine load and rpm modes. If the fuel corrections are high mainly at idle, chances are good you're dealing with an air leak problem, either a vacuum leak, or an air leak upstream of the MAF sensor. The reason is, of course, that as engine load goes up, the leak becomes a much smaller percentage of the total air volume and the fuel correction factor is less. So, even if the long-term fuel trim is at its limit at idle, no code will be set because at the higher load level the computer is able to successfully compensate for the problem. Weighted cells are generally located in the light-load, light-throttle application range, which would correspond to part-throttle cruising conditions. Fuel trim at the end of its travel at this point is supposed to set fuel trim codes.
You can trim a sheeps wool many times, but you can only fleece it once !
Re: fuel trim
Posted by Dave Harney on September 24, 2009, 8:44 am, in reply to "Re: fuel trim" Dave Harney
So, what else can I find with fuel trim values?
Problems come up when base pulse width is not correct for assorted reasons. Fuel pressure too high or too low is common. Low fuel pressure is far more prevalent, due to plugged fuel filters or worn pumps. As the pump pressure starts to decay or the fuel filter begins to clog, the computer will respond by adjusting long-term fuel trim numbers in an attempt to get short-term fuel trim back to zero correction. Often we will see this trend more evident under higher load cells.
Fuel trim is also a great help in diagnosing out-of-spec sensors, both in speed density and mass air flow sensing systems. Very often a careful and thoughtful diagnostic approach is necessary. The customer complaint may be that the engine pings or has poor performance. A look at long-term fuel trim will verify that the computer is adding fuel (positive fuel-trim numbers above 10 or block-learn above 145) to compensate for a lean condition across a range of load conditions. If the vehicle has left- and right-bank oxygen sensors, often comparing fuel trim on both banks can tell me if I have a problem peculiar to one bank, such as a badly restricted fuel injector.
But now I need to do pinpoint testing. I always start with the easiest stuff. First, hang that ancient tool that is so often overlooked, and so easy to use -- the vacuum gauge. No point chasing computer controls if my needle indicates a mechanical problem, now is there? Next, are the sensors that affect base pulse width in spec? This is not always very easy to determine. The best way I know of is to use your scan tool and lab scope on known-good vehicles as often as you can. Record the frequency of a good digital MAF or the waveform of a good analog MAF. Remember, too, that BARO readings, whether directly measured or inferred, have a large impact on base pulse width. MAP sensors can be just as flaky sometimes as mass airflow sensors. Honda Map sensors, for example, can cause problems when out of spec by as little as 2/10ths of a volt.
I generally leave fuel pressure and volume testing for later point due to the inconsistency of manufacturers in providing easy access to tapping into the fuel rail. But don't make the mistake of assuming it is okay. The lack of a Schrader port is no guarantee that a pump is not going to fail!
If all my information at this point looks good, it is time to consider restricted fuel injectors. Unfortunately, this may prove difficult as the location of fuel injection connectors for balance testing is sometimes flat impossible. But I have had restricted injectors set system-lean codes and cause massive drivability problems, so don't rule this one out just because it's hard to check.
Fun factors
In OBD II (and even in some OBD I) applications, the computer itself uses the fuel trim values for diagnostic purposes. The computer will look at fuel trim to determine if evaporative purge, secondary air injection, and even EGR (in some cases) are working. It's helpful to know this because sometimes problems with fuel trim will end up setting codes for apparently non-related systems. Also, some manufactures will use the oxygen sensor downstream of the catalyst for fuel trim if problems develop in the upstream O2 sensor. Techs on iATN have proved that even a heater failure in an upstream O2 sensor can cause the PCM to revert to the rear O2 sensor and not switch back to the front unit even when it comes to life. And don't forget, the computer is depending on an accurate reading from the oxygen sensor to make fuel trim adjustments. Use your gas analyzer to verify that a shifted O2 sensor is not "lying" to the computer, causing it to add or subtract fuel when such action is not called for.
In closing...
Fuel trim numbers, unfortunately, are only available with a scan tool. As we independents know too well, many Asian and Euro vehicle manufactures got dragged kicking and screaming into data stream by OBD II, which leaves us with no fuel trim numbers on many, if not most, pre-OBD II vehicles built by said manufacturers. Another unfortunate fact is that Ford, for some reason I cannot fathom, often provided only short-term fuel trim numbers to the scanner on many of its vehicles before the advent of OBD II. This deprives us of the ability to look for trends in vehicle performance and forces us to rely on a trouble code setting, which may not happen until long after a drivability problem surfaces. So, we have to deal with what information we have and make the best of it. Rest assured, if short-term fuel trim is pegged at one end, long-term will also be pegged even though we can't see it on the scanner. Wise use of this data can steer us in the right direction for starting our diagnostic sequence.
You can trim a sheeps wool many times, but you can only fleece it once !
Re: fuel trim
Posted by Pete Mutlow on September 24, 2009, 9:27 am, in reply to "Re: fuel trim" Peter Mutlow
Hi Dave,
Good read. Thanks
Pete M
Re: fuel trim
Posted by paul goswell on September 24, 2009, 4:06 pm, in reply to "Re: fuel trim" Paul Goswell
thanks dave that was a very interesting read as with the motor trade we learn some thing every day
on September 23, 2009, 10:33 pm, in reply to "