Known good fuel trim for driveability analysis
I am putting up this discussion in regards to this post on a fuel trim problem. diag.net/msg/mxin1kqepl… Many years ago I was introduced to fuel trim and graphing scan tools. This information and system has made me more money with driveability diagnostics than any thing else I have ever used. Since this discussion was about a vehicle with the same platform I thought I would take a known good vehicle, give it a FRTD (flat rate test drive) and capture some good information, then explain; how, when, where and why about the captured data.
When it comes to testing vehicle systems, whether it be an electrical system, a brake system, or a driveability problem, I have found the best and most accurate way to do the testing is to do the testing dynamically, in other words, do a test so the car will tell YOU where the problem is and what the problem is. This is why we use a current probe & lab scope and do volt drop tests when working on electrical problems. It is also why you should be using fuel trim data graphs to point out where fuel related and airflow related problems are.
This vehicle is a 2006 Mercury Mountaineer (Ford explorer) with 4.6 power with 156,000 miles on the clock. This is one of my personal vehicles, and has no driveability problems. This is why I used it, since the need for known good data, whether it be a waveform, or scan data. In the shop I also have a 2010 F150 with 4.6 power. There are enough differences in the two vehicles to keep me from using the known good data from the F150. You need to be testing on known good vehicles to really get a feel for the systems and how they work. Vehicles are much more than steel, nuts & bolts.
I have driven the vehicle until the engine is warmed up, don't expect a cold vehicle to react the same way to fuel trims as a warmed up on will. My FRTD route is one mile out on a crooked road with less than 1/4 mile of straight road. I drive every vehicle on the same road when it comes to capturing my data.
When it comes to this kind of work, I have my favorite scan tools. Most times on a Ford vehicle I will use my IDS, although for simplicity for this post, I used my Snap on tool, and uploaded and edited the captured data with Snap On Shop Stream Connect. Since the Snap on tool does not give me all the needed pids on the enhanced side, I have used OBD2 generic data. This data is a little slow, although it will serve for this discussion.
diag.net/file/fnukiisvh… I have made some notations on the picture. This data is taken over the one mile test drive, but then I can go back and look over the whole capture to see if anything jumps out. In this case, I am interested in the MAF data at engine idle. Since this is a 4.6 engine, I want to see very close to 4.6 G/S of air at idle, with all electrical loads off, the engine at operating speed and the transmission in either park or neutral. In this case, the data is spot on. What I am looking for is to get an idea if the MAF is reporting correctly at idle. If perchance it is NOT, then I can take a glance over at the oxygen sensor voltage, and fuel trim data to give me a good idea about the air flow. If perchance I see what looks like correct air flow at idle and the trims are not correct, then there is always the oxygen sensor data to examine.
Capture 2 is the data zoomed out with a WOT run from a stop.
Some people call this a VE (volumetric efficiency test). What ever you want to call it, the bottom line is you take the vehicle out and spank it hard, and let it tell you if the fuel volume is correct, if there are any exhaust restrictions, if the camshafts are phased properly, if the injectors are flowing correctly, if there is a restricted air filter, just a few things like this. Wow,, so far, I haven't even had to lift the hood to get a great indication of what is going on under the hood.
In this capture I have highlighted two things. I like the way the long term and short term fuel trims are trending. Yes, pay close attention to what the trims do when you move the throttle and when the transmission shifts. Little things like this can make the difference between a long drawn our diagnostic process and a short accurate one. Since the oxygen sensors, (both front and rear) voltage has gone over .900V at WOT, I know for sure there is enough fuel being injected to satisfy the needs of the engine that is running its heart out. OH BTW, if the engine wont stand a WOT run, it needs to be replaced any way :)
By using a system such as this to gather information, it will save you a lot of time poking around looking for problems. Let the vehicle tell you, instead of you going out looking.
Albin, FRTD (flat rate test drive) has become one of my go to tests since I first read it in one of your Motor Age articles. The VE test calculator is a relatively newer addition I've begun doing as well. It's an excellent testing technique to the point that I've adopted the basic concept for off-road equipment as well. It's a work in progress but has helped me cut down diagnostic time on
Good morning Albin, thank you. This is an excellent example of presenting a procedure; clear and consise.
Hi Albin, I've seen the "rule of thumb" regarding engine liters and the expected grams per second which I think is ok for quick math but, have you ever wondered how the controller is using that info to formulate a number it can deal with? We're measuring the air flow against a time factor but we're not considering the engine exact speed. The determining factor value is measured in grams and
WOW Scott, that is great information. A little over 50 years ago, I had my heart set on being a civil engineer. that never happened, and I got into heavy equipment repair & maintenance, which led to many other things, before I landed in the automotive world in the spring of 1992. I have always wished I had some back ground in engineering of things such as this. At my age, this wish is long
If you correct for the time and location of the test (1171 ft above sea level, 55deg air @ 50% humidity), the VE is around 86.6%. Now, I'm assuming (and we all know what happens when you do that) that the engine factor is based on a 0% loss due to ring leakage, so would a 10% leakage rate lower the physical max fill used in the calculations by the same percentage or does that not factor into it?
Scott, I have been trying out your concept (idle G/CYL spec) and I can't seem to make it work. Not sure what I am missing but hoping you can enlighten me. I have a lot of data and it doesn't seem to match up with your 0.10 - 0.12 idle spec. Today had had a horrible bad MAF sensor and G/CYL was 0.1122. See the graphs. What am I doing wrong? diag.net/file/f45ctdx2x…
Hi Randy, Your math is correct. Do you have the data from the new MAF? I'm curious to see what you have. Here are a few of my recent observations and based on them it looks like I'll need to refactor my baseline reference and include "your mileage may vary". 2017 Rav 4 = 0.104 g/cyl @ idle 2010 Mini = 0.138 g/cyl @ idle 2010 Cadillac 3.6L = 0.149 g/cyl @ idle Do you have larger or better…
I don’t have after graphs yet, still waiting on repair. To be honest, I have a very hard time with images on this site. They are much better resolution on my laptop and I must be doing something wrong. I could not get the attach image button to work until I deleted my reply and started over. Then it put low quality images up. It took so long I just gave up and posted. I guess I need…
On the image issue, what browser are you using? The tools we've deployed here required a modern browser. Have you tried using Chrome? I'm not normally looking to measure VE @ idle, just observing the cylinder air # when I suspect FT is being affected by MAF output. I just took a look at the calibration for the desired airflow from a 2000 Camaro 5.7L. At 104c it's looking for 6.2 grams/second…
I used chrome with very poor internet connection. Could something be compressing them due to slow internet speed? Last week I had a technician asking about it for a Hyundai Tiburon, not sure of the year right now, but it had it for the cat. Once you do the math and calculate the VE, you simple use the load PID to get in the range. He didn’t even know what the spec meant but once I did the math…
The slow connection could be the issue but I'm not sure. Can you email me the originals and I'll replace what you added earlier? I'd love to get a better look.
Can’t you have Thomas drop by to solve it? 😀 I’ll send you files tomorrow.