ECMTuning - wiki

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.

<h1>Fuel-trims and How They Work</h1> <div class="level1"> <p> Like any control system, there are two basic fuel modes in the DSM ECU: <b>open loop</b> and <b>closed loop</b> </p> <p> In a nut shell, while running <b>open loop</b> the ECU just assumes it knows how much fuel is really going in and it assumes that accurate airflow data is being presented to it. It does not check these assumptions against anything. It just calculates how much fuel is need to achieve a certain A/F ratio based on airflow data. It looks this target A/F ratio up in an <b>open loop</b> fuel table indexed by RPM and engine load. </p> <p> In <b>closed loop</b> mode, the ECU checks itself by "watching" the front O2 sensor feedback. The factory front O2 sensor switches abruptly around 14.7:1 A/F ratio. The ECU tries to adjust fuel so that the O2 sensor swings back and forth around this target. To do that, the ECU has to have some way to vary its basic fuel calculation. It uses fuel trims for this purpose. Think of these variables as dials if you want. Turn the dial one way and the ECU adds fuel, turn the dial the other way and the ECU reduces fuel. </p> <p> While running <b>closed loop</b> mode, the ECU uses what's called a "short term fuel trim" (STFT) variable for quick changes. This is the variable is uses to force the O2 sensor to cycle. In a perfect setup, this variable would swing from positive to negative values of about equal amplitude. As it's swinging up and down, the front O2 sensor should swing up and down. This is the basic <b>closed loop</b> mode of operation. </p> <p> The ECU will also watch what's going on with the short term fuel trim variable. If conditions are right and the short term variable has stayed positive or negative long enough, the ECU will move the long term STFT pattern into the long term variable (LTFT). This is called a <b>learning mode</b>. Let's say instead of swinging around 0%, STFT is swinging up and down around -5%. The ECU will eventually take note of this and move that -5% adjustment into the appropriate LTFT variable and then zero out the STFT variable. Since both are factored in, the end result will be the same but STFT should now swing around 0%. </p> <p> The ECU uses at least a couple different LTFT variables depending primarily on airflow. Lower airflow values seen at idle will typically result in the LTFT Lo variable being used (and updated). Higher airflow values seen at cruise will result in LTFT Mid (2G) or LTFT Hi (1G) typically being used. </p> </div> <h1>Details</h1> <div class="level1"> <p> There are several reasons the ECU might switch into or out of <b>closed loop</b> mode. In addition, even while running in <b>closed loop</b> mode, the ECU may or may not be running in a mode that allows learned STFT adjustments to propogate into the LTFT variables (<b>learn mode</b>). So there are really three different modes of operation. </p> <ul> <li class="level1"> Open loop (STFT and LTFT ignored)</li> <li class="level1"> Closed loop (STFT used but LTFT not updated)</li> <li class="level1"> Learn mode (STFT used and long term patterns moved into LTFT)</li> </ul> <p> We'll attempt to document some of the conditions that will force the ECU into or out of these modes. First, it's probably no surprise that these conditions are different between 1G and 2G ECU codebases. So the conditions will be listed separately for each. </p> <p> The ECU will force open loop mode for a number of different reasons. The following is intended to list out the more common reasons. If any <b>one</b> of these is true, then the ECU will run in full open loop mode. That is, STFT will not be used and, of course, <b>learn mode</b> will not be active either. </p> </div> <h2>1G Open loop mode (any one true)</h2> <div class="level2"> <ul> <li class="level1"> Throttle position too high (varies by RPM)</li> <li class="level1"> Airflow too low (units are tricky, but it's very low)</li> <li class="level1"> Coolant temp < 87F</li> <li class="level1"> The O2 sensor did not switch around 0.5v for over 20 seconds while running in closed loop mode. In that case, the ECU locks itself in open loop mode.</li> </ul> </div> <h2>2G Open loop mode (any one true)</h2> <div class="level2"> <ul> <li class="level1"> Throttle position too high (varies by RPM)</li> <li class="level1"> Airflow too low (units are tricky, but it's very low)</li> <li class="level1"> Coolant temp < 51F</li> <li class="level1"> Coolant temp > 228F</li> <li class="level1"> The O2 sensor did not switch around 0.5v for over 128 seconds while running in closed loop mode. In that case, the ECU locks itself in open loop mode.</li> </ul> <p> All of the following have to be true for a full 5 seconds. If any single one is not true even for a millisecond, the timer starts over and trends in STFT will *not* be moved into LTFT. </p> </div> <h2>1G Learn mode (all must be true)</h2> <div class="level2"> <ul> <li class="level1"> Coolant temp >= 190F</li> <li class="level1"> Intake air temp < 123F</li> <li class="level1"> Baro >= 22.9 inHg</li> <li class="level1"> Baro &lt;= 31.6 inHg</li> </ul> </div> <h2>2G Learn mode (all must be true)</h2> <div class="level2"> <ul> <li class="level1"> Coolant temp >= 180F</li> <li class="level1"> Intake air temp < 133F</li> <li class="level1"> Baro >= 22.9 inHg</li> <li class="level1"> Baro &lt;= 31.6 inHg</li> </ul> </div>