Ha JP, ik rij ook op LPG en dat gaat bijna altijd goed... Op de momenten dat het wat minder ging op LPG heb ik onderstaande instructie uit deze post op het forum gevolgd:
link naar post
Misschien heb je hier iets aan om vanaf "0" te beginnen met een diagnose.
Mijn LPG ECU volgt de benzine ECU (Prins VSI 1 installatie), voor onderhoud etc. van de LPG installatie ga ik naar Kees Verhulst (www.verdampere.nl) maar dat is voor jou best een afstand...
Nu de tekst: (bij "and/or turning the ignition key" doe ik voor het weer aankoppelen van de min pool op de accu een seconde of 20 "starten" met de sleutel...)
Hallo volvo355
Importance of Reseting Fuel Trim (LTFT).
Postby Brucebo » 08 Jun 2009, 20:46
It is extremely important to reset your Long Term Fuel Trim (LTFT) after any work that could possibly affect the engine fuel and ignition system. This incudes, but is not necessarily limited to:
Oxygen Sensor
MAF Sensor
MAP Sensor
IAT Sensor
Coolant Temp Sensor
Throttle
TPS Sensor
IAC Valve
Any Vacuum Line
Fuel Pump
Fuel Injectors
Fuel Pressure Regulator
Purge Valve
Cannister Shutoff Valve
Spark Plugs
This is done by disconnecting the battery, waiting for 10 minutes or so, and/or turning the ignition key to the starter position to discharge the system. Note, the code "Erase" button on scanners does not reset the LTFT, at least not on my S70. Not reseting the fuel trims provides for a small chance that your adaptive fuel map could fail to converge and/or become unstable.
Bare with me, this article somewhat technical.
The fuel map is a two dimensional table of values, sort of like an Excel spreadsheet, for how much fuel to squirt, based on engine RPM and load value. It consists of the "base fuel map" (BFM) which is hard coded into the ECM, and the adaptive LTFT. In addition there is the Short Term Fuel Trim (STFT). This is a dynamic real time variable fed back to the ECM by the Oxygen Sensor(s).
The BFM is unchanging and is based on the automotive engineer's, or aftermarket upgrader's best estimate of what the fuel map should be, given nominal engine design, fuel, and sensor calibration. Added to the BFM is the LTFT, which adjusts for sensor drift, engine wear, fuel, etc. The LTFT is basically a long term running average of the STFT. If the BFM was perfectly adjusted, and unless something is broken it is usually pretty close, the LTFT would be zero everywhere in the table. It SHOULD be fairly small values everywhere in the table. Once the LTFT is fully adapted, the STFT should swing about zero.
The fuel map needs to be a "smooth" or "continuous" function. By this I mean that small changes in engine operating conditions should produce only small or no change in the fuel map value. For example, all else being the same, for a fuel map value at 2000 RPM, you DON"T what the map value at 2001 RPM to be something completely different! This would wreak havoc on your engine.
Here's where the necessity of reseting the LTFT comes in. If you have a bad sensor, or a loose vacuum line, etc. Your Oxygen sensor will send back large adjustment values of STFT. These will slowly get added to the LTFT and it will adapt. Here's the rub. Not all the values in the LTFT will adapt at the same rate. A value will only get adapted when you're driving in that region of the fuel map (combination of RPM and load). Most drivers are creatures of habit and rarely cover their whole fuel map in any reasonable period of time. Your fuel map will start to diverge, become discontinuous and, especially during Wide Open Throttle (WOT) when there is no feedback from the O2S, you will experience surges and hesitations as you swing through the fuel map. Even during non-WOT your O2S feedback takes at lease 1/2 to 1 second to respond so you may still experience surges and hesitations.
Now let's say you go and fix your mechanical problem but don't reset your LTFT. These fuel map discontinuities are still there! They MAY adapt back over time, but this could take awhile, and in the mean time you will still experience surges and hesitations. But worse things may happen. The fuel map adapting ability is really only reliable for "SMALL" corrections. In electronics it's call "small signal analysis". If the adaptive algorithms are presented with LARGE correction imputes, it is entirely possible that the adaption may diverge and/or become unstable through positive feedback. Let me give a somewhat contrived example. Let's say at 2000 RPM the LTFT is +18% and at 2050 RPM the LTFT is -18%. The engine hits 2000 and gets some more fuel and revs to 2050. The O2S reads "rich" and sends a negative correction, but since there is some time delay it sends the negative correction to the entry for 2050 and the LTFT goes to -20%. In the meantime as the engine hits 2050 it now gets less fuel and goes back to 2000, the O2S reads "lean" and the process happens in reverse, and so on ... . And the owner is pulling his hair out looking for "vacuum leaks", or replacing "bad" sensors when all they really needed to do was unhook the battery for a few minutes.
-Bruce
want wel ingewikkeld!
