Help needed with rich engine condition

From : heather hiscox

Q: hi ive been trying to diagnose an excessively rich condition in my 92 350 pickup 5.9l tbi. the engine runs so rich that the oil turns pure black immediately after changing it. the oxy. sensor signal voltage runs .7 ..9 1.1 with short spurts to .2 .3. in the haynes manual it says make sure the vacuum is reaching the fuel pressure regulator and it shows a picture with the vacuum needle pointing well up. of course i get no vacuum here how can there be when the vacuum port is at the top of the throttle body. i tried connecting the regulator vacuum line to a port lower on the throttle body but now the engine runs excessively lean and the oxy sensor now reads very low ..02 .03. another thing it says the map sensor signal voltage should be 1.5 - 2.5 at idle and 4 to 5 at high rpm. but mine reads about 1.1 - 1.3 at all rpms. but if i pull off the vacuum hose it reads 4.7 volts. isnt this the opposite of what it should do but it checks out ok if i take it off and hook up a 5v supply and suck on the vacuum hose. i checked all the vacuum hose routings and they look ok. all other voltages read ok and the exhaust checks out ok using the quick vacuum change when throttle is released method. for some reason there is no self diagnosis on this truck. the check engine light glows just before starting but no amount of turning the key with produce flashes. any help will be appreciated. heather .

Replies:

From : neil nelson

heather hiscox hh@eui.doe.edu wrote hi ive been trying to diagnose an excessively rich condition in my 92 350 pickup 5.9l tbi. the engine runs so rich that the oil turns pure black immediately after changing it. the oxy. sensor signal voltage runs .7 .9 1.1 with short spurts to .2 .3. in the haynes manual it says make sure the vacuum is reaching the fuel pressure regulator and it shows a picture with the vacuum needle pointing well up. of course i get no vacuum here how can there be when the vacuum port is at the top of the throttle body. i tried connecting the regulator vacuum line to a port lower on the throttle body but now the engine runs excessively lean and the oxy sensor now reads very low .02 .03. the fuel pressure regulator on a dodge truck tbi system is strictly mechanical just like a gm tbi. the port to the top of the throttle body is nothing more than a drain connection in case the regulator diaphragm should leak fuel. as usual haynes prints wrong info. another thing it says the map sensor signal voltage should be 1.5 - 2.5 at idle and 4 to 5 at high rpm. but mine reads about 1.1 - 1.3 at all rpms. but if i pull off the vacuum hose it reads 4.7 volts. isnt this the opposite of what it should do but it checks out ok if i take it off and hook up a 5v supply and suck on the vacuum hose. i checked all the vacuum hose routings and they look ok. map signal sounds okay. low voltage with high vacuum high voltage with low vacuum. you probably arent seeing the voltage go high because youre not putting the engine under enough load. all other voltages read ok and the exhaust checks out ok using the quick vacuum change when throttle is released method. for some reason there is no self diagnosis on this truck. the check engine light glows just before starting but no amount of turning the key with produce flashes. maybe there arent any codes stored yet. any help will be appreciated. your tbi system is built for dodge by holley. they have a very good reputation for worn injector pintels which creates the rich condition youre experiencing. replacing the injectors has always solved this problem. ive seen these injectors go bad in a years time. .

From : heather hiscox

thanks ill try that - and a new regulator too. heather .

From : big al

thanks ill try that - and a new regulator too. heather for what its worth on my 90 d150 the o2 sensor was nuts and it made the engine go full rich. have the computer scanned and see what comes out. dont make the mistake i made. buy the o2 sensor from the dealer or get a bosch. i bought one at checker and it failed in about 18 months. then i chased my tail trying to figure out what in my truck caused the o2 sensor to fail twice. btw on my truck the sensor usually read correctly. what happened is at idle it would falsely tell the computer the engine was lean the computer would richen up the mixture and it would foul the sensor. then it would always tell the computer the mixture was lean. i could remove the sensor clean it out and it would work again until i let it idle long enough to foul the sensor. i replaced the sensor again and the problem went away. al .

From : heather hiscox

hi well i have the meter hooked up in the cab and can see what the o2 sensor reads when i drive and most of the time it reads .9v. it occasionally shows .2v when i coast but never a medium voltage. this shows that the o2 sensor is doing its job and telling the ecu that things are too rich. what i dont know is whether the ecu is acting on this information to shorten the fuel pulses. does anyone know if there is a way to check this ... short of buying an oscilloscope thanks heather .

From : big al

hi well i have the meter hooked up in the cab and can see what the o2 sensor reads when i drive and most of the time it reads .9v. it occasionally shows .2v when i coast but never a medium voltage. this shows that the o2 sensor is doing its job and telling the ecu that things are too rich. what i dont know is whether the ecu is acting on this information to shorten the fuel pulses. does anyone know if there is a way to check this ... short of buying an oscilloscope thanks heather i use a otc scanner. it shows injector pulse width. seems you have some electronics skill. here is what i did when i was chasing my tail. i got a aa cell and made a divider out of two resistors. careful there is a bias current coming from the ecm so it can detect an open sensor. i used a dvm and trimmed the circuit so i could emulate rich or lean o2 readings. just disconnect the o2 connector and wire in your circuit. i was able to drive the ecm full lean and full rich and the ecm compensates the mixture. the way i did it the two resistors were in series. across one i had about .5 volts lean. and across the other i had about .9 volts rich. with a properly working system the o2 sensor output is oscillating from rich to lean. it acts more like a switch than a linear sensor. give the ecm a rich signal then monitor the sensor and see how it reacts. voltage should go down indicating a lean condition. and of course visa versa. al .

From : heather hiscox

oh i see what you mean .. just leave the o2 sensor connected to the meter but disconnect it from the ecu then hook the variable voltage to the ecu signal line and see if it varies the oxygen. i have a variable voltage box for testing transistors. thanks ill try it. heather i use a otc scanner. it shows injector pulse width. seems you have some electronics skill. here is what i did when i was chasing my tail. i got a aa cell and made a divider out of two resistors. careful there is a bias current coming from the ecm so it can detect an open sensor. i used a dvm and trimmed the circuit so i could emulate rich or lean o2 readings. just disconnect the o2 connector and wire in your circuit. i was able to drive the ecm full lean and full rich and the ecm compensates the mixture. the way i did it the two resistors were in series. across one i had about .5 volts lean. and across the other i had about .9 volts rich. with a properly working system the o2 sensor output is oscillating from rich to lean. it acts more like a switch than a linear sensor. give the ecm a rich signal then monitor the sensor and see how it reacts. voltage should go down indicating a lean condition. and of course visa versa. al .

From : heather hiscox

hi guess what. i fixed it! i tried that sending a dummy o2 signal straight to the ecu but there was no response whatsoever. so i started thinking there must be something wrong with the smec itself. so i took it off took it apart cleaned the dust out of it and put it back together. i started the truck and it seemed to be working now so i took it for a spin all the time watching the hi-z fet voltmeter and now the needle was moving slowly back and forth instead of being stuck on .9 or jumping erratically. i can only conclude the tiny voltage from the o2 sensor wasnt getting through the 60-way connector and taking it apart and reassembling it helped to make the connection again. not only that but my diagnostic light is now working again and flashes 12. im going to take that 60-way connector apart again and have a closer look at the contacts ... maybe spray some cleaner in it. thanks for the help heather .

From : neil nelson

heather hiscox hh@eui.doe.edu wrote hi well i have the meter hooked up in the cab and can see what the o2 sensor reads when i drive and most of the time it reads .9v. it occasionally shows .2v when i coast but never a medium voltage. this shows that the o2 sensor is doing its job and telling the ecu that things are too rich. sounds like the ecu cant compensate enough. what i dont know is whether the ecu is acting on this information to shorten the fuel pulses. does anyone know if there is a way to check this ... short of buying an oscilloscope thanks this can be determined with a scan tool by observing the adaptive fuel readings if theyre counting negative the ecu is trying to make corrections. if you can latch onto another tbi truck that -isnt -running rich comparative injector signal reading can be made with something as simple as a dwell meter... .