Wideband issues... what would cause this?
04-12-2012, 10:25 PM
#1
4th Gear Member
Thread Starter
Join Date: Nov 2006
Location:
Posts: 1,229
Wideband issues... what would cause this?
Ok this isn't on my Mustang, but it's all the same. If it matters it's on my 93 Taurus SHO. I have a used PLX M250 wideband that I just put in. I drove the car a few miles up and down my road (sensor was working and reading 12-14) and then the wideband read full lean (I thought exhaust leak) so I took the car home checked it out and have no leak. So I let it sit for a couple weeks. This past weekend I fire the car up and the wideband is working perfect again.
Currently the car is running pig rich (not tuned yet) and I also mounted the wideband probably 14-18 inches from the head near my factory o2's. I later read that the wideband should be about 24" from the head as anything closer can cause premature failure of the sensor. Is this true and or do you think this is why the sensor went full lean on me and just stayed there (possibly overheated and tweaked)? The car's been a work in progress for almost 6 years and I would like this summer to be the one, so any input in appreciated.
TIA
Currently the car is running pig rich (not tuned yet) and I also mounted the wideband probably 14-18 inches from the head near my factory o2's. I later read that the wideband should be about 24" from the head as anything closer can cause premature failure of the sensor. Is this true and or do you think this is why the sensor went full lean on me and just stayed there (possibly overheated and tweaked)? The car's been a work in progress for almost 6 years and I would like this summer to be the one, so any input in appreciated.
TIA
04-13-2012, 04:18 AM
#2
TECH SAVANT
Join Date: Dec 2006
Location: Saint Augustine, FL
Posts: 10,938
The Bosch LSU4 sensor specified maximum temperatures are 930°C (1706°F) at the sensor tip, and 570°C (1058°F) at the body hex. In practice my experience has been that anything above 900°F at the sensor body will cause it to behave erratically.
The options are to use a longer bung (Innovate ships a 1" stainless steel¹ bung with their units), or use their absurdly priced heat sink, or fashion a heat sink to go between the sensor body and the bung.
-------------------------------------------------
¹ - Stainless steel has much lower thermal conductivity than mild steel.
The options are to use a longer bung (Innovate ships a 1" stainless steel¹ bung with their units), or use their absurdly priced heat sink, or fashion a heat sink to go between the sensor body and the bung.
-------------------------------------------------
¹ - Stainless steel has much lower thermal conductivity than mild steel.
04-15-2012, 12:01 AM
#3
4th Gear Member
Thread Starter
Join Date: Nov 2006
Location:
Posts: 1,229
Interesting stuff. Would I be better off plugging the hole I have now and putting another bung on 24 inches down my pipe? I just want to be sure heat is my problem before I go through all this. I should try to find someone with a heat sensing gun to see if the body is actually getting that hot.
04-15-2012, 06:11 AM
#4
TECH SAVANT
Join Date: Dec 2006
Location: Saint Augustine, FL
Posts: 10,938
What you described "(sensor was working and reading 12-14) and then the wideband read full lean" is exactly how an overheated WB sensor would behave.
Most inexpensive (I.e. < $150) IR thermometers won't read temperatures that high, though some will. A high-end infrared thermometer would do it...
Most inexpensive (I.e. < $150) IR thermometers won't read temperatures that high, though some will. A high-end infrared thermometer would do it...
Last edited by cliffyk; 04-15-2012 at 06:14 AM.
04-15-2012, 11:08 AM
#5
4th Gear Member
Thread Starter
Join Date: Nov 2006
Location:
Posts: 1,229
What you described "(sensor was working and reading 12-14) and then the wideband read full lean" is exactly how an overheated WB sensor would behave.
Most inexpensive (I.e. < $150) IR thermometers won't read temperatures that high, though some will. A high-end infrared thermometer would do it...
Most inexpensive (I.e. < $150) IR thermometers won't read temperatures that high, though some will. A high-end infrared thermometer would do it...
Also side note, why do the narrow bands work ok so close to the head but widebands cannot?
04-15-2012, 12:16 PM
#7
TECH SAVANT
Join Date: Dec 2006
Location: Saint Augustine, FL
Posts: 10,938
Damn. Kinda had a feeling. I wasn't thinking when I welded the bung on. I just ***-U-MEed it should go where the factory sensors were so that's where I put it. Do you think I should just plug the hole I have now and move it further down stream or should I try some of those methods you sent above? Do they work well or is it really better to move the sensor down where it should be?
Also side note, why do the narrow bands work ok so close to the head but widebands cannot?
The processor uses the output of the NB sensor to see if a sample of the exhaust gas (held in a small chamber) is rich or lean. It then adjust the voltage (level and polarity) and current to the pump cell to cause it to add or remove O² from the sample chamber, until the NB sensor reports the sample as being stoichiometric--I.e. lambda (λ) 1.0 or 14.7:1 AFR for gasoline.
Because the processor knows what voltage (of what polarity) and what current (for how long) was required to make the sample λ1.0 it also knows what the λ of the original sample was--and it generates a voltage output (often 0-5V) proportional to the exhaust gas' original AFR...
04-15-2012, 06:18 PM
#8
4th Gear Member
Thread Starter
Join Date: Nov 2006
Location:
Posts: 1,229
I would try the heat sink first, you can get a 6" x 6" piece of 14 ga here for $11.00--I can't imagine the shipping could be too much...
A wideband sensor contains a narrow band sensor and an electrochemical oxygen "pump" cell, and requires a microprocessor controller to function. It's the characteristics of the oxygen pump that constrains the sensor's maximum operating temperature, as compared to a NB sensor.
The processor uses the output of the NB sensor to see if a sample of the exhaust gas (held in a small chamber) is rich or lean. It then adjust the voltage (level and polarity) and current to the pump cell to cause it to add or remove O² from the sample chamber, until the NB sensor reports the sample as being stoichiometric--I.e. lambda (λ) 1.0 or 14.7:1 AFR for gasoline.
Because the processor knows what voltage (of what polarity) and what current (for how long) was required to make the sample λ1.0 it also knows what the λ of the original sample was--and it generates a voltage output (often 0-5V) proportional to the exhaust gas' original AFR...
A wideband sensor contains a narrow band sensor and an electrochemical oxygen "pump" cell, and requires a microprocessor controller to function. It's the characteristics of the oxygen pump that constrains the sensor's maximum operating temperature, as compared to a NB sensor.
The processor uses the output of the NB sensor to see if a sample of the exhaust gas (held in a small chamber) is rich or lean. It then adjust the voltage (level and polarity) and current to the pump cell to cause it to add or remove O² from the sample chamber, until the NB sensor reports the sample as being stoichiometric--I.e. lambda (λ) 1.0 or 14.7:1 AFR for gasoline.
Because the processor knows what voltage (of what polarity) and what current (for how long) was required to make the sample λ1.0 it also knows what the λ of the original sample was--and it generates a voltage output (often 0-5V) proportional to the exhaust gas' original AFR...
04-15-2012, 06:28 PM
#9
TECH SAVANT
Join Date: Dec 2006
Location: Saint Augustine, FL
Posts: 10,938
45+ years of mechanical engineering did the trick...
Thread
Thread Starter
Forum
Replies
Last Post
4ValveGrabber
4.6L (1996-2004 Modular) Mustang
5
07-13-2006 01:29 AM
callmeebryan@tps
Parts For Sale
1
02-13-2006 09:37 PM
