jmciver
09-17-2007, 12:36 AM
I posted this on another forum, but since I come here a lot I thought I would post it here as well.....
My little science project entails the following:
1. Install a wide band O2 sensor in the car using the stock headers.
2. Replace one stock O2 sensor to retain use of a stock "bung" location.
3. No O2 sensor related CELs
The hardware I chose to go with was the PLX Devices M-300 Wideband O2 Sensor controller which uses the common Bosch LSU 4.2 wideband O2 sensor. I chose this unit for the following reasons:
1. Moderately priced ($300).
2. The gauge display matched what I wanted to do in the interior of the car (big deciding factor here for me).
3. Controller is designed to send a 0-1V signal to the ECU, thus allowing removal of one of the stock pre-cat sensors.
When I turned the ignition on first, as expected, the PLX unit came on and indicated "Air" after ~40 seconds for the sensor to heat up. After starting the engine, I received "Lean" indications. However, this is expected since the stock air pump runs for the first few minutes after you start the car for emmisions reasons. Once the air pump turned off, A/F ratio went down to between 14-15 while at idle and very low throttle. No CELs.
After soldering all of the connections I started the car again. This time I got a CEL, the O2 Heater circuit fault for Bank 1, Sensor 1. This was expected since that is the sensor I replaced and it was still disconnected at this time.
According to the Bentley manual, the reason I didn't get the CEL the first time I started the car is because it takes at least two drive cycles for some faults to throw a CEL. The wideband "simulated" narrowband signal to the ECU continued to track with the other 3 stock narrowband voltage signals (the scantool I bought for my CarPC is finally starting to pay for itself :) ).
Next step is to tackle the O2 heater circuit problem. The way I see it, there are two ways to solve this problem:
1. Wire the heater terminals at the O2 sensor harness connector to a 20 ohm, 10 watt resistor (2 10 ohm, 10 watt resistors wired in series will do the trick as well).
2. Connect the stock O2 sensor, using the internal heating circuit, and cut the two signal wires. Then mount the stock O2 sensor some place in the engine compartment.
The problem with option 1 is that the resistors get hot. Therefore creativity must be used when both packaging them and mounting them to keep them from getting in contact with anyting that could possibly melt. You will also need some way of tapping the wires, doable, but not very pretty or easy.
I like option 2 better because I can use the stock O2 harness (no splicing necessary. The heating element is self contained within the O2 sensor, providing more flexibility with mounting. The bad thing of course is that the two signal wires to the stock sensor will have to be cut.
I decided to take option 2 and use the heating element inside my stock sensor. I ended up cutting two wires in the stock O2 sensor wiring (between the O2 sensor and the harness connector) that were the sensor power and sensor signal wires. That way the stock ECU still gets a O2 sensor heating element signal from a stock O2 sensor. If I ever wanted to re-use that stock sensor, all I would have to do is solder those wires back together.
I temporarily mounted the stock O2 sensor in the well area on the passenger side firewall. Since the O2 sensor gets hot due to the heating elements inside of it, I plan to get a bung and fabricate some type of bracket and use one of the existing screw holes inside the well. That way it will be permanently mounted in a way that will keep it from touching anything and allow the heat it generates to dissipate.
The verdict? No CELs, the ECU is happy since it gets the 0-1V O2 signal it wants, all wiring from the ECU to the stock O2 harness is in tact should I go back to stock (but I don't see that happening :D ), and the wideband does it's job.
As a whole I am pleased with the results. Getting the O2 sensor out, then back in was a serious PITA :mad , but obviously doable. E36 guys have been doing this trick for years (using a narrowband signal from a wideband controller to feed the stock ECU). It is good to know that the same trick works for the E46, at least the M3 anyway (in my research I did not find an E46 M3 that had done this before). So for all you E46 M3 guys who are boosted out there who want an A/F gauge but want to keep your stock headers (like me :cool ), there is an alternative.
My DIY website update with pictures of the install to follow soon. Installed pics of the display inside the car can be seen in this thread here (http://forums.bimmerforums.com/forum/showthread.php?t=826314).
My little science project entails the following:
1. Install a wide band O2 sensor in the car using the stock headers.
2. Replace one stock O2 sensor to retain use of a stock "bung" location.
3. No O2 sensor related CELs
The hardware I chose to go with was the PLX Devices M-300 Wideband O2 Sensor controller which uses the common Bosch LSU 4.2 wideband O2 sensor. I chose this unit for the following reasons:
1. Moderately priced ($300).
2. The gauge display matched what I wanted to do in the interior of the car (big deciding factor here for me).
3. Controller is designed to send a 0-1V signal to the ECU, thus allowing removal of one of the stock pre-cat sensors.
When I turned the ignition on first, as expected, the PLX unit came on and indicated "Air" after ~40 seconds for the sensor to heat up. After starting the engine, I received "Lean" indications. However, this is expected since the stock air pump runs for the first few minutes after you start the car for emmisions reasons. Once the air pump turned off, A/F ratio went down to between 14-15 while at idle and very low throttle. No CELs.
After soldering all of the connections I started the car again. This time I got a CEL, the O2 Heater circuit fault for Bank 1, Sensor 1. This was expected since that is the sensor I replaced and it was still disconnected at this time.
According to the Bentley manual, the reason I didn't get the CEL the first time I started the car is because it takes at least two drive cycles for some faults to throw a CEL. The wideband "simulated" narrowband signal to the ECU continued to track with the other 3 stock narrowband voltage signals (the scantool I bought for my CarPC is finally starting to pay for itself :) ).
Next step is to tackle the O2 heater circuit problem. The way I see it, there are two ways to solve this problem:
1. Wire the heater terminals at the O2 sensor harness connector to a 20 ohm, 10 watt resistor (2 10 ohm, 10 watt resistors wired in series will do the trick as well).
2. Connect the stock O2 sensor, using the internal heating circuit, and cut the two signal wires. Then mount the stock O2 sensor some place in the engine compartment.
The problem with option 1 is that the resistors get hot. Therefore creativity must be used when both packaging them and mounting them to keep them from getting in contact with anyting that could possibly melt. You will also need some way of tapping the wires, doable, but not very pretty or easy.
I like option 2 better because I can use the stock O2 harness (no splicing necessary. The heating element is self contained within the O2 sensor, providing more flexibility with mounting. The bad thing of course is that the two signal wires to the stock sensor will have to be cut.
I decided to take option 2 and use the heating element inside my stock sensor. I ended up cutting two wires in the stock O2 sensor wiring (between the O2 sensor and the harness connector) that were the sensor power and sensor signal wires. That way the stock ECU still gets a O2 sensor heating element signal from a stock O2 sensor. If I ever wanted to re-use that stock sensor, all I would have to do is solder those wires back together.
I temporarily mounted the stock O2 sensor in the well area on the passenger side firewall. Since the O2 sensor gets hot due to the heating elements inside of it, I plan to get a bung and fabricate some type of bracket and use one of the existing screw holes inside the well. That way it will be permanently mounted in a way that will keep it from touching anything and allow the heat it generates to dissipate.
The verdict? No CELs, the ECU is happy since it gets the 0-1V O2 signal it wants, all wiring from the ECU to the stock O2 harness is in tact should I go back to stock (but I don't see that happening :D ), and the wideband does it's job.
As a whole I am pleased with the results. Getting the O2 sensor out, then back in was a serious PITA :mad , but obviously doable. E36 guys have been doing this trick for years (using a narrowband signal from a wideband controller to feed the stock ECU). It is good to know that the same trick works for the E46, at least the M3 anyway (in my research I did not find an E46 M3 that had done this before). So for all you E46 M3 guys who are boosted out there who want an A/F gauge but want to keep your stock headers (like me :cool ), there is an alternative.
My DIY website update with pictures of the install to follow soon. Installed pics of the display inside the car can be seen in this thread here (http://forums.bimmerforums.com/forum/showthread.php?t=826314).