- Joined
- May 28, 2001
From over on the Bullet.
Wideband sensor discussion..A bit lenghty, but interesting, at least to me.
• If you experience a failed WBO2 sensor on a fairly new EFI installation (low miles), ensure the current WBO2 sensor location isn't causing damage by exhaust condensation thermally shocking the sensor at startup and/or during the warmup period. If you suspect it is, install a sensor bung at a better location, use a taller O2 sensor bung or use an Innovate Motorsports HBX-1 to move the sensor probe out of the direct exhaust stream. I've been successfully using HBX-1 Heat-Sink Bung Extenders for many years.
↑ This is referring to the "wet" gases flowing through the exhaust system while the engine is cold. On most vehicles, you can literally see water spitting out of the tailpipe(s). It's imperative to prevent this water from contacting the WBO2 sensor, because it will thermally shock the heated sensor. OEM engineers go to great lengths to locate the WBO2 at its optimum location. Anyway, a poor sensor location can actually blow water directly at the sensor probe. Look at the routing direction (angle) of the exhaust pipe just ahead of the sensor. Does it "direct" the water right at the sensor or away from it?"
Originally Posted by Klaus Allmendinger, VP of Engineering, Innovate Motorsports
Differences between Bosch LSU4/4.2 and NTK UEGO
1) Response speed:
Measured by the sensors own response delay between pump cell and sense cell-
Bosch ~5 msec
NTK ~50 msec
I'm referring to the internal delay of the sensor between a change of pump current to change of measurement cell output for a constant AFR gas, not the delay between gas change and measurement cell output. The delay I stated is the impulse response of the control input (pump current) to measurement output.
2) Back Pressure sensitivity:
Bosch sensor has about 1/3 the NTK's pressure sensitivity.
Pressure sensitivity means that the sensor reads richer than reality in a rich mixture, leaner in lean mixture.
3) Temperature sensitivity:
NTK sensor is fairly insensitive to temperature either at bung or EGT, can run with constant heater voltage.
Bosch sensors are very sensitive and needs to have precisely controlled cell temperatures. Bosch sensors are sensitive to housing temperature.
4) Heating time:
NTK ~60 seconds
Bosch ~20 seconds
Warmup times are greatly influenced by additional heating by exhaust gas and can be shorter than what I stated. The numbers I posted are warmup times for the sensor in 20°C (68°F) still gas, sensor at 20°C (68°F), heating to full operating temp (useful measurements). Warmup cycle controlled to max warmup ramp as per respective manufacturer specs. In an actual engine situation (additional heating by exhaust gas) the Bosch can warm up in as low as 10 sec while still staying within the warmup ramp specs. The NTK can warmup in 33 secs or less. But these numbers are very much dependent on the engine situation (rpm, EGT) and therefore hard to compare and control.
5) Thermal shock:
There's no difference. The Bosch sensor is actually shrouded heavier than the NTK, which protects it a little better. The only difference I have seen is that the NTK often does not fail as dramatically and obviously as the Bosch, but can give you readings that are 1-3 AFR off.
This generally true from my experience with the NTK. It fails just as often and for the same reasons as the Bosch unit, but the failures are less noticeable. But it IS a lot slower due to its construction and can handle more mechanical abuse (banging it around).
Differences between Bosch LSU4 (066) and LSU4.2 (057/05
1) Heater response time:
066 sensor has higher thermal mass and responds slower to heat input. This makes the 057/058 sensor a little bit more challenging to control, heater PID must react faster.
No problem in the LM-1, because it's already designed for the faster response of the 057.
2) Sensor tolerances:
066 sensors have tighter tolerances between sensors. This is not an issue with the LM-1 because it is calibrates to the individual sensor when doing a free air calibration.
3) 066 sensor is slightly more tolerant to overheating of the sensor housing at the bung. It takes longer to give completely erroneous values, but has the same specs.
4) 066 sensor is more expensive.
5) Connector is different.
Heat & Back Pressure Aspects:
It's all really simple. A narrowband sensor does not care about heat or back-pressure very much. It just needs to be at it's minimum temp (~300°C/572°F) to operate. It switches at 14.7 AFR between a low voltage (lean) and high voltage (rich). That's all it does. If people post that they have different readings on an NB sensor before and after turbo, it's because they use the NB sensor for what it can't do, namely measure AFR outside it's 14.7 +- ~0.3 AFR range. It's output voltage on the rich side just varies mainly with EGT (and just a little with AFR), that's all.
A wideband sensor IS sensitive to back pressure. Some more, some less. With back pressure a WB reads richer than reality on the rich side of stoichiometric, leaner than reality on the lean side. The NTK sensor is actually much more sensitive in that respect than the Bosch. The NTK sensor on the other hand is less sensitive to heat, and could theoretically take the pre-turbo heat. But the ECU needs to compensate for the wrong reading due to back pressure. OEMs do that by finding the appropriate compensation factors based on engine state.
The gist of it is that you can mount a narrowband sensor before the turbo, because it can take the heat and is ignored at WOT anyway. But even a heated NBO2 does not have a heater strong enough without help from EGTs to keep at it's operating temp in all conditions.
Mount a WB sensor downstream, for accurate measurement and sensor life. In the location designed for a NBO2 you will typically be asking for trouble.
Regarding Bosch LSU4.2 Location:
The Bosch LSU4.2 sensor has a specified housing temperature (at the bung) of max 560°C (1040°F). Exceeding this can cause problems because the heater in the sensor can no longer be precisely controlled. It does not destroy the sensor (typically), but AFR readings will be inaccurate because of uncontrolled sensor temperatures.
On many turbo cars the bung temperatures are higher than that. Same can happen with superchargers, wrapped/coated headers and pipes.
Narrowband O2 sensors are not very sensitive to heat, because they only need to work as a switch, not as a measurement device.
In many cases the bung temperatures of a NBO2 sensor location are much higher than the Bosch sensor can tolerate.
We (Innovate) found that although the sensor head can handle up to 1560°F EGT, bung temperatures are typically the bigger problem. That's why we recommend a heat sink for high bung temp applications.
The sensor head is temperature controlled. If heated to or above its operating temperature it can no longer be controlled (the sensor heater cannot cool). In that case in an OEM application for the sensor (closed loop WB control) the ECU goes open loop. A WB meter cannot go open loop (unfortunately) and either becomes inaccurate or shows an error (the LM-1 shows an error).
• As for warmup of the sensor:
This is key to the lifespan of the sensor. The LM-1 uses a controlled heat-up profile at the max. allowable heat-up rate that Bosch specifies for the sensor. Heat-up time in room temperature air is about 20 sec.
Sensors get destroyed for four reasons:
1) Carbon fouling
Happens when the sensor is left unpowered during engine warmup or running continuously at excessively rich mixtures (<10 AFR).
2) Lead fouling
Lead will at any temperature over time coat the pump cell ceramics and prevent it from working.
3) Penetrants
Things like WD-40, even traces of it, will destroy the sensor instantly because of a chemical reaction between the penetrant and the sensor ceramics.
The sensors get their reference air through the cable sheath. Penetrants can work their way through the cable into the sensor. After all, that's what penetrants are for.
4) Running the sensor outside specified temperatures.
This does not destroy the sensor instantly, but reduces its lifespan significantly.
Regards, Klaus"
Wideband sensor discussion..A bit lenghty, but interesting, at least to me.
• If you experience a failed WBO2 sensor on a fairly new EFI installation (low miles), ensure the current WBO2 sensor location isn't causing damage by exhaust condensation thermally shocking the sensor at startup and/or during the warmup period. If you suspect it is, install a sensor bung at a better location, use a taller O2 sensor bung or use an Innovate Motorsports HBX-1 to move the sensor probe out of the direct exhaust stream. I've been successfully using HBX-1 Heat-Sink Bung Extenders for many years.
↑ This is referring to the "wet" gases flowing through the exhaust system while the engine is cold. On most vehicles, you can literally see water spitting out of the tailpipe(s). It's imperative to prevent this water from contacting the WBO2 sensor, because it will thermally shock the heated sensor. OEM engineers go to great lengths to locate the WBO2 at its optimum location. Anyway, a poor sensor location can actually blow water directly at the sensor probe. Look at the routing direction (angle) of the exhaust pipe just ahead of the sensor. Does it "direct" the water right at the sensor or away from it?"
Differences between Bosch LSU4/4.2 and NTK UEGO
1) Response speed:
Measured by the sensors own response delay between pump cell and sense cell-
Bosch ~5 msec
NTK ~50 msec
I'm referring to the internal delay of the sensor between a change of pump current to change of measurement cell output for a constant AFR gas, not the delay between gas change and measurement cell output. The delay I stated is the impulse response of the control input (pump current) to measurement output.
2) Back Pressure sensitivity:
Bosch sensor has about 1/3 the NTK's pressure sensitivity.
Pressure sensitivity means that the sensor reads richer than reality in a rich mixture, leaner in lean mixture.
3) Temperature sensitivity:
NTK sensor is fairly insensitive to temperature either at bung or EGT, can run with constant heater voltage.
Bosch sensors are very sensitive and needs to have precisely controlled cell temperatures. Bosch sensors are sensitive to housing temperature.
4) Heating time:
NTK ~60 seconds
Bosch ~20 seconds
Warmup times are greatly influenced by additional heating by exhaust gas and can be shorter than what I stated. The numbers I posted are warmup times for the sensor in 20°C (68°F) still gas, sensor at 20°C (68°F), heating to full operating temp (useful measurements). Warmup cycle controlled to max warmup ramp as per respective manufacturer specs. In an actual engine situation (additional heating by exhaust gas) the Bosch can warm up in as low as 10 sec while still staying within the warmup ramp specs. The NTK can warmup in 33 secs or less. But these numbers are very much dependent on the engine situation (rpm, EGT) and therefore hard to compare and control.
5) Thermal shock:
There's no difference. The Bosch sensor is actually shrouded heavier than the NTK, which protects it a little better. The only difference I have seen is that the NTK often does not fail as dramatically and obviously as the Bosch, but can give you readings that are 1-3 AFR off.
This generally true from my experience with the NTK. It fails just as often and for the same reasons as the Bosch unit, but the failures are less noticeable. But it IS a lot slower due to its construction and can handle more mechanical abuse (banging it around).
Differences between Bosch LSU4 (066) and LSU4.2 (057/05
1) Heater response time:
066 sensor has higher thermal mass and responds slower to heat input. This makes the 057/058 sensor a little bit more challenging to control, heater PID must react faster.
No problem in the LM-1, because it's already designed for the faster response of the 057.
2) Sensor tolerances:
066 sensors have tighter tolerances between sensors. This is not an issue with the LM-1 because it is calibrates to the individual sensor when doing a free air calibration.
3) 066 sensor is slightly more tolerant to overheating of the sensor housing at the bung. It takes longer to give completely erroneous values, but has the same specs.
4) 066 sensor is more expensive.
5) Connector is different.
Heat & Back Pressure Aspects:
It's all really simple. A narrowband sensor does not care about heat or back-pressure very much. It just needs to be at it's minimum temp (~300°C/572°F) to operate. It switches at 14.7 AFR between a low voltage (lean) and high voltage (rich). That's all it does. If people post that they have different readings on an NB sensor before and after turbo, it's because they use the NB sensor for what it can't do, namely measure AFR outside it's 14.7 +- ~0.3 AFR range. It's output voltage on the rich side just varies mainly with EGT (and just a little with AFR), that's all.
A wideband sensor IS sensitive to back pressure. Some more, some less. With back pressure a WB reads richer than reality on the rich side of stoichiometric, leaner than reality on the lean side. The NTK sensor is actually much more sensitive in that respect than the Bosch. The NTK sensor on the other hand is less sensitive to heat, and could theoretically take the pre-turbo heat. But the ECU needs to compensate for the wrong reading due to back pressure. OEMs do that by finding the appropriate compensation factors based on engine state.
The gist of it is that you can mount a narrowband sensor before the turbo, because it can take the heat and is ignored at WOT anyway. But even a heated NBO2 does not have a heater strong enough without help from EGTs to keep at it's operating temp in all conditions.
Mount a WB sensor downstream, for accurate measurement and sensor life. In the location designed for a NBO2 you will typically be asking for trouble.
Regarding Bosch LSU4.2 Location:
The Bosch LSU4.2 sensor has a specified housing temperature (at the bung) of max 560°C (1040°F). Exceeding this can cause problems because the heater in the sensor can no longer be precisely controlled. It does not destroy the sensor (typically), but AFR readings will be inaccurate because of uncontrolled sensor temperatures.
On many turbo cars the bung temperatures are higher than that. Same can happen with superchargers, wrapped/coated headers and pipes.
Narrowband O2 sensors are not very sensitive to heat, because they only need to work as a switch, not as a measurement device.
In many cases the bung temperatures of a NBO2 sensor location are much higher than the Bosch sensor can tolerate.
We (Innovate) found that although the sensor head can handle up to 1560°F EGT, bung temperatures are typically the bigger problem. That's why we recommend a heat sink for high bung temp applications.
The sensor head is temperature controlled. If heated to or above its operating temperature it can no longer be controlled (the sensor heater cannot cool). In that case in an OEM application for the sensor (closed loop WB control) the ECU goes open loop. A WB meter cannot go open loop (unfortunately) and either becomes inaccurate or shows an error (the LM-1 shows an error).
• As for warmup of the sensor:
This is key to the lifespan of the sensor. The LM-1 uses a controlled heat-up profile at the max. allowable heat-up rate that Bosch specifies for the sensor. Heat-up time in room temperature air is about 20 sec.
Sensors get destroyed for four reasons:
1) Carbon fouling
Happens when the sensor is left unpowered during engine warmup or running continuously at excessively rich mixtures (<10 AFR).
2) Lead fouling
Lead will at any temperature over time coat the pump cell ceramics and prevent it from working.
3) Penetrants
Things like WD-40, even traces of it, will destroy the sensor instantly because of a chemical reaction between the penetrant and the sensor ceramics.
The sensors get their reference air through the cable sheath. Penetrants can work their way through the cable into the sensor. After all, that's what penetrants are for.
4) Running the sensor outside specified temperatures.
This does not destroy the sensor instantly, but reduces its lifespan significantly.
Regards, Klaus"
