Open Loop Tuning W/O Wide Band
07-14-2004, 05:35 PM
#61
Originally Posted by Brains
Oh damn, yea I forgot.. Dynos are now obsolete, since we have those extended PID's .. 
(never mind that mine says the motor makes 365HP and runs mid 11's at 3720 lbs. N/A)
(also never mind what the PIDs show with a nitrous motor)
(never mind that mine says the motor makes 365HP and runs mid 11's at 3720 lbs. N/A)
(also never mind what the PIDs show with a nitrous motor)
07-14-2004, 06:46 PM
#62
TECH Addict
Join Date: Jun 2002
Location: Las Vegas, NV
Posts: 2,428
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by Brains
Oh damn, yea I forgot.. Dynos are now obsolete, since we have those extended PID's ..
(never mind that mine says the motor makes 365HP and runs mid 11's at 3720 lbs. N/A)
(also never mind what the PIDs show with a nitrous motor)
(never mind that mine says the motor makes 365HP and runs mid 11's at 3720 lbs. N/A)
(also never mind what the PIDs show with a nitrous motor)
What a coincidence that JR and I are the only EASE users in this thread.
07-14-2004, 07:36 PM
#63
TECH Senior Member
iTrader: (7)
Join Date: Jan 2002
Location: Katy, TX
Posts: 12,754
Likes: 0
Received 0 Likes
on
0 Posts
Ease is a good scan tool package, no doubt about that... But lets not get JR jumping up and down like he did on the LS1Edit list about how its so good it'll make your coffee in the morning too (and how everything else sucks).
I dug up your sig, because I was TRULY curious as to what setup you were tuning that you were able to get a successful result with so little "outside" data. I wagered it'd be quite close to stock, and I was right I could also tune your car equally as well as you have w/o using a wideband -- because the factory tables still match the overall state of the motor. Adjusting for the larger injectors is quite simple - its a linear adjustment, simply scale up/down until the LTrims come back into line. The rest of your additions (LS6 intake, headers, etc.) don't skew your tune to the point of running poorly at all. You must be turning some pretty serious RPM to approach 100% DC with stock injectors @ 326RWHP, and the stock cam definitely isn't making power up in those RPM's
I don't doubt for a second that your car runs well -- and rightfully it SHOULD !! The point is, just because those methods worked on your car it doesn't mean they'll work on every car. There's a LOT of very experienced folks in this thread (moreso than JR) that have tuned combinations far in excess of what he has (or I have). Its pretty pig-headed to just shoot them down with meager attempt to discredit (ie. picking subsets of quotes from GM tech manuals). You have to remember, those manuals are written on a KNOWN setup -- they designed and tuned it, and wrote the manual AFTERWARDS Once we go and change all that, that may or may not apply in part or in entirety.
I dug up your sig, because I was TRULY curious as to what setup you were tuning that you were able to get a successful result with so little "outside" data. I wagered it'd be quite close to stock, and I was right
I could also tune your car equally as well as you have w/o using a wideband -- because the factory tables still match the overall state of the motor. Adjusting for the larger injectors is quite simple - its a linear adjustment, simply scale up/down until the LTrims come back into line. The rest of your additions (LS6 intake, headers, etc.) don't skew your tune to the point of running poorly at all. You must be turning some pretty serious RPM to approach 100% DC with stock injectors @ 326RWHP, and the stock cam definitely isn't making power up in those RPM's I don't doubt for a second that your car runs well -- and rightfully it SHOULD !! The point is, just because those methods worked on your car it doesn't mean they'll work on every car. There's a LOT of very experienced folks in this thread (moreso than JR) that have tuned combinations far in excess of what he has (or I have). Its pretty pig-headed to just shoot them down with meager attempt to discredit (ie. picking subsets of quotes from GM tech manuals). You have to remember, those manuals are written on a KNOWN setup -- they designed and tuned it, and wrote the manual AFTERWARDS
Once we go and change all that, that may or may not apply in part or in entirety.
07-14-2004, 08:28 PM
#64
TECH Enthusiast
Join Date: Apr 2003
Location: silver Spring MD
Posts: 561
Likes: 0
Received 0 Likes
on
0 Posts
Hi folks
I have to agree with Brains and also point out there are people like me - just trying to learn how to tune and need all the help we can get, including any widget we can get to make the job easier. With the amount of money I spend on software and now a WB O2 sensor, I could have taken the car to a tuner and have the car done 3 times but that is no fun. I don’t plan on doing this for anybody else but will help once I am knowledgeable. This is my hobby and very learning experience for me.
I really can’t believe how people can not see that the WB will help some people but maybe other like JR does not need it. I guess the biggest disappointment with all the comments is there people who still don’t think a WB O2 is useful or would not concede the point. With that closed approach you will be limited in life.
Bill
I have to agree with Brains and also point out there are people like me - just trying to learn how to tune and need all the help we can get, including any widget we can get to make the job easier. With the amount of money I spend on software and now a WB O2 sensor, I could have taken the car to a tuner and have the car done 3 times but that is no fun. I don’t plan on doing this for anybody else but will help once I am knowledgeable. This is my hobby and very learning experience for me.
I really can’t believe how people can not see that the WB will help some people but maybe other like JR does not need it. I guess the biggest disappointment with all the comments is there people who still don’t think a WB O2 is useful or would not concede the point. With that closed approach you will be limited in life.
Bill
07-14-2004, 09:28 PM
#65
On The Tree
Join Date: Oct 2003
Posts: 175
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by Team ZR-1
LOW OCTANE timing table then you should not be mucking around with the PCM for its simple knowledge that the C5 has a bad gas (low octane) protection and resets that function when gas has been added.
FACT research GM service documents, they even clearly say not to do certain tests when the fuel is under a quarter of a tank for the PCM functons differently when low fuel or bad fuel conditions.
FACT research GM service documents, they even clearly say not to do certain tests when the fuel is under a quarter of a tank for the PCM functons differently when low fuel or bad fuel conditions.
Do us all a favour and read your GM notes or EASE manual on the term "Octane Adaptive Scaler".
07-16-2004, 03:45 PM
#66
TECH Fanatic
Join Date: Mar 2003
Location: La Porte, TX
Posts: 1,839
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by Bink
Anyone remember the LS1Edit mailing list fiascos?
Yes, I do remember a couple of them.
Do you remember the 5 gallons of gas to throw the PCM back to the High Octane table?? Fetumpsh was tenacious as a pit bull on dispelling that one.
Excellent, accurate and persuasive info.
Thanks guys.
joel
Yes, I do remember a couple of them.
Do you remember the 5 gallons of gas to throw the PCM back to the High Octane table?? Fetumpsh was tenacious as a pit bull on dispelling that one.
Excellent, accurate and persuasive info.
Thanks guys. joelSo what was the definitive answer behind that one???? Is is it really a myth, or does an empty tank cause the car to go into the low octane table?
07-16-2004, 07:37 PM
#67
On The Tree
Join Date: Oct 2003
Posts: 175
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by WS6snake-eater
So what was the definitive answer behind that one???? Is is it really a myth, or does an empty tank cause the car to go into the low octane table?
Visual Proof no such thing as the empty tank syndrome exists - http://www.efilive.com/forum/viewtopic.php?t=282
Remember, some LS1 applications don't even monitor fuel levels, yet run twin spark maps.
03-27-2005, 11:53 PM
#68
Wow! Such anymosity !
Maybe it would be helpful to explain how these things ( Oxygen sensors )actually work from the insde ( for those who don't already know )? Maybe I'll just make things worse.
The narrowband Oxygen sensors work on the Nernst priciple. They are not AFR sensors, or even Lambda sensors. They just sense oxygen molecules at their sensor membrane (the "Nernst cell"). One side of the membrane is exposed to free air. the other side to the exhaust. If there is no free oxygen in the exhaust, some is pulled through the membrane. These molecules crossing the membrane cause a voltage ( or charge, more properly ) according to the Nernst equation. It's kinda like a battery. As long as there is no "load" on it, it doen't matter so much just how many molecules are moving.
The basic idea is, if there is any free oxygen in the the exhaust at all, you are running lean of stochiometric. These cells are, ideally, switches; either on or off. 0volts if there is no Oxygen present and a voltage, dependant on temperature and pressure, if it is. Generally this "full scale" voltage is somewhere in the neighborhood of 1.0v but it can be 700mv to 1300mv depending on temperature and pressure. Since we don't live in a laboratory, the switching is not quite a perfect on/off, with nothign in between. So, there is a region called an ambigious region that surrounds the switching point. This type of curve is called an instability about a point. So, at 14.7 AFR, the point where there is no more oxygen in the exhaust, the output of the Nernst cell wanders around the midpoint of it's output range. Usually about 450-500mv. As you move away from that point, the output quickly saturates twords either 0v or full scale output. In theory, if the Nernst cell in the sensor was only one molecule in size, the sensor would switch perfectly from 0v to 1v ( or whatever the max is based on temp and presure ) and there would be no ambigious range. The percieved "linear range" is just the effect of different molecules of the Nernst cell fighting with eachother. Some see an Oxygen molecule, some don't.
That's, partly, why the range is unstable. Changes in temperature change the "full scale" output. Changes in pressure affect how many molecules of the Nernst cell actually see Oxygen molecules. If you are a Chem major ( and are bored and out of beer ), you might want to look up the Nernst equations. or not.
http://www.science.uwaterloo.ca/~cch.../nernsteq.html
Temperature controlled narrowbands do a little better, as the full scale voltage is more predictable. But they still have the instability and pressure sensitivities. And that varies from part to part.
Widebands kinda cheat. A wideband sensor has a narrowband sensor inside of it. It also has what is called a "pump cell". ( that's the 5th wire on a 5-wire sensor) This is a ceramic membrane that will actually "pump" Oxygen ions across it based on electrical current. I works both ways. that is, put a positive voltage across it, and it pumps oxygen ions into the Nernst cell. Put a negative voltage and it pumps Oxygen ions out of the cell. The amount of oxygen it pumps is proportional ( more or less ) to the electrical current applied. It is acually the same type of material as the Nernst cell.
The Wideband system constantly tries to make the Nersnt cell read a perfect 14.7 AFR by using the pump cell to either add Oxygen to burn off unburnt fuel or pumping extra oxygen out of a lean exhaust. The amount of current it takes to get the Nernst cell back to 14.7 ( where it is predictable )determines the actual AFR. Thats why you can get a wide stable linear range from a wideband unit; even though the Nernst cell itself can not resolve values very far from 14.7AFR. The Nernst cell in a wideband sensor is ALWAYS at 14.7AFR. It's the pump cell that is linear.
You can kinda do the same thing by injecting air or fuel into the exhaust stream right before a narrowband sensor. If you add fresh air, the sensor will think its leaner than it is and the ECU will try to richen up the mixture to burn off the extra oxygen you are adding. If you add fuel, the opposite happens. The ECU leans out the mixture until there is just enough free oxygen in the exhaust to burn of the extra fuel. In both cases, its trying to get back to 14.7, as read by the Nernst cell; just like a wideband.
I'm sleepy now....
Maybe it would be helpful to explain how these things ( Oxygen sensors )actually work from the insde ( for those who don't already know )? Maybe I'll just make things worse.
The narrowband Oxygen sensors work on the Nernst priciple. They are not AFR sensors, or even Lambda sensors. They just sense oxygen molecules at their sensor membrane (the "Nernst cell"). One side of the membrane is exposed to free air. the other side to the exhaust. If there is no free oxygen in the exhaust, some is pulled through the membrane. These molecules crossing the membrane cause a voltage ( or charge, more properly ) according to the Nernst equation. It's kinda like a battery. As long as there is no "load" on it, it doen't matter so much just how many molecules are moving.
The basic idea is, if there is any free oxygen in the the exhaust at all, you are running lean of stochiometric. These cells are, ideally, switches; either on or off. 0volts if there is no Oxygen present and a voltage, dependant on temperature and pressure, if it is. Generally this "full scale" voltage is somewhere in the neighborhood of 1.0v but it can be 700mv to 1300mv depending on temperature and pressure. Since we don't live in a laboratory, the switching is not quite a perfect on/off, with nothign in between. So, there is a region called an ambigious region that surrounds the switching point. This type of curve is called an instability about a point. So, at 14.7 AFR, the point where there is no more oxygen in the exhaust, the output of the Nernst cell wanders around the midpoint of it's output range. Usually about 450-500mv. As you move away from that point, the output quickly saturates twords either 0v or full scale output. In theory, if the Nernst cell in the sensor was only one molecule in size, the sensor would switch perfectly from 0v to 1v ( or whatever the max is based on temp and presure ) and there would be no ambigious range. The percieved "linear range" is just the effect of different molecules of the Nernst cell fighting with eachother. Some see an Oxygen molecule, some don't.
That's, partly, why the range is unstable. Changes in temperature change the "full scale" output. Changes in pressure affect how many molecules of the Nernst cell actually see Oxygen molecules. If you are a Chem major ( and are bored and out of beer ), you might want to look up the Nernst equations. or not.
http://www.science.uwaterloo.ca/~cch.../nernsteq.html
Temperature controlled narrowbands do a little better, as the full scale voltage is more predictable. But they still have the instability and pressure sensitivities. And that varies from part to part.
Widebands kinda cheat. A wideband sensor has a narrowband sensor inside of it. It also has what is called a "pump cell". ( that's the 5th wire on a 5-wire sensor) This is a ceramic membrane that will actually "pump" Oxygen ions across it based on electrical current. I works both ways. that is, put a positive voltage across it, and it pumps oxygen ions into the Nernst cell. Put a negative voltage and it pumps Oxygen ions out of the cell. The amount of oxygen it pumps is proportional ( more or less ) to the electrical current applied. It is acually the same type of material as the Nernst cell.
The Wideband system constantly tries to make the Nersnt cell read a perfect 14.7 AFR by using the pump cell to either add Oxygen to burn off unburnt fuel or pumping extra oxygen out of a lean exhaust. The amount of current it takes to get the Nernst cell back to 14.7 ( where it is predictable )determines the actual AFR. Thats why you can get a wide stable linear range from a wideband unit; even though the Nernst cell itself can not resolve values very far from 14.7AFR. The Nernst cell in a wideband sensor is ALWAYS at 14.7AFR. It's the pump cell that is linear.
You can kinda do the same thing by injecting air or fuel into the exhaust stream right before a narrowband sensor. If you add fresh air, the sensor will think its leaner than it is and the ECU will try to richen up the mixture to burn off the extra oxygen you are adding. If you add fuel, the opposite happens. The ECU leans out the mixture until there is just enough free oxygen in the exhaust to burn of the extra fuel. In both cases, its trying to get back to 14.7, as read by the Nernst cell; just like a wideband.
I'm sleepy now....
03-28-2005, 05:07 PM
#70
Originally Posted by dr.mike
Wow! Such anymosity !
Maybe it would be helpful to explain how these things ( Oxygen sensors )actually work from the insde ( for those who don't already know )? Maybe I'll just make things worse.
The narrowband Oxygen sensors work on the Nernst priciple. They are not AFR sensors, or even Lambda sensors. They just sense oxygen molecules at their sensor membrane (the "Nernst cell"). One side of the membrane is exposed to free air. the other side to the exhaust. If there is no free oxygen in the exhaust, some is pulled through the membrane. These molecules crossing the membrane cause a voltage ( or charge, more properly ) according to the Nernst equation. It's kinda like a battery. As long as there is no "load" on it, it doen't matter so much just how many molecules are moving.
The basic idea is, if there is any free oxygen in the the exhaust at all, you are running lean of stochiometric. These cells are, ideally, switches; either on or off. 0volts if there is no Oxygen present and a voltage, dependant on temperature and pressure, if it is. Generally this "full scale" voltage is somewhere in the neighborhood of 1.0v but it can be 700mv to 1300mv depending on temperature and pressure. Since we don't live in a laboratory, the switching is not quite a perfect on/off, with nothign in between. So, there is a region called an ambigious region that surrounds the switching point. This type of curve is called an instability about a point. So, at 14.7 AFR, the point where there is no more oxygen in the exhaust, the output of the Nernst cell wanders around the midpoint of it's output range. Usually about 450-500mv. As you move away from that point, the output quickly saturates twords either 0v or full scale output. In theory, if the Nernst cell in the sensor was only one molecule in size, the sensor would switch perfectly from 0v to 1v ( or whatever the max is based on temp and presure ) and there would be no ambigious range. The percieved "linear range" is just the effect of different molecules of the Nernst cell fighting with eachother. Some see an Oxygen molecule, some don't.
That's, partly, why the range is unstable. Changes in temperature change the "full scale" output. Changes in pressure affect how many molecules of the Nernst cell actually see Oxygen molecules. If you are a Chem major ( and are bored and out of beer ), you might want to look up the Nernst equations. or not.
http://www.science.uwaterloo.ca/~cch.../nernsteq.html
Temperature controlled narrowbands do a little better, as the full scale voltage is more predictable. But they still have the instability and pressure sensitivities. And that varies from part to part.
Widebands kinda cheat. A wideband sensor has a narrowband sensor inside of it. It also has what is called a "pump cell". ( that's the 5th wire on a 5-wire sensor) This is a ceramic membrane that will actually "pump" Oxygen ions across it based on electrical current. I works both ways. that is, put a positive voltage across it, and it pumps oxygen ions into the Nernst cell. Put a negative voltage and it pumps Oxygen ions out of the cell. The amount of oxygen it pumps is proportional ( more or less ) to the electrical current applied. It is acually the same type of material as the Nernst cell.
The Wideband system constantly tries to make the Nersnt cell read a perfect 14.7 AFR by using the pump cell to either add Oxygen to burn off unburnt fuel or pumping extra oxygen out of a lean exhaust. The amount of current it takes to get the Nernst cell back to 14.7 ( where it is predictable )determines the actual AFR. Thats why you can get a wide stable linear range from a wideband unit; even though the Nernst cell itself can not resolve values very far from 14.7AFR. The Nernst cell in a wideband sensor is ALWAYS at 14.7AFR. It's the pump cell that is linear.
You can kinda do the same thing by injecting air or fuel into the exhaust stream right before a narrowband sensor. If you add fresh air, the sensor will think its leaner than it is and the ECU will try to richen up the mixture to burn off the extra oxygen you are adding. If you add fuel, the opposite happens. The ECU leans out the mixture until there is just enough free oxygen in the exhaust to burn of the extra fuel. In both cases, its trying to get back to 14.7, as read by the Nernst cell; just like a wideband.
I'm sleepy now....
Maybe it would be helpful to explain how these things ( Oxygen sensors )actually work from the insde ( for those who don't already know )? Maybe I'll just make things worse.
The narrowband Oxygen sensors work on the Nernst priciple. They are not AFR sensors, or even Lambda sensors. They just sense oxygen molecules at their sensor membrane (the "Nernst cell"). One side of the membrane is exposed to free air. the other side to the exhaust. If there is no free oxygen in the exhaust, some is pulled through the membrane. These molecules crossing the membrane cause a voltage ( or charge, more properly ) according to the Nernst equation. It's kinda like a battery. As long as there is no "load" on it, it doen't matter so much just how many molecules are moving.
The basic idea is, if there is any free oxygen in the the exhaust at all, you are running lean of stochiometric. These cells are, ideally, switches; either on or off. 0volts if there is no Oxygen present and a voltage, dependant on temperature and pressure, if it is. Generally this "full scale" voltage is somewhere in the neighborhood of 1.0v but it can be 700mv to 1300mv depending on temperature and pressure. Since we don't live in a laboratory, the switching is not quite a perfect on/off, with nothign in between. So, there is a region called an ambigious region that surrounds the switching point. This type of curve is called an instability about a point. So, at 14.7 AFR, the point where there is no more oxygen in the exhaust, the output of the Nernst cell wanders around the midpoint of it's output range. Usually about 450-500mv. As you move away from that point, the output quickly saturates twords either 0v or full scale output. In theory, if the Nernst cell in the sensor was only one molecule in size, the sensor would switch perfectly from 0v to 1v ( or whatever the max is based on temp and presure ) and there would be no ambigious range. The percieved "linear range" is just the effect of different molecules of the Nernst cell fighting with eachother. Some see an Oxygen molecule, some don't.
That's, partly, why the range is unstable. Changes in temperature change the "full scale" output. Changes in pressure affect how many molecules of the Nernst cell actually see Oxygen molecules. If you are a Chem major ( and are bored and out of beer ), you might want to look up the Nernst equations. or not.
http://www.science.uwaterloo.ca/~cch.../nernsteq.html
Temperature controlled narrowbands do a little better, as the full scale voltage is more predictable. But they still have the instability and pressure sensitivities. And that varies from part to part.
Widebands kinda cheat. A wideband sensor has a narrowband sensor inside of it. It also has what is called a "pump cell". ( that's the 5th wire on a 5-wire sensor) This is a ceramic membrane that will actually "pump" Oxygen ions across it based on electrical current. I works both ways. that is, put a positive voltage across it, and it pumps oxygen ions into the Nernst cell. Put a negative voltage and it pumps Oxygen ions out of the cell. The amount of oxygen it pumps is proportional ( more or less ) to the electrical current applied. It is acually the same type of material as the Nernst cell.
The Wideband system constantly tries to make the Nersnt cell read a perfect 14.7 AFR by using the pump cell to either add Oxygen to burn off unburnt fuel or pumping extra oxygen out of a lean exhaust. The amount of current it takes to get the Nernst cell back to 14.7 ( where it is predictable )determines the actual AFR. Thats why you can get a wide stable linear range from a wideband unit; even though the Nernst cell itself can not resolve values very far from 14.7AFR. The Nernst cell in a wideband sensor is ALWAYS at 14.7AFR. It's the pump cell that is linear.
You can kinda do the same thing by injecting air or fuel into the exhaust stream right before a narrowband sensor. If you add fresh air, the sensor will think its leaner than it is and the ECU will try to richen up the mixture to burn off the extra oxygen you are adding. If you add fuel, the opposite happens. The ECU leans out the mixture until there is just enough free oxygen in the exhaust to burn of the extra fuel. In both cases, its trying to get back to 14.7, as read by the Nernst cell; just like a wideband.
I'm sleepy now....
Thank you.
03-28-2005, 05:44 PM
#71
11 Second Club
iTrader: (16)
Join Date: May 2004
Location: MS
Posts: 2,223
Likes: 0
Received 0 Likes
on
0 Posts
Too much for me to read but...
fyi 800 on my car is like 14 sometimes.. and after getting the wideband i have seen how bad off o2's are. One day 900 can be 12.8 and other days it can be 13.8. I dont know about any other car but mine o2's are no where near accurate and I wouldnt trust them ever again..
fyi 800 on my car is like 14 sometimes.. and after getting the wideband i have seen how bad off o2's are. One day 900 can be 12.8 and other days it can be 13.8. I dont know about any other car but mine o2's are no where near accurate and I wouldnt trust them ever again..
03-28-2005, 06:06 PM
#72
FormerVendor
iTrader: (1)
Join Date: Feb 2002
Location: Waldorf, MD
Posts: 3,059
Likes: 0
Received 0 Likes
on
0 Posts
The million dollar question is how accurate is the WB sensor and can it be completely trusted for what it is doing. Some on this forum have argued the inaccuracies of the WB so maybe the professional can put the skepticism to rest
03-29-2005, 12:33 AM
#74
Remember, I work for Innovate ( LM-1 makers ) when reading this:
Widebands are generally very accurate when you take them out of the box. The sensors are characterized at the factory. When they are being tested at the factory, they are exposed to a calibrated gas and then a "calibration resistor" on the sensor is "laser trimmed" to a value tha standardizes the output of the sensor. When the cal resistor is taken into account in the controler circuit, all the sensors read about the same. So, a nice new sensor is probably spot-on.
The problem comes with sensor aging. As the sensors get older, their outputs start to drift. ( Actually its the efficiency of the pump cell that changes ). Eventually, they kinda stabilize. But, by then, they usually do not match the cal resistor anymore. The richer you run, the faster they drift out of calibration. On an OEM application that spends 95% of its time at stoich, this is not so much of an issue. On a tuner car running 12.0AFR, the sensor can drift out of calibration in less than 4 hours.
This is why the LM-1 has a "Free Air Calibration" mode. The LM-1 actually does not even look at the calibration resistor. This is because it is not an analog device. Analog controlers use the calibration resistor in their feedback loops; to control the pump cell current. The cal resistor never changes. But, the pump cell does...
The LM-1 uses a completely different method to control the sensor. it uses a form of PWM. It feeds a digital squarewave signal to the pumpcell ( 50mA one way or the other ). When its high it pumps one way, and when its low it pumps the other way.
The Nernst cell creates the squarewave (kinda). When the nernst cell switches on-off, the square wave switches the opposite way. So, it oscillates. Based on that, we know that at a perfect 50% duty cycle... the pump is effectivly nullified ( equal amounts of o2 being pumped in/out). So, we know we are right at 14.7AFR. And that never changes. 50% dutycyle = stoich ( 14.7AFR ) always.
When you do a free air calibrate, the LM-1 tests what duty cycle it needs to force the sensor back to stoich when the sensor is exposed to free air. ( remember the nernst cell always runs a stoich). The oxygen content of free air is a known value. 20.9%
Ok.... now we have 2 points. we know the duty cycles at 14.7 and 20.9%. For the sensor being calibrated. From there it is easy to calculate a function to convert duty cycle directly to AFR. And it matches the sensor's current state. This value is stored in the LM-1's flash memory.
Sensor calibrated.
So, even if the sensor has aged, drifted, etc. Once you do a "free air cal", its like having a new sensor again.
Some people find this to be a pain in the ***. The 1st time you use a new sensor on an LM-1 you have to do a free air cal. It won't give a reading until you do. For what should be obvious reasons now. It never trusts ( or even looks at ) the cal resistor. So, it has no idea how to read the sensor until it's calibrated.
I reccomend doing a free air calibration as often as you can. Especially if you are logging data. Just do a free air cal before you start your sessions. And your data will always be consistant.
This is the Innovate patent ( the PWM squarewave dutycycle air cal thing ). It is more accurate and MUCH MUCH FASTER than the analog method of running variable current through the cal resistor and pump cell.
It is so fast that it can see changes in AFR between individual cyninders; at idle anyway. If you look at the output of an LM-1 or (even faster ) LC-1 with an oscilloscope, you will see what looks like noise. But if you trigger the scope off of the spark wire for a cylinder, you will see the "noise" start to line up... on a per cylinder basis. At least at idle
Widebands are generally very accurate when you take them out of the box. The sensors are characterized at the factory. When they are being tested at the factory, they are exposed to a calibrated gas and then a "calibration resistor" on the sensor is "laser trimmed" to a value tha standardizes the output of the sensor. When the cal resistor is taken into account in the controler circuit, all the sensors read about the same. So, a nice new sensor is probably spot-on.
The problem comes with sensor aging. As the sensors get older, their outputs start to drift. ( Actually its the efficiency of the pump cell that changes ). Eventually, they kinda stabilize. But, by then, they usually do not match the cal resistor anymore. The richer you run, the faster they drift out of calibration. On an OEM application that spends 95% of its time at stoich, this is not so much of an issue. On a tuner car running 12.0AFR, the sensor can drift out of calibration in less than 4 hours.
This is why the LM-1 has a "Free Air Calibration" mode. The LM-1 actually does not even look at the calibration resistor. This is because it is not an analog device. Analog controlers use the calibration resistor in their feedback loops; to control the pump cell current. The cal resistor never changes. But, the pump cell does...
The LM-1 uses a completely different method to control the sensor. it uses a form of PWM. It feeds a digital squarewave signal to the pumpcell ( 50mA one way or the other ). When its high it pumps one way, and when its low it pumps the other way.
The Nernst cell creates the squarewave (kinda). When the nernst cell switches on-off, the square wave switches the opposite way. So, it oscillates. Based on that, we know that at a perfect 50% duty cycle... the pump is effectivly nullified ( equal amounts of o2 being pumped in/out). So, we know we are right at 14.7AFR. And that never changes. 50% dutycyle = stoich ( 14.7AFR ) always.
When you do a free air calibrate, the LM-1 tests what duty cycle it needs to force the sensor back to stoich when the sensor is exposed to free air. ( remember the nernst cell always runs a stoich). The oxygen content of free air is a known value. 20.9%
Ok.... now we have 2 points. we know the duty cycles at 14.7 and 20.9%. For the sensor being calibrated. From there it is easy to calculate a function to convert duty cycle directly to AFR. And it matches the sensor's current state. This value is stored in the LM-1's flash memory.
Sensor calibrated.
So, even if the sensor has aged, drifted, etc. Once you do a "free air cal", its like having a new sensor again.
Some people find this to be a pain in the ***. The 1st time you use a new sensor on an LM-1 you have to do a free air cal. It won't give a reading until you do. For what should be obvious reasons now. It never trusts ( or even looks at ) the cal resistor. So, it has no idea how to read the sensor until it's calibrated.
I reccomend doing a free air calibration as often as you can. Especially if you are logging data. Just do a free air cal before you start your sessions. And your data will always be consistant.
This is the Innovate patent ( the PWM squarewave dutycycle air cal thing ). It is more accurate and MUCH MUCH FASTER than the analog method of running variable current through the cal resistor and pump cell.
It is so fast that it can see changes in AFR between individual cyninders; at idle anyway. If you look at the output of an LM-1 or (even faster ) LC-1 with an oscilloscope, you will see what looks like noise. But if you trigger the scope off of the spark wire for a cylinder, you will see the "noise" start to line up... on a per cylinder basis. At least at idle
03-29-2005, 03:01 AM
#75
Some on this forum have argued the inaccuracies of the WB so maybe the professional can put the skepticism to rest
Long live The Amazing Randi!
10-10-2007, 05:43 PM
#77
TECH Regular
I know this is an old thread, but I did a few searches and this is all I could find so.....
I am looking for data that puts a narrowband and wideband log next to each other.
I have seen just 1 instance where anyone even posted data and it was at the beginning of this thread way back in 2004. I am surprised to see that EVERYONES observations are in the > .9 volt range. I ALWAYS get the same consistent reading of .88 and if I added say 5% fuel my next reading will be .89.
The OEM tune, which most agree is rich, only rarely goes to .90. of my NB.
Instead of making a point of trying to prove the narrow band is inaccurate would someone please post a narrowband and wideband log next to each other. I'd like to see the data myself. I will not be buying a WB, I just want the data. Thank you.
I am looking for data that puts a narrowband and wideband log next to each other.
I have seen just 1 instance where anyone even posted data and it was at the beginning of this thread way back in 2004. I am surprised to see that EVERYONES observations are in the > .9 volt range. I ALWAYS get the same consistent reading of .88 and if I added say 5% fuel my next reading will be .89.
The OEM tune, which most agree is rich, only rarely goes to .90. of my NB.
Instead of making a point of trying to prove the narrow band is inaccurate would someone please post a narrowband and wideband log next to each other. I'd like to see the data myself. I will not be buying a WB, I just want the data. Thank you.
10-11-2007, 07:10 AM
#79
TECH Regular
"I ALWAYS get the same consistent reading of .88 and if I added say 5% fuel my next reading will be .89."
I agree that the band is very small. I.E. a change of .01 is significant.
But it is there, it is consistant. But I want to know what AFR .87v is. and .88 and .89 and .90 on a specific car.
I agree that the band is very small. I.E. a change of .01 is significant.
But it is there, it is consistant. But I want to know what AFR .87v is. and .88 and .89 and .90 on a specific car.
10-13-2007, 11:37 AM
#80
Teching In
iTrader: (1)
Join Date: Aug 2005
Location: Memphis, Tn
Posts: 38
Likes: 0
Received 0 Likes
on
0 Posts
"I ALWAYS get the same consistent reading of .88 and if I added say 5% fuel my next reading will be .89."
I agree that the band is very small. I.E. a change of .01 is significant.
But it is there, it is consistant. But I want to know what AFR .87v is. and .88 and .89 and .90 on a specific car.
I agree that the band is very small. I.E. a change of .01 is significant.
But it is there, it is consistant. But I want to know what AFR .87v is. and .88 and .89 and .90 on a specific car.