ecoboostsho
New member
Learned Octane Ratio (LOR)
Not being able to understand what the LOR (A gauge we have in the Torque App) was doing or even if the formula was right (I was pretty confident it wasn’t) was driving me a bit crazy so I decided to do some experimenting. This is a long post so I apologize in advance but stick with it – I think this is a good gauge to have working when all is said and done.
First a good refresher on what the LOR gauge is "supposed" to do would be in order. I believe it is approximating the octane level in your gas - not through a sensor - but by monitoring knock (and maybe other?) parameters in the ECU. If it sees it knocking away then it assumes you've got some bad gas or you put in 87 Octane and it adjusts the timing to compensate by multiplying the LOR number by the appropriate timing table and thus reducing your overall timing. This should result in less knock and a happy motor. A great analogy (Thanks Larry!) is that the LOR is sort of like the Long Term Fuel Trim in that it learns over time from the Short Term Fuel Trims and comes up with an adjustment value. In this case however the Knock gauge is the “real time” feedback loop and LOR is the learned value over time that will try and prevent knock in the first place.
The Experiment:
I put the car back to stock as I was planning on running different Octane fuels to try and understand the results. Needless to say I didn’t think throwing 87 octane in with my 93 tune was a great idea. I didn’t really like doing it stock but I did this to try and figure out how it worked.
Step one after going back to stock was to watch the values change as it learned and I did this with 93 octane and a mix of E-85. I calculated my mix at about E16.
Timeout: The other piece of background information that I want to insert here was the Cobb tuning article posted by another member that indicated how the numbers should work in theory…a value of -1 was said to indicate the “best” possible ratio you could achieve. If the value moved back toward 0 or to +1 then the car was ultimately removing timing to avoid Knock. It used values from part throttle and full throttle. They call it “Octane Adjust Ratio” but I believe it is the same thing…I may even rename the PID in Torque. http://cobbtuning.zendesk.com/hc/en-us/articles/204049644-Octane-Adjust-Ratio-OAR-
Back to the experiment: So as I stated above I started with E16 so I could get the car to a theoretical -1 and let the car go from there. After tracking the points and doing some math it was clear that the numbers were signed binary numbers but thankfully Torque handles this. The formula I came up with was:
(Signed(A)*256+B)/16384
Custom PID is 2203E8
(If anyone needs assistance in where to type this in just PM me or refer to the Torque PIDs thread in this same forum.) You will need to DELETE the existing gauge off of the display and THEN edit the PID or you will get weird results.
I used the Digital display gauge and one important point (if using that gauge) is to override the default # of decimal places and use a value = 2. Just press on the gauge on the screen and go to “display configuration” and select “decimal places”. If you don’t do this it will start rounding on you and you may not realize it has changed at all.
My car started “learning” and moved toward -1 and eventually achieved that value after a few hours of driving the car. I did get some part throttle knock and at some point the value actually “settled” at -.98 for about a week and then finally back to -1 after about a full tank of gas.
Next up: 87 Octane
I put in 10 gallons of 87 Octane with about a ¼ tank (roughly 5 gallons or so) while not exactly 87 Octane combined I’m guessing it was probably in the 89-90 total range. Well my results were immediate and dramatic. Once I used up whatever 93+ was in the lines the car immediately started knocking during part throttle acceleration on a freeway on ramp. I saw a constant 4-5 degrees of knock at part throttle and then just couldn’t take it anymore (i.e. I chickened out) so I backed off the throttle. The LOR instantly went from -1 to -.75 and the car immediately felt slower (I’m not kidding I could tell…) but it was no longer knocking nearly as much when attempting the same type of acceleration. I only drove it for a few miles like this and expect it to continue to move towards 0 but time will tell what happens and I will update the results as I get them. I can’t imagine what kind of stupid gas it would take to make it positive number but I don’t know that I want to find out. The good news is that it appears very sensitive and reacts right away to save the engine in conjunction with the Knock gauge.
After driving some more on the blended crappy gas I can report back that the car briefly tried to go from -.75 to -.83 and that lasted until the next highway on ramp run under part throttle and it shot back to -0.67 when I hit 4 degrees of knock. I can definitiely say the gauge is working well. So far any knock that hits 4 degrees or higher seems to cause a fairly immediate adjustment to the LOR gauge. The car seems to tolerate up to 3 degrees of knock as “normal” and doesn’t make any immediate adjustments to the LOR value.
Drove the car for a full week with the 87 Octane. I just put 93 octane back in it and after a few trips the LOR shot to -0.99 which is almost perfect. It did have a little of the 87 left in it so I’m running something a little less than pure 93. Car is still stock but will be retuning shortly.
I will probably try 91 on my 93 tune to see if the LOR value is indeed utilized by the “tuned” ECU. I can’t imagine it isn’t but you never know…
So in closing I highly recommend you remove the previous Learned Octane Ratio gauge (Also accept my apologies since I figured out the first one based on another Scan tool that apparently had it wrong as well).
Also these are all based on my research – I welcome any constructive feedback or data points you might experience so we can improve this and/or verify that this is currently “correct”. I am 99% more confident in this formula than the last one however, that doesn’t make it perfect. Just understand that there ARE situations where you could get readings that indicate “bad gas” or “Low Octane” that might actually be other problems. What I mean by that is if a knock sensor is going bad it may show excessive knock even on 93 octane fuel. In this case the problem isn’t the fuel but something with your car – but it will show up as a LOR that is moving toward 0 or +1 instead of -1.
Happy motoring!
PS. The car sucks on a stock tune and 87 octane…
Not being able to understand what the LOR (A gauge we have in the Torque App) was doing or even if the formula was right (I was pretty confident it wasn’t) was driving me a bit crazy so I decided to do some experimenting. This is a long post so I apologize in advance but stick with it – I think this is a good gauge to have working when all is said and done.
First a good refresher on what the LOR gauge is "supposed" to do would be in order. I believe it is approximating the octane level in your gas - not through a sensor - but by monitoring knock (and maybe other?) parameters in the ECU. If it sees it knocking away then it assumes you've got some bad gas or you put in 87 Octane and it adjusts the timing to compensate by multiplying the LOR number by the appropriate timing table and thus reducing your overall timing. This should result in less knock and a happy motor. A great analogy (Thanks Larry!) is that the LOR is sort of like the Long Term Fuel Trim in that it learns over time from the Short Term Fuel Trims and comes up with an adjustment value. In this case however the Knock gauge is the “real time” feedback loop and LOR is the learned value over time that will try and prevent knock in the first place.
The Experiment:
I put the car back to stock as I was planning on running different Octane fuels to try and understand the results. Needless to say I didn’t think throwing 87 octane in with my 93 tune was a great idea. I didn’t really like doing it stock but I did this to try and figure out how it worked.
Step one after going back to stock was to watch the values change as it learned and I did this with 93 octane and a mix of E-85. I calculated my mix at about E16.
Timeout: The other piece of background information that I want to insert here was the Cobb tuning article posted by another member that indicated how the numbers should work in theory…a value of -1 was said to indicate the “best” possible ratio you could achieve. If the value moved back toward 0 or to +1 then the car was ultimately removing timing to avoid Knock. It used values from part throttle and full throttle. They call it “Octane Adjust Ratio” but I believe it is the same thing…I may even rename the PID in Torque. http://cobbtuning.zendesk.com/hc/en-us/articles/204049644-Octane-Adjust-Ratio-OAR-
Back to the experiment: So as I stated above I started with E16 so I could get the car to a theoretical -1 and let the car go from there. After tracking the points and doing some math it was clear that the numbers were signed binary numbers but thankfully Torque handles this. The formula I came up with was:
(Signed(A)*256+B)/16384
Custom PID is 2203E8
(If anyone needs assistance in where to type this in just PM me or refer to the Torque PIDs thread in this same forum.) You will need to DELETE the existing gauge off of the display and THEN edit the PID or you will get weird results.
I used the Digital display gauge and one important point (if using that gauge) is to override the default # of decimal places and use a value = 2. Just press on the gauge on the screen and go to “display configuration” and select “decimal places”. If you don’t do this it will start rounding on you and you may not realize it has changed at all.
My car started “learning” and moved toward -1 and eventually achieved that value after a few hours of driving the car. I did get some part throttle knock and at some point the value actually “settled” at -.98 for about a week and then finally back to -1 after about a full tank of gas.
Next up: 87 Octane
I put in 10 gallons of 87 Octane with about a ¼ tank (roughly 5 gallons or so) while not exactly 87 Octane combined I’m guessing it was probably in the 89-90 total range. Well my results were immediate and dramatic. Once I used up whatever 93+ was in the lines the car immediately started knocking during part throttle acceleration on a freeway on ramp. I saw a constant 4-5 degrees of knock at part throttle and then just couldn’t take it anymore (i.e. I chickened out) so I backed off the throttle. The LOR instantly went from -1 to -.75 and the car immediately felt slower (I’m not kidding I could tell…) but it was no longer knocking nearly as much when attempting the same type of acceleration. I only drove it for a few miles like this and expect it to continue to move towards 0 but time will tell what happens and I will update the results as I get them. I can’t imagine what kind of stupid gas it would take to make it positive number but I don’t know that I want to find out. The good news is that it appears very sensitive and reacts right away to save the engine in conjunction with the Knock gauge.
After driving some more on the blended crappy gas I can report back that the car briefly tried to go from -.75 to -.83 and that lasted until the next highway on ramp run under part throttle and it shot back to -0.67 when I hit 4 degrees of knock. I can definitiely say the gauge is working well. So far any knock that hits 4 degrees or higher seems to cause a fairly immediate adjustment to the LOR gauge. The car seems to tolerate up to 3 degrees of knock as “normal” and doesn’t make any immediate adjustments to the LOR value.
Drove the car for a full week with the 87 Octane. I just put 93 octane back in it and after a few trips the LOR shot to -0.99 which is almost perfect. It did have a little of the 87 left in it so I’m running something a little less than pure 93. Car is still stock but will be retuning shortly.
I will probably try 91 on my 93 tune to see if the LOR value is indeed utilized by the “tuned” ECU. I can’t imagine it isn’t but you never know…
So in closing I highly recommend you remove the previous Learned Octane Ratio gauge (Also accept my apologies since I figured out the first one based on another Scan tool that apparently had it wrong as well).
Also these are all based on my research – I welcome any constructive feedback or data points you might experience so we can improve this and/or verify that this is currently “correct”. I am 99% more confident in this formula than the last one however, that doesn’t make it perfect. Just understand that there ARE situations where you could get readings that indicate “bad gas” or “Low Octane” that might actually be other problems. What I mean by that is if a knock sensor is going bad it may show excessive knock even on 93 octane fuel. In this case the problem isn’t the fuel but something with your car – but it will show up as a LOR that is moving toward 0 or +1 instead of -1.
Happy motoring!
PS. The car sucks on a stock tune and 87 octane…
