No Device: Wide Band or Narrow Band ?
#42
I'm sure it is. I think the point that Steve was trying to make is in regards to a slower data rate of a particular 02 sensor (Bosch WB) compared to NB.
Also another thing I noticed is that folks using WB TT maybe under the impression that their bike is constantly being tuned.
I don't think that is the case.
Once you get an engine dialed in and a completed tune flashed that's where it should stay. The only thing the 02s should be doing after this point is verifying and helping to correct for minor changes. This is particularly helpful on engines as they get older and clock a lot of miles.
At least that's how it works on your average car.
Also another thing I noticed is that folks using WB TT maybe under the impression that their bike is constantly being tuned.
I don't think that is the case.
Once you get an engine dialed in and a completed tune flashed that's where it should stay. The only thing the 02s should be doing after this point is verifying and helping to correct for minor changes. This is particularly helpful on engines as they get older and clock a lot of miles.
At least that's how it works on your average car.
I'm sure some shops think it's great. I can sell you a tuner, and a tune. Just like before. Now with this new TT. Before the bike goes out the door. I now can upsell you something else that I will get labor for install and money from the mark up of the product. Sounds like they manufactured a false need for the product they are now going to sell you. ........and they will call it Target Tune. Even though it is not meant to tune your bike. The only thing it address's is the fuel side during steady state cruise during closed loop operation. It does nothing in the open loop operations. Your bike is not in closed loop 100% of the time with TT. The same rules apply as when it was handled with NB's. Acceleration table is still going to kick it out of closed loop. Deceleration table will do the same. Same with PE mode. How does TT feed back for the WOT/ high rpm area with PE mode on? All goes back to the VE tables that were tuned correctly in the beginning.
#43
If we take the FACTS and look them over people may be, more able to understand the truth about things when it comes to the WB debate. All the "I like this one better" cannot get around the simple facts.
Let start with an engine at idle to make this simple.
1000 RPM = 8.33 exhaust events per cylinder FACT
1000 RPM = a total 240 mS for one cylinder event FACT
A Bosch WB (LSU4.x) sensor is a NB sensor with an additional circuit added to it FACT
Since it is a NB+, all the debate of it being better quality or whatever goes right out the window FACT
Today's factory NB sensor reads ~ 20 times faster than a Bosch WB sensor FACT
Working with a 4 cycle engine the exhaust cycle is about 25% of the total time. Depending on the camshaft slightly less, to slightly more, but 25% of the time is a good average of them all.
240mS x .25 = 60mS total exhaust flow
So you have 60ms of time to get a reading of the exhaust from one exhaust event @ 1000RPM or the accuracy goes out the widow FACT
If you try and read the exhaust before or after the valve is closed the accuracy goes out the window FACT
Response time of the Bosch LSU4.x sensor is ~250mS FACT
So you have a exhaust flow from the cylinder of 60 mS and a sensor that needs ~250 to get an accurate reading. It is therefor impossible for the sensor itself to accurately read any single cylinder event @ 1000RPM FACT
These are the facts, and it's not hard to see from them that the WB cannot handle a single cylinder running at 1000 RPM, let alone a 4, 6 or 8 cylinder engine that is running 5000 + RPM and provide accurate information about that cylinder event . This is just why any good tuner knows better than to rely on them, UNLESS they can control the engine RPM and load. This is just what a dyno tuner does, he controls the engine so that it stays at the same RPM and Load for a long enough period of time, to allow the sensor reading to get stable(averaged). They then know that the reading they are getting, is only one part of what they need, to make adjustments.
Let start with an engine at idle to make this simple.
1000 RPM = 8.33 exhaust events per cylinder FACT
1000 RPM = a total 240 mS for one cylinder event FACT
A Bosch WB (LSU4.x) sensor is a NB sensor with an additional circuit added to it FACT
Since it is a NB+, all the debate of it being better quality or whatever goes right out the window FACT
Today's factory NB sensor reads ~ 20 times faster than a Bosch WB sensor FACT
Working with a 4 cycle engine the exhaust cycle is about 25% of the total time. Depending on the camshaft slightly less, to slightly more, but 25% of the time is a good average of them all.
240mS x .25 = 60mS total exhaust flow
So you have 60ms of time to get a reading of the exhaust from one exhaust event @ 1000RPM or the accuracy goes out the widow FACT
If you try and read the exhaust before or after the valve is closed the accuracy goes out the window FACT
Response time of the Bosch LSU4.x sensor is ~250mS FACT
So you have a exhaust flow from the cylinder of 60 mS and a sensor that needs ~250 to get an accurate reading. It is therefor impossible for the sensor itself to accurately read any single cylinder event @ 1000RPM FACT
These are the facts, and it's not hard to see from them that the WB cannot handle a single cylinder running at 1000 RPM, let alone a 4, 6 or 8 cylinder engine that is running 5000 + RPM and provide accurate information about that cylinder event . This is just why any good tuner knows better than to rely on them, UNLESS they can control the engine RPM and load. This is just what a dyno tuner does, he controls the engine so that it stays at the same RPM and Load for a long enough period of time, to allow the sensor reading to get stable(averaged). They then know that the reading they are getting, is only one part of what they need, to make adjustments.
#44
Here is where it works. This happens to be about 4 min of operation.
http://www.nbs-stl.com/CRX/Autocross...and%20Time.png
What's your point?
http://www.nbs-stl.com/CRX/Autocross...and%20Time.png
What's your point?
Maybe offer the same graph for a typical Harley ridden on the street?
I believe the picture would be a lot different than a CRX racing around cones for best time.
Bob
#45
I'm sure it is. I think the point that Steve was trying to make is in regards to a slower data rate of a particular 02 sensor (Bosch WB) compared to NB.
Also another thing I noticed is that folks using WB TT maybe under the impression that their bike is constantly being tuned.
I don't think that is the case.
Once you get an engine dialed in and a completed tune flashed that's where it should stay. The only thing the 02s should be doing after this point is verifying and helping to correct for minor changes. This is particularly helpful on engines as they get older and clock a lot of miles.
At least that's how it works on your average car.
Also another thing I noticed is that folks using WB TT maybe under the impression that their bike is constantly being tuned.
I don't think that is the case.
Once you get an engine dialed in and a completed tune flashed that's where it should stay. The only thing the 02s should be doing after this point is verifying and helping to correct for minor changes. This is particularly helpful on engines as they get older and clock a lot of miles.
At least that's how it works on your average car.
My point is to stop the mis-information that is being passed out. I think Bosch Broad Band sensors are fine when used correctly and by someone who understands the limits. In the case of a HD Twin Cam you have to look at it like two one cylinder engines with a common crankshaft and common intake manifold. With the stock NB sensor it can respond and see real changes in the entire closed loop range cycle by cycle, a WB cannot. No getting around the simple facts. Now if your bike uses a 4 pin DLC connector you have a J1850 system and the stock HD code only allows data to be put out ~ 2 times per second, so you miss seeing everything else that is really going on in the data. Remember the DLC communications was NOT designed to provide real time tuning information. While it does give helpful information, it runs way to slow to show the real picture even at idle, let alone at higher RPM's. If your bike is the later 6 pin DLC connector then your data rate goes up to ~ 12 times per second with stock HD code. Much better but still not near enough for higher RPM's. So there is an art, for lack of better words, to getting the job done. One of the tricks the dyno operators use is to slow the engine down by holding it in a steady state condition. This allows for the data coming out of the ECM and sensors to be a truer average of what is going on. Mind you its not perfect but much better.
As far as using a stock HD delphi ECM and using WB sensors the ECM is slowly making adjustments but it does NOT remember any of them! That is a key point, it makes an adjustment based on the current condition the engine is operating in. So let's take the typical bike rider who starts out from a stop sign. You let the clutch out and give it throttle to accelerate the engine up in RPM then let off throttle and shift. You repeat the process in each gear until you reach whatever speed you are looking for. So the ECM is adjusting in real time but the data coming out will not show it and you have a WB sensor not capable of giving the ECM what it needs cycle by cycle. So the engine is at 1000 RPM when you let the clutch out and the ECM is making an adjustment based on the information coming from a WB in this case. Let say the setting in the ECM is off 10%, so it issues a 10% correction but you now start to accelerate the engine and the WB tells the ECM its now 20% off, so the ECM makes a 20% correction and the previous 10% correction is gone forever. So now the engine continues up in RPM and this process happens over and over again until you reach the engine RPM you want and you pull the clutch in and shift. Engine RPM drops and the process starts all over again. Nothing is ever learned by the ECM so it starts over and over again the entire time the engine is running!
The reason for even having this development code in the ECM in the first place was to help create a new tune on a dyno where the engine would be held in a steady state condition and the VE values could be properly adjusted by the calibrator. Test, adjust and move on. Once the VE tables are properly calibrated the ECM is now running the show and the closed loop feed back system is turned on to allow feed back for trimming over the life of the engine. The NB system operates differently than the development code does in the ECM. It adjust real time, cycle by cycle and then stores the adjustments for anytime the engine returns to that operating condition. The memorized values are allowed to be slowly adjusted over time.
This is the heart of a Closed Loop system and they have worked this way for years in both the car and motorcycle world so it's nothing new or trick that HD has done. The fact that some are trying to misuse the development code and telling people it does things it clearly is not capable of is NOT coming from HD/Delphi.
The following 6 users liked this post by Steve Cole:
breakman (05-01-2016),
Gordon61 (05-02-2016),
IdahoHacker (05-02-2016),
JustDave13 (05-01-2016),
Stud Duck (10-03-2016),
and 1 others liked this post.
#46
There is some really good info in this thread, thanks to all. For arguments sake we'll say the nb's are more accurate in their range. With that being said, has anyone taken one bike, Harley only, tuned it with nb to the best of their ability on a dyno, gone out ridden at least 100 miles, hopefully data logging, then switch the same bike over to wb' retune to the best of their ability, gone out for the 100 mile data run and compare it? Is there a big enough difference for concern? And, is there a definite advantage one over the other. Of course the last question, will the riders butt dyno feel the difference, if both tunes are using the same afr/lambda range.
The results I have seen with this isn't real exciting. Comparing a CL in the cruise area calibration to a OL tune. I saw no difference in engine or oil temp. Fuel mileage dropped about 10 from 43. Plugs were not as clean.
Did it ride and perform better in the cruise area? Was it more responsive? Could I even tell a difference? No. How could I quantify any felt change at steady state? Is the HD bathtub style head sophisticated enough to know the difference between 14.3 and 14 or 13.8? How much more fuel is that actually equate to. Is the math there to support if that amount is going to have an affect on power output or heat absorption? How much of an affect?
Lets look at what happens when you roll on pass on the highway. As soon as you open throttle. You loose vacuum and MAP goes up. We are now operating in a different area of the AFR table, VE table, spark table....... and that area of the table can be set up the same in both of the test calibration. Regardless of how the steady state cruise area is.
If 13.8 is a better starting point to transition to 13.2 with throttle opening. Your acceleration tables needs some work.
The following users liked this post:
Wmitz (05-01-2016)
#47
My point is to stop the mis-information that is being passed out. I think Bosch Broad Band sensors are fine when used correctly and by someone who understands the limits. In the case of a HD Twin Cam you have to look at it like two one cylinder engines with a common crankshaft and common intake manifold. With the stock NB sensor it can respond and see real changes in the entire closed loop range cycle by cycle, a WB cannot. No getting around the simple facts. Now if your bike uses a 4 pin DLC connector you have a J1850 system and the stock HD code only allows data to be put out ~ 2 times per second, so you miss seeing everything else that is really going on in the data. Remember the DLC communications was NOT designed to provide real time tuning information. While it does give helpful information, it runs way to slow to show the real picture even at idle, let alone at higher RPM's. If your bike is the later 6 pin DLC connector then your data rate goes up to ~ 12 times per second with stock HD code. Much better but still not near enough for higher RPM's. So there is an art, for lack of better words, to getting the job done. One of the tricks the dyno operators use is to slow the engine down by holding it in a steady state condition. This allows for the data coming out of the ECM and sensors to be a truer average of what is going on. Mind you its not perfect but much better.
As far as using a stock HD delphi ECM and using WB sensors the ECM is slowly making adjustments but it does NOT remember any of them! That is a key point, it makes an adjustment based on the current condition the engine is operating in. So let's take the typical bike rider who starts out from a stop sign. You let the clutch out and give it throttle to accelerate the engine up in RPM then let off throttle and shift. You repeat the process in each gear until you reach whatever speed you are looking for. So the ECM is adjusting in real time but the data coming out will not show it and you have a WB sensor not capable of giving the ECM what it needs cycle by cycle. So the engine is at 1000 RPM when you let the clutch out and the ECM is making an adjustment based on the information coming from a WB in this case. Let say the setting in the ECM is off 10%, so it issues a 10% correction but you now start to accelerate the engine and the WB tells the ECM its now 20% off, so the ECM makes a 20% correction and the previous 10% correction is gone forever. So now the engine continues up in RPM and this process happens over and over again until you reach the engine RPM you want and you pull the clutch in and shift. Engine RPM drops and the process starts all over again. Nothing is ever learned by the ECM so it starts over and over again the entire time the engine is running!
That's not good, without learned adaptives the set up is useless.
The reason for even having this development code in the ECM in the first place was to help create a new tune on a dyno where the engine would be held in a steady state condition and the VE values could be properly adjusted by the calibrator. Test, adjust and move on. Once the VE tables are properly calibrated the ECM is now running the show and the closed loop feed back system is turned on to allow feed back for trimming over the life of the engine. The NB system operates differently than the development code does in the ECM. It adjust real time, cycle by cycle and then stores the adjustments for anytime the engine returns to that operating condition. The memorized values are allowed to be slowly adjusted over time.
This is the heart of a Closed Loop system and they have worked this way for years in both the car and motorcycle world so it's nothing new or trick that HD has done. The fact that some are trying to misuse the development code and telling people it does things it clearly is not capable of is NOT coming from HD/Delphi.
As far as using a stock HD delphi ECM and using WB sensors the ECM is slowly making adjustments but it does NOT remember any of them! That is a key point, it makes an adjustment based on the current condition the engine is operating in. So let's take the typical bike rider who starts out from a stop sign. You let the clutch out and give it throttle to accelerate the engine up in RPM then let off throttle and shift. You repeat the process in each gear until you reach whatever speed you are looking for. So the ECM is adjusting in real time but the data coming out will not show it and you have a WB sensor not capable of giving the ECM what it needs cycle by cycle. So the engine is at 1000 RPM when you let the clutch out and the ECM is making an adjustment based on the information coming from a WB in this case. Let say the setting in the ECM is off 10%, so it issues a 10% correction but you now start to accelerate the engine and the WB tells the ECM its now 20% off, so the ECM makes a 20% correction and the previous 10% correction is gone forever. So now the engine continues up in RPM and this process happens over and over again until you reach the engine RPM you want and you pull the clutch in and shift. Engine RPM drops and the process starts all over again. Nothing is ever learned by the ECM so it starts over and over again the entire time the engine is running!
That's not good, without learned adaptives the set up is useless.
The reason for even having this development code in the ECM in the first place was to help create a new tune on a dyno where the engine would be held in a steady state condition and the VE values could be properly adjusted by the calibrator. Test, adjust and move on. Once the VE tables are properly calibrated the ECM is now running the show and the closed loop feed back system is turned on to allow feed back for trimming over the life of the engine. The NB system operates differently than the development code does in the ECM. It adjust real time, cycle by cycle and then stores the adjustments for anytime the engine returns to that operating condition. The memorized values are allowed to be slowly adjusted over time.
This is the heart of a Closed Loop system and they have worked this way for years in both the car and motorcycle world so it's nothing new or trick that HD has done. The fact that some are trying to misuse the development code and telling people it does things it clearly is not capable of is NOT coming from HD/Delphi.
It seems to me that a lot of folks don't understand or realize that the NB sensors are perfectly fine for determining accurate VEs throughout the whole map while running a tuning session (I use basic auto tune with my Power Vision).
Once the VEs are memorized and written that part of the tune is set, you could throw away the 02 sensors and your bike will run just as good off of the VEs
Last edited by JustDave13; 05-01-2016 at 04:39 PM.
#49
Like I mentioned in bold if the TT using WB sensors can't be used to incorporate learned values then it does nothing or even worse it just messes things up by confusing the ECM.
It seems to me that a lot of folks don't understand or realize that the NB sensors are perfectly fine for determining accurate VEs throughout the whole map while running a tuning session (I use basic auto tune with my Power Vision).
Once the VEs are memorized and written that part of the tune is set, you could throw away the 02 sensors and your bike will run just as good off of the VEs
It seems to me that a lot of folks don't understand or realize that the NB sensors are perfectly fine for determining accurate VEs throughout the whole map while running a tuning session (I use basic auto tune with my Power Vision).
Once the VEs are memorized and written that part of the tune is set, you could throw away the 02 sensors and your bike will run just as good off of the VEs
Now, someone is starting to understand that you are only using the NB sensor to arrive at the proper VE value. Place the ECM in a range the NB is deadly accurate and adjust the VE values to get the target mixture, now the system is calibrated. Notice I did NOT say tuned, but calibrated. Once you have that, you can set the ECM to run back in Open loop in the high load areas and you will have a good tune done. Then, the weather, and fuel must remain exactly the same forever and the combustion process must be exactly as it was the day you tested. Could it be better, sure but you would need a dyno and several other pieces of equipment to be sure that the accuracy is any better.
#50
The results I have seen with this isn't real exciting. Comparing a CL in the cruise area calibration to a OL tune. I saw no difference in engine or oil temp. Fuel mileage dropped about 10 from 43. Plugs were not as clean.
Did it ride and perform better in the cruise area? Was it more responsive? Could I even tell a difference? No. How could I quantify any felt change at steady state? Is the HD bathtub style head sophisticated enough to know the difference between 14.3 and 14 or 13.8? How much more fuel is that actually equate to. Is the math there to support if that amount is going to have an affect on power output or heat absorption? How much of an affect?
Lets look at what happens when you roll on pass on the highway. As soon as you open throttle. You loose vacuum and MAP goes up. We are now operating in a different area of the AFR table, VE table, spark table....... and that area of the table can be set up the same in both of the test calibration. Regardless of how the steady state cruise area is.
If 13.8 is a better starting point to transition to 13.2 with throttle opening. Your acceleration tables needs some work.
Did it ride and perform better in the cruise area? Was it more responsive? Could I even tell a difference? No. How could I quantify any felt change at steady state? Is the HD bathtub style head sophisticated enough to know the difference between 14.3 and 14 or 13.8? How much more fuel is that actually equate to. Is the math there to support if that amount is going to have an affect on power output or heat absorption? How much of an affect?
Lets look at what happens when you roll on pass on the highway. As soon as you open throttle. You loose vacuum and MAP goes up. We are now operating in a different area of the AFR table, VE table, spark table....... and that area of the table can be set up the same in both of the test calibration. Regardless of how the steady state cruise area is.
If 13.8 is a better starting point to transition to 13.2 with throttle opening. Your acceleration tables needs some work.