i hope to have a wealth of info over the next 2 weeks. I do not know who else has high boost BBSC , low comp obd II with SSBB.I have fjo wideband onboard with data logging and real time hookup to my ssbb via my laptop. I also have boost gauge/fuel pressure with peak hold. should be able to provide answers to alot of questions that have been presented here over the past months.
BBSC New SS Box Thread
- Thread starter kpond
- Start date
Yellow Rose
Suspended
- Joined
- 22 November 2001
- Messages
- 2,256
.....I do not know who else has high boost BBSC, low comp obd II with SSBB. I have fjo wideband onboard with data logging and real time hookup to my ssbb via my laptop. I also have boost gauge / fuel pressure.....
I do - yes to all of the above.
I do - yes to all of the above.
Yellow Rose
Suspended
- Joined
- 22 November 2001
- Messages
- 2,256
The 47° number you saw under light cruising conditions corresponds exactly with factory Honda low cam ignition map numbers in the 17 to 20 inches of vacuum range all the way to 5000 RPM.
Devin - I'm a bit confused.....the graph you posted shows timing to be between 23° and 29°.....how does that relate to 47°?
Devin - I'm a bit confused.....the graph you posted shows timing to be between 23° and 29°.....how does that relate to 47°?
AndyVecsey said:
The OEM graph I posted applies to full throttle (1 atmosphere) conditions from 3000 to redline. The Honda map under cruise between 17 and 20 inches in the RPM range you specified uses values around 47°.
More graphs Andy 
.
Here is a plot that explains the range we are talking about. Approximately 18" vacuum, light throttle cruise, 1250 RPM - 6250 RPM
It would take me forever to post every vacuum vs. RPM map (100's), but I hope you understand a little better what I meant.
. Here is a plot that explains the range we are talking about. Approximately 18" vacuum, light throttle cruise, 1250 RPM - 6250 RPM
It would take me forever to post every vacuum vs. RPM map (100's), but I hope you understand a little better what I meant.
Yellow Rose
Suspended
- Joined
- 22 November 2001
- Messages
- 2,256
Pssst.....Take A Look
http://baschboost.org/forums/viewtopic.php?p=252#252
First off, I need to make something very clear. What you are about to read was not derived all by me. The theory basis was provided by someone else that is widely recognized have having tremendous talent in the area of electronics, with him being an electrical engineer. I have added my experiences, opinions and observations.
Within the Split Second software are two maps. Map A is the fuel map and Map B is the timing map. Let’s begin with the latter, because it is easier to understand. Within a particular cell, that value represents the amount of timing that is being retarded (taken away) from the factory timing at that “snapshot” of load vs RPM. The amount of timing retard increases with RPM and the rate at which the timing retard increases, increases with rising load / boost. You can see this trend very clearly in Map B. I’ve created three timing maps - conservative, moderate and aggressive...currently running the conservative timing map.
As I understand timing, the correct math for a forced induction NSX could be to follow the rule of thumb of 1° retard per PSI of boost. So with my old 4.2" pulley that made 8 PSI at redline, the total timing at full throttle would be 28° advance minus 8° retard equals 20° total timing. This compares well the range of numbers that I've read for a boosted NSX of 18° to 22°. There is an NSX making huge HP at 27° of total timing; however, he has a huge aftercooler so he can get away with more aggressive timing.
Unlike the AEM where you tell it what you want the total timing to be, with the SS you just tell it how much timing to retard. The confirmation would be with a timing light on the dyno and / or a Honda scanner. I've tried one Honda scanner (that is outdated and repaired often) but I did not believe the result. If I can't get a believable reading from another scanner, I'll do the timing light trick. Somebody once told me that MB verified timing in this manner.
Now for the fuel map discussion. The OEM’s MAP sensor calibration is different from the SS’s MAP sensor calibration. The whole idea behind the fuel map is to modify the SS’s MAP sensor voltage so that the ECU thinks it is still being fed a signal from the factory MAP sensor. Sounds easy in concept, and to a point it is. A value in a particular boost column is how much the SS’s MAP sensor voltage is modified before the signal is conveyed to the ECU. A value of 10 means there is no signal modification. A value of 11 means the SS’s signal is increased 0.25 V. A value of 9 means the SS’s signal is decreased by 0.25 V. For the most part, increasing the value in the fuel map increases the richness of the air-fuel ratio.....to a point. I had experienced a situation where I was approaching a lean condition towards redline. Increasing the fuel map value in the corresponding boost columns had no affect on the air-fuel ratio. I had numbers in the 10’s and 11’s, but the air-fuel ratio was no responding. That’s when I wiped the slate clean and looked into just how the SS fuel map works.
You know how there is a value of 13.8 in the 500 RPM in 0 PSI column? Wonder how it is arrived at? At standard atmospheric pressure, the OEM’s MAP sensor to the ECU is 2.825 V. This is the signal voltage the ECU expects to see as the ignition key is turned, just before the engine fires. At standard atmospheric pressure, the SS’s MAP sensor is 1.864 V. The difference is 0.961 V. If you divide 0.961 by 0.25 (the same 0.25 mentioned in the paragraph before this one) you get 3.844. Round off to one decimal point to be consistent with the SS’s values and add to 10 (the same 10 mentioned in the paragraph before this one) you get 13.8. I measured the input voltage to the ECU from the modified SS output voltage is 2.65 V - a difference of 6%. The voltage measurement was made with a twenty year old analog voltmeter. When the key is turned to the ignition position, it is very likely that the stock ECU measures the MAP voltage to ensure that the MAP sensor itself is not broken. It may also perform a calibration of the MAP sensor against the barometric pressure sensor inside the ECU. This type of "sensor checking" is very common with ECUs and is used to identify bad sensors.
The ECU’s algorithm knows that if the incoming voltage is 2.825 V, this is at 0” Hg which means no intake manifold vacuum which means WOT (if the RPM is any number greater than zero). Since the ECU is built around a NA engine, when it gets out of closed loop (vacuum) and goes into open loop (boost) it defaults to an injector pulse width look-up table that Japan wrote to correspond with WOT. With the BBSC / SS, not only is the ECU pulsing the injector longer under WOT vs cruising conditions, the larger injectors are delivering the required amount of fuel to ensure the target air-fuel ratio.
If you know the injector size and what pressure the fuel pump delivers, you can calculate what the theoretical voltage modification should be for each ½ PSI boost column in Map A to yield a certain air-fuel ratio. I emphasize the words “theoretical” and “should” because the real world is different.
For my car (500 injectors, 54 PSI fuel pressure, target 11.8:1 AFR) the calculated modification values are 8.8 and 7.9 in the 0 PSI and 13 PSI columns, respectively. If you plotted these two datapoints and connected them with a straight line, the values for the ½ PSI through 12½ PSI columns fall exactly on that line. On one hand I felt optimistic with these values; however, on the other hand I also felt skeptical. Why? Because up this point the values in the boost columns of my fuel map were in the 9’s and 10’s. Nonetheless, I loaded the calculated values and went for a drive.
As I was puttering through the neighborhood, there was a street about one-half-mile long with speed bumps every few hundred feet. On this street I left it in second gear. As I passed a speed bump, I gently gave it gas but the car stumbled before revving up to speed. Kinda like "vruh vruh vruh vroooooom". At light acceleration (not even approaching 0 PSI nor boost) staying in the vacuum range, the air-fuel ratio would go lean to 16:1. This was at no more than about 32-3500 RPM and at very light load. Say, did I mention that it is a good thing I have on on-board FJO?
When I got home, I pulled out some previous fuel maps and every single one of them had a 10.7 under the 0 PSI column. So I globally increased the calculated values in the boost columns by 21.6% such that the 8.8 is now a 10.7 at 0 PSI and all of the linearly reduced values were also ramped up by 21.6%. A subsequent full throttle FJO recording then showed the air-fuel ratio averaged out to 11.6:1. And by maintaining linearly reducing voltage modification factor as boost goes up, the air-fuel ratio graph is by far the smoothest my car has produced.
The only thing that I can think of is this. If you plug in negative 14 PSI (full vacuum) in the voltage modification equation, 11.2 is the calculated value. In the vacuum columns of my fuel map, the cell value is 11.1 across the board. Coincidence or reality? Vacuum columns are used under steady-state cruising. Apparently the ECU freaked out when jumping from 11.2 to 8.8 as it goes from vacuum to boost; however, a transition from 11.2 to 10.7 is easier to accept. Therefore, the SS is telling the ECU to consider even a partial vacuum as full vacuum. Why, I don’t know.
Clearly, there are hidden algorithms inside the ECU that come into play if certain “out of acceptable range” values are entered into the fuel map. There is another facet that comes into play which I suspect may be the source of tuning gremlins. I have a 3.0L OBD-II and it has never thrown a CEL without cause. The two times it did, I had erroneous values in the fuel map. Fix map, run again and the ECU is smiling. However, I suspect that the factory changed wiring on the 3.2L OBD-II cars and that may be the obstacle. Although the OBD-I cars are also 3.0L, for whatever reason the factory changed numbering convention in the wiring connectors from 1994 to 1995. So in my mind, it is entirely plausible that the factory changed something (either in the wiring or in the ECU) again from 1996 to 1997.
The slope of the SS’s MAP voltage graph vs the OEM’s MAP voltage graph is such that under boost, the SS output voltage is greater than the modified* OEM’s MAP output voltage. And the values diverge as boost increases, thus the need for the fuel map values to get smaller and smaller as boost climbs. (*This is based on 500 injectors instead of the factory 270 injectors.) Keep in mind that this is based on theory and does not account for VVIS operation nor VTEC operation. At certain RPM points where my FJO shows a slight leanness or richness, the fuel map value deviates from the straight-line value. The only way to verify how much adjustment is needed is with a wide-band oxygen sensor. My current air-fuel ratio is 11.8:1 average. There is a very slight bump here and a dip there in the plot of the data recording, but it is not like a trend tendency to go rich or lean and never return. Once the respective RPM point is passed, the AFR returns to 11.8:1 all the way to redline. .
If the mathematics of the electronics are this straightforward, you would think that the only variable component that changes with boost is the size of the injector and / or the fuel pressure. Seems like there ought to be only two pre-programmed fuel maps up to the SS's limit of 16 PSI - one for 440 injectors and one for 500 (or 550) injectors. Build the map, pull the clock fuse to reset the ECU, load the map and it should be good to go. This seems like a reasonable assumption. Yet, everyone talks bout "tuning" each engine individually. This shouldn't be necessary since there are no large differences between our NSX engines in terms of volumetric efficiency, even with aftermarket headers & exhaust. The OEMs don't "tune" each vehicle off the production line, even the really high power turbo vehicles.
Since the SS box does not control the fuel enrichment directly, the relationship between theory and reality can deviate quickly, as evidenced a few paragraphs ago. For example, what happens when you are at part throttle driving up a hill at constant speed and the boost is a few PSI (not maximum boost). The MAP voltage that the ECU sees will be considerably less than WOT, so the ECU may attempt to run in closed loop mode. This is bad for the engine since the ECU will attempt to lean out the air-fuel ratio to bring it back to stoichiometric. This may explain some of the engine failures.
I have a copy of the “Killer” fuel / timing map. The date stamp on the file is July 26, 2002. July 2002 was the timetable when MB was making site installs for the BBSC in Florida. I recall a thread where MB tuned an NSX to 403 RWHP and in that communiqué he said something along the lines of "And it had a killer graph." Don't know if he meant the air-fuel ratio curve or what, but perhaps that's how the map file was named Killer.
A few days ago I created a new map based on my calculations that netted an average 11.6:1 AFR. This was with the 13 PSI column with a value of 7.9 and the 0 PSI column at 10.7 PSI. I held the 10.7 constant, changed the 7.9 to an 8.9 and autofilled in between. This dropped to average 11.2:1 AFR. After starring at the before-and-after FJO recording plots, I arrived at this observation. Raising the 13 PSI column by 1 (7.9 to 8.9) dropped the AFR by 0.4:1 average. But I wanted 12:1 average. So by logic, if I reduce the 13 PSI column by 1 (7.9 to 6.9) then it stands to reason that the average should go the other direction by 0.4:1 to the then target of 12:1. Again, I held the 10.7 in the 0 PSI fixed and autofilled the columns in between.
I looked at what Killer’s numbers are. They are 12.5 at 0 PSI and 10.5 at 6.5 PSI. Plotting the datapoint yields a straight line, which is not surprising. You are not going to believe this. I plot my 10.7 at 0 PSI and 6.9 at 13 PSI, and the two lines (between my map and the Killer map) are identically parallel. The shift accounts for the fact that my lower numbers mean more fuel. Why? I have the Novi 2000 and the early Florida installations were Novi 1000. Smaller blower size = less air inducted = less fuel required = different map values.
Well gee wiz, Andy. Your smack certainly is enticing, are you willing to share? Hell no! Just kidding. Go to http://www.baschboost.org/downloads and you will see three files. The first file is the SS software. The second file is my current fuel / timing map. The third file is the resulting air-fuel ratio from file #2.
STRONG CAVEAT - I am in no way suggesting that my fuel / timing map is the “solve all, fix all” that will result in your car’s desired performance. Additionally, I am not advocating that you carte blanche load maps into your SS without care. The only intention of this post is to somewhat explain how the SS works.
http://baschboost.org/forums/viewtopic.php?p=252#252
First off, I need to make something very clear. What you are about to read was not derived all by me. The theory basis was provided by someone else that is widely recognized have having tremendous talent in the area of electronics, with him being an electrical engineer. I have added my experiences, opinions and observations.
Within the Split Second software are two maps. Map A is the fuel map and Map B is the timing map. Let’s begin with the latter, because it is easier to understand. Within a particular cell, that value represents the amount of timing that is being retarded (taken away) from the factory timing at that “snapshot” of load vs RPM. The amount of timing retard increases with RPM and the rate at which the timing retard increases, increases with rising load / boost. You can see this trend very clearly in Map B. I’ve created three timing maps - conservative, moderate and aggressive...currently running the conservative timing map.
As I understand timing, the correct math for a forced induction NSX could be to follow the rule of thumb of 1° retard per PSI of boost. So with my old 4.2" pulley that made 8 PSI at redline, the total timing at full throttle would be 28° advance minus 8° retard equals 20° total timing. This compares well the range of numbers that I've read for a boosted NSX of 18° to 22°. There is an NSX making huge HP at 27° of total timing; however, he has a huge aftercooler so he can get away with more aggressive timing.
Unlike the AEM where you tell it what you want the total timing to be, with the SS you just tell it how much timing to retard. The confirmation would be with a timing light on the dyno and / or a Honda scanner. I've tried one Honda scanner (that is outdated and repaired often) but I did not believe the result. If I can't get a believable reading from another scanner, I'll do the timing light trick. Somebody once told me that MB verified timing in this manner.
Now for the fuel map discussion. The OEM’s MAP sensor calibration is different from the SS’s MAP sensor calibration. The whole idea behind the fuel map is to modify the SS’s MAP sensor voltage so that the ECU thinks it is still being fed a signal from the factory MAP sensor. Sounds easy in concept, and to a point it is. A value in a particular boost column is how much the SS’s MAP sensor voltage is modified before the signal is conveyed to the ECU. A value of 10 means there is no signal modification. A value of 11 means the SS’s signal is increased 0.25 V. A value of 9 means the SS’s signal is decreased by 0.25 V. For the most part, increasing the value in the fuel map increases the richness of the air-fuel ratio.....to a point. I had experienced a situation where I was approaching a lean condition towards redline. Increasing the fuel map value in the corresponding boost columns had no affect on the air-fuel ratio. I had numbers in the 10’s and 11’s, but the air-fuel ratio was no responding. That’s when I wiped the slate clean and looked into just how the SS fuel map works.
You know how there is a value of 13.8 in the 500 RPM in 0 PSI column? Wonder how it is arrived at? At standard atmospheric pressure, the OEM’s MAP sensor to the ECU is 2.825 V. This is the signal voltage the ECU expects to see as the ignition key is turned, just before the engine fires. At standard atmospheric pressure, the SS’s MAP sensor is 1.864 V. The difference is 0.961 V. If you divide 0.961 by 0.25 (the same 0.25 mentioned in the paragraph before this one) you get 3.844. Round off to one decimal point to be consistent with the SS’s values and add to 10 (the same 10 mentioned in the paragraph before this one) you get 13.8. I measured the input voltage to the ECU from the modified SS output voltage is 2.65 V - a difference of 6%. The voltage measurement was made with a twenty year old analog voltmeter. When the key is turned to the ignition position, it is very likely that the stock ECU measures the MAP voltage to ensure that the MAP sensor itself is not broken. It may also perform a calibration of the MAP sensor against the barometric pressure sensor inside the ECU. This type of "sensor checking" is very common with ECUs and is used to identify bad sensors.
The ECU’s algorithm knows that if the incoming voltage is 2.825 V, this is at 0” Hg which means no intake manifold vacuum which means WOT (if the RPM is any number greater than zero). Since the ECU is built around a NA engine, when it gets out of closed loop (vacuum) and goes into open loop (boost) it defaults to an injector pulse width look-up table that Japan wrote to correspond with WOT. With the BBSC / SS, not only is the ECU pulsing the injector longer under WOT vs cruising conditions, the larger injectors are delivering the required amount of fuel to ensure the target air-fuel ratio.
If you know the injector size and what pressure the fuel pump delivers, you can calculate what the theoretical voltage modification should be for each ½ PSI boost column in Map A to yield a certain air-fuel ratio. I emphasize the words “theoretical” and “should” because the real world is different.
For my car (500 injectors, 54 PSI fuel pressure, target 11.8:1 AFR) the calculated modification values are 8.8 and 7.9 in the 0 PSI and 13 PSI columns, respectively. If you plotted these two datapoints and connected them with a straight line, the values for the ½ PSI through 12½ PSI columns fall exactly on that line. On one hand I felt optimistic with these values; however, on the other hand I also felt skeptical. Why? Because up this point the values in the boost columns of my fuel map were in the 9’s and 10’s. Nonetheless, I loaded the calculated values and went for a drive.
As I was puttering through the neighborhood, there was a street about one-half-mile long with speed bumps every few hundred feet. On this street I left it in second gear. As I passed a speed bump, I gently gave it gas but the car stumbled before revving up to speed. Kinda like "vruh vruh vruh vroooooom". At light acceleration (not even approaching 0 PSI nor boost) staying in the vacuum range, the air-fuel ratio would go lean to 16:1. This was at no more than about 32-3500 RPM and at very light load. Say, did I mention that it is a good thing I have on on-board FJO?
When I got home, I pulled out some previous fuel maps and every single one of them had a 10.7 under the 0 PSI column. So I globally increased the calculated values in the boost columns by 21.6% such that the 8.8 is now a 10.7 at 0 PSI and all of the linearly reduced values were also ramped up by 21.6%. A subsequent full throttle FJO recording then showed the air-fuel ratio averaged out to 11.6:1. And by maintaining linearly reducing voltage modification factor as boost goes up, the air-fuel ratio graph is by far the smoothest my car has produced.
The only thing that I can think of is this. If you plug in negative 14 PSI (full vacuum) in the voltage modification equation, 11.2 is the calculated value. In the vacuum columns of my fuel map, the cell value is 11.1 across the board. Coincidence or reality? Vacuum columns are used under steady-state cruising. Apparently the ECU freaked out when jumping from 11.2 to 8.8 as it goes from vacuum to boost; however, a transition from 11.2 to 10.7 is easier to accept. Therefore, the SS is telling the ECU to consider even a partial vacuum as full vacuum. Why, I don’t know.
Clearly, there are hidden algorithms inside the ECU that come into play if certain “out of acceptable range” values are entered into the fuel map. There is another facet that comes into play which I suspect may be the source of tuning gremlins. I have a 3.0L OBD-II and it has never thrown a CEL without cause. The two times it did, I had erroneous values in the fuel map. Fix map, run again and the ECU is smiling. However, I suspect that the factory changed wiring on the 3.2L OBD-II cars and that may be the obstacle. Although the OBD-I cars are also 3.0L, for whatever reason the factory changed numbering convention in the wiring connectors from 1994 to 1995. So in my mind, it is entirely plausible that the factory changed something (either in the wiring or in the ECU) again from 1996 to 1997.
The slope of the SS’s MAP voltage graph vs the OEM’s MAP voltage graph is such that under boost, the SS output voltage is greater than the modified* OEM’s MAP output voltage. And the values diverge as boost increases, thus the need for the fuel map values to get smaller and smaller as boost climbs. (*This is based on 500 injectors instead of the factory 270 injectors.) Keep in mind that this is based on theory and does not account for VVIS operation nor VTEC operation. At certain RPM points where my FJO shows a slight leanness or richness, the fuel map value deviates from the straight-line value. The only way to verify how much adjustment is needed is with a wide-band oxygen sensor. My current air-fuel ratio is 11.8:1 average. There is a very slight bump here and a dip there in the plot of the data recording, but it is not like a trend tendency to go rich or lean and never return. Once the respective RPM point is passed, the AFR returns to 11.8:1 all the way to redline. .
If the mathematics of the electronics are this straightforward, you would think that the only variable component that changes with boost is the size of the injector and / or the fuel pressure. Seems like there ought to be only two pre-programmed fuel maps up to the SS's limit of 16 PSI - one for 440 injectors and one for 500 (or 550) injectors. Build the map, pull the clock fuse to reset the ECU, load the map and it should be good to go. This seems like a reasonable assumption. Yet, everyone talks bout "tuning" each engine individually. This shouldn't be necessary since there are no large differences between our NSX engines in terms of volumetric efficiency, even with aftermarket headers & exhaust. The OEMs don't "tune" each vehicle off the production line, even the really high power turbo vehicles.
Since the SS box does not control the fuel enrichment directly, the relationship between theory and reality can deviate quickly, as evidenced a few paragraphs ago. For example, what happens when you are at part throttle driving up a hill at constant speed and the boost is a few PSI (not maximum boost). The MAP voltage that the ECU sees will be considerably less than WOT, so the ECU may attempt to run in closed loop mode. This is bad for the engine since the ECU will attempt to lean out the air-fuel ratio to bring it back to stoichiometric. This may explain some of the engine failures.
I have a copy of the “Killer” fuel / timing map. The date stamp on the file is July 26, 2002. July 2002 was the timetable when MB was making site installs for the BBSC in Florida. I recall a thread where MB tuned an NSX to 403 RWHP and in that communiqué he said something along the lines of "And it had a killer graph." Don't know if he meant the air-fuel ratio curve or what, but perhaps that's how the map file was named Killer.
A few days ago I created a new map based on my calculations that netted an average 11.6:1 AFR. This was with the 13 PSI column with a value of 7.9 and the 0 PSI column at 10.7 PSI. I held the 10.7 constant, changed the 7.9 to an 8.9 and autofilled in between. This dropped to average 11.2:1 AFR. After starring at the before-and-after FJO recording plots, I arrived at this observation. Raising the 13 PSI column by 1 (7.9 to 8.9) dropped the AFR by 0.4:1 average. But I wanted 12:1 average. So by logic, if I reduce the 13 PSI column by 1 (7.9 to 6.9) then it stands to reason that the average should go the other direction by 0.4:1 to the then target of 12:1. Again, I held the 10.7 in the 0 PSI fixed and autofilled the columns in between.
I looked at what Killer’s numbers are. They are 12.5 at 0 PSI and 10.5 at 6.5 PSI. Plotting the datapoint yields a straight line, which is not surprising. You are not going to believe this. I plot my 10.7 at 0 PSI and 6.9 at 13 PSI, and the two lines (between my map and the Killer map) are identically parallel. The shift accounts for the fact that my lower numbers mean more fuel. Why? I have the Novi 2000 and the early Florida installations were Novi 1000. Smaller blower size = less air inducted = less fuel required = different map values.
Well gee wiz, Andy. Your smack certainly is enticing, are you willing to share? Hell no! Just kidding. Go to http://www.baschboost.org/downloads and you will see three files. The first file is the SS software. The second file is my current fuel / timing map. The third file is the resulting air-fuel ratio from file #2.
STRONG CAVEAT - I am in no way suggesting that my fuel / timing map is the “solve all, fix all” that will result in your car’s desired performance. Additionally, I am not advocating that you carte blanche load maps into your SS without care. The only intention of this post is to somewhat explain how the SS works.
Now that was a post!
You are a good man to share.
You are a good man to share.
well done Andy, that represents alot of time and research and is greatly appreciated. Devin can plug in your formulas the first week in july when he tunes my car. I would put together a "theoretical" map now but I dont know what my final fuel pressure will be. Sure would be nice if we could plug in a few values and create a steady afr like you have done.
