Power FC Accel vs. TPS1 tuning strategies
Accel vs. TPS1 tuning strategies
Well guys, I've been playing around with Accel vs. TPS1 and Accelerate Injector settings to fix a slight tip-in problem on my series 4 Turbo II, which only has a narrow-range TPS. Most/all of you have a series 5 or series 6 with the full range TPS so you may not have as much trouble with irritating tip-in lean spots as I have, but I want to propose a strategy here for tuning it that has been working for me. The strategy consists of first setting your TPS, and making sure your INJ/basemap is ok as a preliminary step. Then you try to ascertain what type (how quick the pedal has to move) of throttle movements cause a lean stumble so that you can rescale the input and setting columns of Accel vs. TPS1, followed by some fine tuning of the Accelerate Injector settings.
TPS adjustment info is readily available, and getting your fuel map tuned decently mostly involves datalogging the map with steady throttle movements (as opposed to quick stabs which utilize a lot of the Accel settings and only make tuning more confusing). Chuck's notes discuss this. For s4 Turbo II users I have my VTA1 and VTA2 showing the same voltage. I tied in the full range TPS wire together with the narrow range on the Banzai Racing Adapter. Voltage is .75ish with the car warmed up and no throttle, and 4.77ish at WOT. This is on a brand new s4 TPS from Mazda.
Here are the stock Apex'i settings for Accel vs. TPS 1 (units are out of 256) :
Input - Setting
103 - 256
37 - 230
20 - 25
The first thing I did was change the setting for the lowest set from 25 to 50 , after I had enriched all the Accelerate Injector settings around .2 msec . That helped some, but the car still has a few random stumbles where the wideband pegs off-the-charts lean. It's nothing that seriously affects performance, but it's just kind of annoying, so I decided to experiment a little with the settings. Here are my settings currently, which are a great improvement over before but are still not perfect:
Input - Setting
70 - 256
26 - 170
8 - 6
What I have essentially done here is have the accel enrichments kick in earlier, to eliminate those complete lean spots, and then apply less fuel so it then doesn't peg rich. Remember though that if your TPS and fuel map aren't set properly you are going to make the car run worse before it runs better.
To figure these numbers out, the first thing I did was try very, very quick and slight (pedal barely depressed) throttle movements to see if the wideband pegs lean, which is indicative IMO of the "input" setting on Accel vs. TPS 1 not set correctly. You can do it sitting in your driveway, but I prefer to do it around 20-25 mph in 2nd with no traffic around. I kept lowering the lowest "input" setting until I no longer had any dead lean spots with these tiny, quick throttle stabs. I then had to lower the "setting" number significantly, lower than stock, to keep it from dumping in fuel. I used a similar strategy with "medium" throttle movements, the middle row of the Accel TPS1 stuff. I kept lowering the input setting until I wasn't seeing any super lean spots. Then I trimmed the "setting" values. I repeated the same strategy with the top set of numbers. Note that I had no scientific basis for judging what a small, medium, or large throttle movement is, except that I looked at the VTA voltages in the commander so I could make some educated guesses.
Finally, I went back into accelerate injector. I especially played around with the 1000 rpm settings because my tip-in off idle was now too rich.
My current settings (I think I may lean these out a little bit more):
5000 7.70 1.00
4000 7.70 1.00
3000 8.10 1.50
2000 8.10 1.80
1000 6.70 2.00
So in conclusion, I've found that a useful strategy if you have a few nagging tip-in lean spots is to use the "input" column to rescale the Accel Vs TPS1 setting to first eliminate any dead lean areas. Then you trim out the fuel with the "setting" column so it doesn't peg rich, with final adjustments made with the Accelerate Injector function.
Any thoughts on this? I think the main thing that prompted me to try this is that I have only a narrow range TPS, and I am in fact a little OCD about tuning these days.
TPS adjustment info is readily available, and getting your fuel map tuned decently mostly involves datalogging the map with steady throttle movements (as opposed to quick stabs which utilize a lot of the Accel settings and only make tuning more confusing). Chuck's notes discuss this. For s4 Turbo II users I have my VTA1 and VTA2 showing the same voltage. I tied in the full range TPS wire together with the narrow range on the Banzai Racing Adapter. Voltage is .75ish with the car warmed up and no throttle, and 4.77ish at WOT. This is on a brand new s4 TPS from Mazda.
Here are the stock Apex'i settings for Accel vs. TPS 1 (units are out of 256) :
Input - Setting
103 - 256
37 - 230
20 - 25
The first thing I did was change the setting for the lowest set from 25 to 50 , after I had enriched all the Accelerate Injector settings around .2 msec . That helped some, but the car still has a few random stumbles where the wideband pegs off-the-charts lean. It's nothing that seriously affects performance, but it's just kind of annoying, so I decided to experiment a little with the settings. Here are my settings currently, which are a great improvement over before but are still not perfect:
Input - Setting
70 - 256
26 - 170
8 - 6
What I have essentially done here is have the accel enrichments kick in earlier, to eliminate those complete lean spots, and then apply less fuel so it then doesn't peg rich. Remember though that if your TPS and fuel map aren't set properly you are going to make the car run worse before it runs better.
To figure these numbers out, the first thing I did was try very, very quick and slight (pedal barely depressed) throttle movements to see if the wideband pegs lean, which is indicative IMO of the "input" setting on Accel vs. TPS 1 not set correctly. You can do it sitting in your driveway, but I prefer to do it around 20-25 mph in 2nd with no traffic around. I kept lowering the lowest "input" setting until I no longer had any dead lean spots with these tiny, quick throttle stabs. I then had to lower the "setting" number significantly, lower than stock, to keep it from dumping in fuel. I used a similar strategy with "medium" throttle movements, the middle row of the Accel TPS1 stuff. I kept lowering the input setting until I wasn't seeing any super lean spots. Then I trimmed the "setting" values. I repeated the same strategy with the top set of numbers. Note that I had no scientific basis for judging what a small, medium, or large throttle movement is, except that I looked at the VTA voltages in the commander so I could make some educated guesses.
Finally, I went back into accelerate injector. I especially played around with the 1000 rpm settings because my tip-in off idle was now too rich.
My current settings (I think I may lean these out a little bit more):
5000 7.70 1.00
4000 7.70 1.00
3000 8.10 1.50
2000 8.10 1.80
1000 6.70 2.00
So in conclusion, I've found that a useful strategy if you have a few nagging tip-in lean spots is to use the "input" column to rescale the Accel Vs TPS1 setting to first eliminate any dead lean areas. Then you trim out the fuel with the "setting" column so it doesn't peg rich, with final adjustments made with the Accelerate Injector function.
Any thoughts on this? I think the main thing that prompted me to try this is that I have only a narrow range TPS, and I am in fact a little OCD about tuning these days.
Here are my latest settings:
65 256
23 150
6 9
Accelerate injector (decay boxes are stock):
7.4
7.4
8.0
8.0
6.1
This was accompanied by a lot of adjustments to the fuel map.
I've also been working a little more on figuring out how to tune cells that are heavily affected by both Accel enrichments as well as the basemap. This is for around town driving and throttle response when going WOT. What I've found is that a good way to get an idea of whether you are getting any significant Accel enrichments (as opposed to fuel being mostly determined by just the fuel map) is to set up 4 graphs: datalog Advanced InjFrPr , TPS voltage, wideband, and RPM , all with their own separate scales so that visually the graphs make sense and don't confuse you. Then use the ghost feature in FC Edit to highlight the cells in the INJ/Basemap. Trace through the cells with the slider (have a bunch of watches open as well to view all this data). Look for the lean and rich spots that you are trying to fix. If you see a lean spot, see what InjFrPr reads and look at the curve on the chart. If there is a spike in this curve, and datalogged pulsewidths are much higher than what is in the BASE * INJ (in the base tab) cell(s), then you know that there is a good chance that accel enrichments are at least kicking in. Be mindful of the other corrective factors on pulsewidth such as temperature maps.
How does this help you? If you see steady throttle voltage, steady pulsewidth close to BASE * INJ (be careful because under enough load the secondarys kick in), and the cell is too rich, then try changing the basemap because Accel probably isn't doing too much. In another example, if you see pulsewidth spiking and AFR's are too lean, then you may still need to adjust Accel settings in some way. The PFC is adding fuel due to a change in throttle position, but not enough (or the fuel map is just way off. confusing, isn't it?) What gets tricky though is that MAP could be changing rapidly as well so the car is going into another cell and that is in turn increasing pulsewidth. That is why it is important to compare logged pulsewidth to BASE * INJ pulsewidth by looking at the map. I just wish I could figure out a way to datalog which input box under Accel vs. TPS1 that the PFC is using at any given point so that rescaling it could be more precise.
Hope this helps a little bit.
65 256
23 150
6 9
Accelerate injector (decay boxes are stock):
7.4
7.4
8.0
8.0
6.1
This was accompanied by a lot of adjustments to the fuel map.
I've also been working a little more on figuring out how to tune cells that are heavily affected by both Accel enrichments as well as the basemap. This is for around town driving and throttle response when going WOT. What I've found is that a good way to get an idea of whether you are getting any significant Accel enrichments (as opposed to fuel being mostly determined by just the fuel map) is to set up 4 graphs: datalog Advanced InjFrPr , TPS voltage, wideband, and RPM , all with their own separate scales so that visually the graphs make sense and don't confuse you. Then use the ghost feature in FC Edit to highlight the cells in the INJ/Basemap. Trace through the cells with the slider (have a bunch of watches open as well to view all this data). Look for the lean and rich spots that you are trying to fix. If you see a lean spot, see what InjFrPr reads and look at the curve on the chart. If there is a spike in this curve, and datalogged pulsewidths are much higher than what is in the BASE * INJ (in the base tab) cell(s), then you know that there is a good chance that accel enrichments are at least kicking in. Be mindful of the other corrective factors on pulsewidth such as temperature maps.
How does this help you? If you see steady throttle voltage, steady pulsewidth close to BASE * INJ (be careful because under enough load the secondarys kick in), and the cell is too rich, then try changing the basemap because Accel probably isn't doing too much. In another example, if you see pulsewidth spiking and AFR's are too lean, then you may still need to adjust Accel settings in some way. The PFC is adding fuel due to a change in throttle position, but not enough (or the fuel map is just way off. confusing, isn't it?) What gets tricky though is that MAP could be changing rapidly as well so the car is going into another cell and that is in turn increasing pulsewidth. That is why it is important to compare logged pulsewidth to BASE * INJ pulsewidth by looking at the map. I just wish I could figure out a way to datalog which input box under Accel vs. TPS1 that the PFC is using at any given point so that rescaling it could be more precise.
Hope this helps a little bit.
Thread
Thread Starter
Forum
Replies
Last Post
demetlaw
3rd Generation Specific (1993-2002)
6
Oct 2, 2015 06:22 PM
