Joined
·
4,423 Posts
Some quick notes for noob StreetTuner users who want to tune their VF39'd STI with a TBE. These comments are for those wanting to produce a better tune for your car but have no clue what to do. I am not an expert and I take no responsibility for anything that you may do to your car. This is your decision. I am not a road racer, so anything below is for 1/4 mile type racing (i.e. relatively short bursts). For extended racing and high speed, high RPM pulls for extended time, more fueling and slightly less timing than suggested would be highly advised. In fact, I would consider using a stock Cobb 2 map for road racing.
First and foremost, install your wide band and log data. At a minimum what do you want to log and why?
What are you looking for in the logs? At first, you just want to really get familiar with things. Get familiar with things? What does that mean? What will I see? What am I looking for? Well ...
The above is just a start. There are a bunch of other things to see too. Furthermore, the above only considers WOT. For part throttle logging you will want to view MAF volts, A/F Learning (long term fuel trim), A/F Correction (short term fuel trim), and even the stock AFR. This type of information can be used to dial in your specific intake or to fine tune the factory intake. More on that later.
So, you've logged data and looked at it ... tens of times. You've learned a bunch of stuff. You are familiar with what's going on. Where do you go from here? This is a tuner's preference. You will see varied suggestions with various reasoning. Personally, I like Boost, then Ignition, then Fueling for a stock VF39 with a TBE (well call this B->I->F). If you have an aftermarket intake then I would suggest tuning that first.
INTAKE TUNING
BOOST
TIMING
FUELING
Also, I wanted to add in that I will be updating this later after adding a Walbro Fuel Pump. I will then be updating it again after adding a Perrin EBCS.
t
Update: Tuning for the Perrin Solenoid
First and foremost, install your wide band and log data. At a minimum what do you want to log and why?
- Throttle Position - (You have to know if/when you are at WOT)
- Load - (Ignition and fuel tables are based on Load)
- Wideband AFR - (You need to know if you are running lean or rich)
- RPM - (Ignition and fuel tables are based on RPM)
- Boost - (Can't tune boost if you don't know what it is)
- Wastegate Duty Cycle - (This determines your boost. The ECM modifies this in an attempt to reach target boost)
- Injector Pulse Width - (Very Important. The STI injectors are undersized from the factory. TBE + EM maxes them)
- Ignition Timing, Dynamic Advance, Dynamic Advance Multiplier and Knock Activity - (Need to log all of these to tune ignition)
What are you looking for in the logs? At first, you just want to really get familiar with things. Get familiar with things? What does that mean? What will I see? What am I looking for? Well ...
- Insert a row beside the Injector Pulse Width (IPW). Add a formula of IPW * RPM / 1200. This is your Injector Duty Cycle (IDC). Populate this formula down the entire column and check out your IDC. At this point you will be saying 'Whoa, they are really high at WOT!'. Welcome to factory STI injectors really suck if you mod your car. Now you know why.
- Look at your Wastegate Duty Cycle (WGDC) and compare it to your actual boost while taking into account the boost target specified in your StreetTuner tune. You'll see some very interesting things at WOT in the upper RPM. What will you see? Well, most of you will see that Cobb's maps never achieve the specified target boost in the upper RPM. You'll see that the WGDC in the log will be maxed at the value specified in your WGDC (high) map in StreetTuner.
- Dynamic Advance (DA) is all over the place ... and will vary on one day compared to the next, or one WOT run compared to the next. You may notice after a couple weeks of logging that your DA is rather small compared to a week or two previous and that your car feels sorta slow. Welcome to the world of Dynamic Advance. Reset the ECU and do a couple WOT runs. Within 3 WOT runs your DA will have completely learned and you will be running max DA (which will most likely be higher than what you just logged).
- Look at the load. At WOT it increases and then decreases. Our VF39's can't support any midhigh-to-high RPM flow so the boost naturally tapers (even if you specify higher boost). Thus, the load decreases as RPM rise. This is an important observation in it's own right. Given that the fuel and ignition maps are based on load, you need to make sure that you look at the StreetTuner maps in their entirety. You can't just look down the column. You have to look from right to left and left to right.
- Look at the load again. You will notice that Cobb's main ignition and fuel tables max out a 2.75 load. A standard Cobb Stage 2 map running 17.5 PSI can hit 3.0-3.10 load. This, IMHO, is unacceptable. A stock Cobb map is already off the chart.
- Check your Ignition Timing. Compare it to your StreetTuner files. You won't be able to fully figure out how you are getting values that you are. You'll also learn that Ignition Timing changes based on Load. Load is proportional to boost. Boost varies based on gear. Thus, you will run different timing in different gears (at the same RPM) when comparing 1st and 2nd to 4th and 5th.
- After a reset the Dynamic Advance Multiplier (aka Ignition Advance Multiplier or IAM) is at 0.5. Simply putting the car in 5th gear at low RPM and holding boost to 4-6 PSI for a few seconds will max it at 1.0 very quickly. Neato Mosquito.
- Knock Activity sucks. It is binary - On and Off. This is a limitation to how you will tune if you want to use this for determining knock. Remember, you are only gathering so many frames a second. It can knock in one fraction of a second and not another. Plus, this assumes you want to depend on the pre-determined ECM boundary for knock (the threshold).
The above is just a start. There are a bunch of other things to see too. Furthermore, the above only considers WOT. For part throttle logging you will want to view MAF volts, A/F Learning (long term fuel trim), A/F Correction (short term fuel trim), and even the stock AFR. This type of information can be used to dial in your specific intake or to fine tune the factory intake. More on that later.
So, you've logged data and looked at it ... tens of times. You've learned a bunch of stuff. You are familiar with what's going on. Where do you go from here? This is a tuner's preference. You will see varied suggestions with various reasoning. Personally, I like Boost, then Ignition, then Fueling for a stock VF39 with a TBE (well call this B->I->F). If you have an aftermarket intake then I would suggest tuning that first.
INTAKE TUNING
You modify the MAF table (Intake Calibration). You basically look at what the A/F learning is at a particular MAF Voltage. Look up that MAF voltage in the Intake Calibration table. Take the value for that voltage and multiply it by the A/F learning.
Make sure that you expand these changes to the adjacent cells and smooth things out a little. You need to modify as many of these as possible, which means you need to log a lot of various driving where you can obtain different MAF Voltages. After making these changes, you have to drive your car around for awhile and let it relearn (or reset the ECU). This is one location where using the cursor tracing feature doesn't help you right away.
Examples:
- +5% A/F learning at 1.20 MAF Volts
Let's Assume the value for 1.20 MAF Volts is 3.1 g/s.
The new value will be 3.1 + (0.05 * 3.1) = 3.255 = 3.26
- -7.2% A/F learning at 1.20 MAF Volts
Let's Assume the value for 1.20 MAF Volts is 3.1 g/s.
The new value will be 3.1 - (0.072 * 3.1) = 3.255 = 2.88
Make sure that you expand these changes to the adjacent cells and smooth things out a little. You need to modify as many of these as possible, which means you need to log a lot of various driving where you can obtain different MAF Voltages. After making these changes, you have to drive your car around for awhile and let it relearn (or reset the ECU). This is one location where using the cursor tracing feature doesn't help you right away.
BOOST
In order to follow the B->I->F you want to first ensure that you are rich across the board once boost hits good (10+ PSI). If you are running greater than 11:1 AFR, (i.e. you are running mid 11s to mid 12s), then I would suggest first tweaking the fuel. You don't want to increase boost when you are running lean. That would just be stupid.
The reason I am suggesting that you verify that you are running rich is because Cobb does something that I do not agree with as specified above. Their fueling tables only go to 2.75 load when at the stock 17.5 PSI settings in Stage 2 you can hit 3.0-3.1 load. Thus, if you push boost up to 18.5+, you are going to be even further off the chart (3.2+).
Let's assume that you are running rich so that you can start with tuning the boost. Most Cobb Stage 2 STI maps (let's assume the 1.10 maps - IIRC, earlier maps were leaner) at sea level with a stock intake will run rich. How do you tweak the boost? What values should you target? This is a sensitive area. Again, lots of tuners have different theories here. I've seen max PSI recommendations ranging from 17.5 to 19.0. However, generally most recommendations fall at about ~18.5 PSI.
This is absolutely not a max boost across all RPM. Why? Well, first of all, the injectors wouldn't support it (as we should now know because we are hitting almost 100% with only 14-15 PSI in the upper RPM. Second, the VF39 can't support that much boost. (Yeah, yeah - some of you are going to say to replace the actuator or add helper springs to help hold boost till redline a little better. We'll get to this later. It's not a bad idea but you have to be extra cautious when doing this.)
18.5 PSI appears to be a safe value to attempt to achieve in the mid RPM. First and foremost, you cannot simply use the Boost Targets table. This is just a target. The ECM has boundaries on how it can utilize the Wastegate Solenoid in order to achieve this boost. Thus, you must also modify your WGDC tables at the same time. Take a look at Cobb's maps and you'll see 17.5 PSI as a midrange target boost. Many StreetTuner users bump those values up to 18.5 PSI, and then set the WGDC table to 75% across the board for WOT and call it a day. Been there, done that. It's not a horrible approach. It does work ok. However, don't start there and don't stop there.
Don't start with 18.5 PSI. Start lower with 17.75 PSI and log. Bump up to 18. Bump up to 18.25. Bump up to 18.5. In each instance, verify your IDCs are ok, your Wideband AFRs are still rich (<11:2), and that you're not pinging. I said don't stop there.
Now we can address a very unique part of boost tuning: the higher RPM. Here is what you will notice with your Cobb Stage 2 setup at sea level with a stock intake: You will never reach boost targets in the upper RPM and even if you did reach them then you can extrapolate to say that your IDCs would be over 100%. You won't reach boost targets for 2 reasons.
I'd love to hit 16.25 PSI at 6400 as specified in the Target Boost map for Cobb Stage 2. In reality, you are going to be lucky to hit 15.5. You can better try to reach these targets by bumping up the WGDC (high) map at 5600 and 6400. With mine bumped up to 85%, I am only hitting 15.0 PSI and have 99% IDCs. I can't imagine hitting 16.25. Rule of thumb is that you never want to have to use really high WGDCs to achieve your target boost. If you think you need to use >90% then the turbo is just too small. Don't put 100% in the WGDC maps and consider not using >90%. Please keep in mind that diminishing returns truly applies to this VF39. Forcing this turbo to produce 16 PSI in the higher RPM (via a helper spring or actuator change) doesn't mean that you are going to be faster. For the VF39, forcing it to do anything above like 15.5 PSI in the upper RPM just means you are blowing really hot air. Keep in mind that 14 PSI of cool air creates more power than 16 PSI of hot air. If you remember that then you are golden. More boost doesn't mean more power.
Also don't underestimate the backpressure you are creating in the pre-turbo plumbing by trying to run higher boost in the upper RPM. Now, with all of that said, adding a helper spring to your VF39 if you are sure you have enough injector to do so isn't a bad idea as long as you keep the boost clamped to no more than 15.5 PSI at redline. However, this is a really sketchy area since you need to be 100% sure you have enough injector and you need to find an appropriate spring that won't produce too much boost. Cool? Cool. Another thing that I want to touch on quickly is optimizing your target boost tables at higher RPM. After pumping up the WGDCs in order to achieve higher boost in the higher RPM, you will become very familiar with the amount of boost that you can produce vs. IDC. I would highly advise that you reduce Cobb's boost values in cases where your turbo isn't capable of producing more boost and your injector duty cycle is about maxed out anyhow.
Ok, I'm done with describing very simple boost tuning on the VF39 with TBE. I know some will reply and say so-and-so said blah, blah, blah and so-and-so said blah, blah, blah. Thanks. Again, "This is a tuner's preference. You will see varied suggestions with various reasoning.". Others will reply and say that 'You just increased boost with no measurable means to record whether you made the car faster or not.'. I fully agree. I would suggest doing this on a dyno or on a drag strip. Did I do that? No. Will I go to the dyno/dragstrip and check my changes later? Absolutely. Do I like how the car drives with the new boost settings? Absolutely.
The reason I am suggesting that you verify that you are running rich is because Cobb does something that I do not agree with as specified above. Their fueling tables only go to 2.75 load when at the stock 17.5 PSI settings in Stage 2 you can hit 3.0-3.1 load. Thus, if you push boost up to 18.5+, you are going to be even further off the chart (3.2+).
Let's assume that you are running rich so that you can start with tuning the boost. Most Cobb Stage 2 STI maps (let's assume the 1.10 maps - IIRC, earlier maps were leaner) at sea level with a stock intake will run rich. How do you tweak the boost? What values should you target? This is a sensitive area. Again, lots of tuners have different theories here. I've seen max PSI recommendations ranging from 17.5 to 19.0. However, generally most recommendations fall at about ~18.5 PSI.
This is absolutely not a max boost across all RPM. Why? Well, first of all, the injectors wouldn't support it (as we should now know because we are hitting almost 100% with only 14-15 PSI in the upper RPM. Second, the VF39 can't support that much boost. (Yeah, yeah - some of you are going to say to replace the actuator or add helper springs to help hold boost till redline a little better. We'll get to this later. It's not a bad idea but you have to be extra cautious when doing this.)
18.5 PSI appears to be a safe value to attempt to achieve in the mid RPM. First and foremost, you cannot simply use the Boost Targets table. This is just a target. The ECM has boundaries on how it can utilize the Wastegate Solenoid in order to achieve this boost. Thus, you must also modify your WGDC tables at the same time. Take a look at Cobb's maps and you'll see 17.5 PSI as a midrange target boost. Many StreetTuner users bump those values up to 18.5 PSI, and then set the WGDC table to 75% across the board for WOT and call it a day. Been there, done that. It's not a horrible approach. It does work ok. However, don't start there and don't stop there.
Don't start with 18.5 PSI. Start lower with 17.75 PSI and log. Bump up to 18. Bump up to 18.25. Bump up to 18.5. In each instance, verify your IDCs are ok, your Wideband AFRs are still rich (<11:2), and that you're not pinging. I said don't stop there.
- Continue to tweak the WGDC high map so that you achieve target boost at the specified high WGDC (or with a few percent to spare) - (this will also serve as a clamp to help control boost creep.)
- You may also consider tweaking the lower WGDC map to provide more room for boost creep. However, this is a double edged sword because it will take longer on an ECU reset to get to the target boost.
Now we can address a very unique part of boost tuning: the higher RPM. Here is what you will notice with your Cobb Stage 2 setup at sea level with a stock intake: You will never reach boost targets in the upper RPM and even if you did reach them then you can extrapolate to say that your IDCs would be over 100%. You won't reach boost targets for 2 reasons.
- The turbo just isn't capable of providing that much.
- The WGDC high map is only 75% in the upper RPM.
I'd love to hit 16.25 PSI at 6400 as specified in the Target Boost map for Cobb Stage 2. In reality, you are going to be lucky to hit 15.5. You can better try to reach these targets by bumping up the WGDC (high) map at 5600 and 6400. With mine bumped up to 85%, I am only hitting 15.0 PSI and have 99% IDCs. I can't imagine hitting 16.25. Rule of thumb is that you never want to have to use really high WGDCs to achieve your target boost. If you think you need to use >90% then the turbo is just too small. Don't put 100% in the WGDC maps and consider not using >90%. Please keep in mind that diminishing returns truly applies to this VF39. Forcing this turbo to produce 16 PSI in the higher RPM (via a helper spring or actuator change) doesn't mean that you are going to be faster. For the VF39, forcing it to do anything above like 15.5 PSI in the upper RPM just means you are blowing really hot air. Keep in mind that 14 PSI of cool air creates more power than 16 PSI of hot air. If you remember that then you are golden. More boost doesn't mean more power.
Also don't underestimate the backpressure you are creating in the pre-turbo plumbing by trying to run higher boost in the upper RPM. Now, with all of that said, adding a helper spring to your VF39 if you are sure you have enough injector to do so isn't a bad idea as long as you keep the boost clamped to no more than 15.5 PSI at redline. However, this is a really sketchy area since you need to be 100% sure you have enough injector and you need to find an appropriate spring that won't produce too much boost. Cool? Cool. Another thing that I want to touch on quickly is optimizing your target boost tables at higher RPM. After pumping up the WGDCs in order to achieve higher boost in the higher RPM, you will become very familiar with the amount of boost that you can produce vs. IDC. I would highly advise that you reduce Cobb's boost values in cases where your turbo isn't capable of producing more boost and your injector duty cycle is about maxed out anyhow.
Example: Cobb specifies 16.25 PSI at 6400 RPM. With a 90% WGDC and my injectors static I am only getting about 15.0 PSI. There is absolutely no point in me having 16.25 specified. The proper thing to do is to reduce this value to a value that is safe. Reduce it to 15.0 PSI because the injectors can't support more anyhow.
Ok, I'm done with describing very simple boost tuning on the VF39 with TBE. I know some will reply and say so-and-so said blah, blah, blah and so-and-so said blah, blah, blah. Thanks. Again, "This is a tuner's preference. You will see varied suggestions with various reasoning.". Others will reply and say that 'You just increased boost with no measurable means to record whether you made the car faster or not.'. I fully agree. I would suggest doing this on a dyno or on a drag strip. Did I do that? No. Will I go to the dyno/dragstrip and check my changes later? Absolutely. Do I like how the car drives with the new boost settings? Absolutely.
TIMING
Now on to Ignition. Egad. This one sucks. There is enough theory here to fill up pages and pages. Lets get to the point here.
We want to run the minimum amount of timing necessary to produce maximum power and we want this timing to be consistent.
You then also need to make the decision as to whether or not you want SOA designed safety built in. In laymen's terms, we need to know what values to shoot for and we need to know how to achieve them. The values I will list here as to what to shoot for on the stock VF39 with TBE are based on other tuner's work. I don't claim to have come up with these. If you don't want to shoot for blind values listed here then you will need access to various knock indicator tools to determine your own values (knock microphone/stethescope, an indicator like a KnockLink, etc).
The values I have seen others use for TOTAL IGNITION TIMING are:
Moving on, the natural next question is how the heck can you achieve these values given the SOA dynamic advance. How do KC A, KC B, and KC C work with the main ignition in order to produce a predictable ignition advance? I have no clue ... and I have truly tried to understand.
I have had the greatest success with #5 and #6. I am currently using #6 for simplicity and easy tuning since KC C isn't a realtime map. It's pretty easy to tune the WOT timing once you can predict what the values will be. Hypothetically, if you need an extra degree of advance at 5600 RPM and 2.50 load then you simply add 1.0 degree of advance in KC A at 5600 RPM and 2.50 load when you are using method #5 or #6. With #5 you need to copy those changes over to KC B and KC C. With #6 you make the change to KC A and forget about KC B and KC C because the new base map replaces all pointers to KC B and KC C with pointers to KC A. It works and it works great.
The only thing that you need to be concerned about is safety. I can't make that determination for you. However, I can say the following: Any additional knock advance is still in an 'extra map' that is dependent on the Ignition Advance Multiplier. You have added no extra advance to the main ignition so the overall 'additive dynamic amount' will still change based on bad gas due to the IAM. This is how ignition is done on the WRX so it can't be that bad. Last, but not least, it is safer than UTEC because you are still using a base ignition table and an additive table. It is as safe as many ECUTek tunes because they are going to do the same thing that I just explained. You make the call.
We want to run the minimum amount of timing necessary to produce maximum power and we want this timing to be consistent.
You then also need to make the decision as to whether or not you want SOA designed safety built in. In laymen's terms, we need to know what values to shoot for and we need to know how to achieve them. The values I will list here as to what to shoot for on the stock VF39 with TBE are based on other tuner's work. I don't claim to have come up with these. If you don't want to shoot for blind values listed here then you will need access to various knock indicator tools to determine your own values (knock microphone/stethescope, an indicator like a KnockLink, etc).
The values I have seen others use for TOTAL IGNITION TIMING are:
- 18° at 4000 RPM
- 22° at 5000 RPM
- 24° at 6000 RPM
- 28° at 6800 RPM
Moving on, the natural next question is how the heck can you achieve these values given the SOA dynamic advance. How do KC A, KC B, and KC C work with the main ignition in order to produce a predictable ignition advance? I have no clue ... and I have truly tried to understand.
See my tests posted at http://www.imprezawrxsti.com/forums/showthread.php?t=27424.
I've had no luck in truly figuring these out. What I have determined is that the knock advance most clearly follows KC A and that KC B appears almost to be a max limit that the ECM can try to achieve given the right gas. I have no clue what KC C is for. Long story short, here are the various ways that people have tuned the ignition advance to my knowledge ...- Smooth A and B. Leave C alone. Increase A and lower B. Add a little to Main Ign if values of 5 or greater are used for KA. (I've seen Cobb do this)
- Leave all of the KC tables alone. Do all your work in main ignition (I've seen Jorge do this - RiftsWRX).
- Smooth A and B. Leave C alone. Increase both A and B. Leave Main Ign alone. (various users).
- Smooth A, B, and C. Subtract from B and C. Leave Main Ign alone. (various users).
- Set all KC tables to be 100% equal. Leave Main Ign alone. (various ECUTek tuners do this).
- Use a custom base map from Cobb that only uses KC A and thus converts the Dynamic Advance to be similar to a WRX. Leave main ignition alone and put all extra spark advance into KC A. This is very similar to #5 above.
I have had the greatest success with #5 and #6. I am currently using #6 for simplicity and easy tuning since KC C isn't a realtime map. It's pretty easy to tune the WOT timing once you can predict what the values will be. Hypothetically, if you need an extra degree of advance at 5600 RPM and 2.50 load then you simply add 1.0 degree of advance in KC A at 5600 RPM and 2.50 load when you are using method #5 or #6. With #5 you need to copy those changes over to KC B and KC C. With #6 you make the change to KC A and forget about KC B and KC C because the new base map replaces all pointers to KC B and KC C with pointers to KC A. It works and it works great.
The only thing that you need to be concerned about is safety. I can't make that determination for you. However, I can say the following: Any additional knock advance is still in an 'extra map' that is dependent on the Ignition Advance Multiplier. You have added no extra advance to the main ignition so the overall 'additive dynamic amount' will still change based on bad gas due to the IAM. This is how ignition is done on the WRX so it can't be that bad. Last, but not least, it is safer than UTEC because you are still using a base ignition table and an additive table. It is as safe as many ECUTek tunes because they are going to do the same thing that I just explained. You make the call.
FUELING
Now on to fueling. This one isn't so bad and is relatively easy to do. In my opinion, this is the easiest thing to do. Again, as with ignition timing, there is enough theory here to fill up pages and pages.
This one is definitely dependent on not only the tuner, but also on how you will be driving your car. I drive mine for short bursts of power: stoplight to stoplight, 1/4 mile, etc. I don't road race or high speed race on the highway. To keep this relatively short:
Another debated topic is if you should run lean on spoolup or rich on spoolup. Again, you will find many opinions on both and I won't claim to be a good opinion. There is enough information on NASIOC about this that would take you days to read. I like the lean on spoolup theory but, honestly, that means nothing because I am no expert.
This one is definitely dependent on not only the tuner, but also on how you will be driving your car. I drive mine for short bursts of power: stoplight to stoplight, 1/4 mile, etc. I don't road race or high speed race on the highway. To keep this relatively short:
- I strive for very high 10:1 AFR at peak torque and then lean it out to lower 11's as RPM increase. For me this means 10.9:1 at peak torque (often peak boost on a VF39 + TBE STI) with the AFRs leaning out to 11.3 toward redline. You'll often see the exact opposite of this: Lower 11's on boost to high 10's at redline.
- The opposite approach is better for road racing since the extra fuel at extended high RPM driving in the higher RPM can help to slow the burn and thus act as an anti-detonation mechanism.
- Some tuners shoot for 11:1 across the board to provide a good compromise on power and safety. The choice is yours.
Another debated topic is if you should run lean on spoolup or rich on spoolup. Again, you will find many opinions on both and I won't claim to be a good opinion. There is enough information on NASIOC about this that would take you days to read. I like the lean on spoolup theory but, honestly, that means nothing because I am no expert.
Also, I wanted to add in that I will be updating this later after adding a Walbro Fuel Pump. I will then be updating it again after adding a Perrin EBCS.
t
Update: Tuning for the Perrin Solenoid
Right now I am caught up in optimizing what I have. Once I get things dialed in to where I like them then I will go to the track and run the stock Cobb Stage 2 stuff and then run my own map too. So, everything right now is road tuning. I have no proof to offer anybody for what I mentioned above in terms of it actually increasing horsepower. SOTP and Butt dynos suck so I won't say the car feels better 
