Boost control pills and ECU tweaks
Jun 8, 2007, 03:03 PM
#1
Boost control pills and ECU tweaks
I started this thread to discuss the options and fine tuning required when replaceing or eliminating BC restrictor pills with stock ECU control .
i'd like to see a definitive guide to setups and what the various changes in pills and pill sizes do. This includes WGA preload, WGA type, pill size and type ( straight drill or venturi style ), pill location, clamps, line ID for the hardware. it includes the table tweaks in the stock ecu to deliver a boost level and any adjustments to help minimise taper.
i think it would be helpful if the setup could be broken down to IX's and VIII's
When posting data, please list any supporting logged data or screen shots of AFR, boost levels, timing : before and after a mod would be nice
i'll start this off by describing pill sizes and location on my 05 mr. These are handmade from 1/4" nylon rod, turbo pill is 56 and BCS pill is 52. Both were machined on a micro-lathe to have a venturi design
| \ / |
| / \ |
|/ \|
the narrowest ID is as listed above. i installed these as follows ( BTW i recd the car with no lines and no pills on it !!!!! ) turbo pill at the T for the WGA oriented so the smaller v is to the turbo. the BCS pill is at the BCS solenoid with the larger v pointing to the BCS . The lines are secured with very small ring clamps, the line itself is made of 3/16 fuel line. The "t' is from Advance auto, and is a ribbed design no restrictor in it. ( yes iknow you an get them with it in. ) . Additionally i have ring clamps retaining the pills in place.
discussion:
why did I use nylon? very easy to cut and modify, and very cheap
What did I use to drill the holes : microlathe and a set of 61-80 bits along with 42-60 bits. McMaster Carr and Grainger is where i purchased the bits.
how did I cut the taper ? 30 and 45 degree cutting bits on the micro lathe
why did i use clamps to retain the pills ? Even with a tight fit the pills were leaking air past them at 20+ psi. they have to be a very tight fit 5/16 would be better OD.
why a venturi design ( similar to stock ) the easy answer is the mitsu engineers went to school, im just a a guy tuning ... seriously I believe they are using the design to smooth the air flow and eliminate a central high speed jet flow that would be harder to control.
Before we get to sizing discussions, let me identify the tables in the VIII ECU to work with: Boost Desired Engine Load #2 ( BDEL ) , Turbo Boost Error Correction ( TBEC ), Boost Control Load offset ( BCLO ) , Wastegate duty cycle #2 ( WGDC ), Boost limit ( BL )
Back on topic. Size discussion:
Let me describe to the best of my knowledge how the control system works.
Signal air is sourced from the compressor nipple and delivered to the wastegate actuator ( WGA ) to open the wastegate, thus stopping the turbo from overboosting. Control over how much and how fast the wastegate opens is by the boost control solenoid ( BCS ), the internal spring pressure of the WGA, and the exhaust gas pressure trying to open the WG .
The BCS works by venting the signal air to the vent line returning to the turbo inlet tube. It is metered air so it needs to go back in there. More air is vented by increasing the WGDC, less air by decreasing the WGDC. Increased WGDC = higher boost, as the signal air that the WGA sees is less, thus it doesnt open as far. decreasing the WGDC allows the WGA to see more signal and open further, thus less boost.
i am not sure of the size of the factory pills, but for this discussion, lets say the turbo pill is 54 and the BCS pill is 52. Without changing any table values, if we use a larger bcs pill, it will allow more air to be vented, thus allowing the WGA to stay closed longer . If we use a smaller BCS pill, it will reduce the air vented, allowing the WGA to open sooner.
So a larger bcs pill allows more boost and may reduce taper. why not just eliminate it ? some tuners do and adjust the tables accordingly.
On to the turbo pill. it passes air from the turbo to the WGA and BCS. If more air is available, the WGA will open more readily, reducing boost. if less air is available, the WGA will be open less readily. So use a smaller turbo pill to increase the boost.
Putting it together: if we went with a 56 turbo pill and 50 BCS pill, we would get more boost due to less air signal, and hold boost more due to venting more air through the bcs.
How does the internal spring pressure of the WG factor in ? Well the factory WG has a 8-12 psi spring in it. Meaning it can hold closed up to 12 psi. The degree of opening is then controlled by the BCS and air available and how much exhaust pressure is trying to open the WG. If we are running factory boost say 17 psi at 3200, the WG may be open 20 degrees, allowing some of the exhaust to bypass, thus slowing the compressor wheel. If we use a smaller turbo pill, the WGA sees less air and will open less, diverting less exhaust and allowing more boost.
Aftermarket WGA have a stronger spring to hold closed more. IIRC the Forge ones can be adjusted from 12 -23 depending on the spring in them ( shameless plug Mike@Forge is gonna owe me ) The downside is that boost will be at the spring pressure ie 18 psi and the BCS wont be able to control that very well since it can only vent so much air.
So how does exhaust pressure have to do in the equation ? At higher rpms, the amount of exhaust flow is much higher, this forces the WG open introducing our friend taper. Remember the air being supplied to the WGA is helping to open it too. So to deal with this we can reduce the air supply to the WGA by increasing the WGDC at higher rpms ( Ralliart roms do this ) or by reducing the amount of air available ( smaller turbo pill ) This is not 100 % effective though as the exhaust continues to build and eventually boost falls off.
End of theory section
see further posts
i'd like to see a definitive guide to setups and what the various changes in pills and pill sizes do. This includes WGA preload, WGA type, pill size and type ( straight drill or venturi style ), pill location, clamps, line ID for the hardware. it includes the table tweaks in the stock ecu to deliver a boost level and any adjustments to help minimise taper.
i think it would be helpful if the setup could be broken down to IX's and VIII's
When posting data, please list any supporting logged data or screen shots of AFR, boost levels, timing : before and after a mod would be nice
i'll start this off by describing pill sizes and location on my 05 mr. These are handmade from 1/4" nylon rod, turbo pill is 56 and BCS pill is 52. Both were machined on a micro-lathe to have a venturi design
| \ / |
| / \ |
|/ \|
the narrowest ID is as listed above. i installed these as follows ( BTW i recd the car with no lines and no pills on it !!!!! ) turbo pill at the T for the WGA oriented so the smaller v is to the turbo. the BCS pill is at the BCS solenoid with the larger v pointing to the BCS . The lines are secured with very small ring clamps, the line itself is made of 3/16 fuel line. The "t' is from Advance auto, and is a ribbed design no restrictor in it. ( yes iknow you an get them with it in. ) . Additionally i have ring clamps retaining the pills in place.
discussion:
why did I use nylon? very easy to cut and modify, and very cheap
What did I use to drill the holes : microlathe and a set of 61-80 bits along with 42-60 bits. McMaster Carr and Grainger is where i purchased the bits.
how did I cut the taper ? 30 and 45 degree cutting bits on the micro lathe
why did i use clamps to retain the pills ? Even with a tight fit the pills were leaking air past them at 20+ psi. they have to be a very tight fit 5/16 would be better OD.
why a venturi design ( similar to stock ) the easy answer is the mitsu engineers went to school, im just a a guy tuning ...
seriously I believe they are using the design to smooth the air flow and eliminate a central high speed jet flow that would be harder to control.Before we get to sizing discussions, let me identify the tables in the VIII ECU to work with: Boost Desired Engine Load #2 ( BDEL ) , Turbo Boost Error Correction ( TBEC ), Boost Control Load offset ( BCLO ) , Wastegate duty cycle #2 ( WGDC ), Boost limit ( BL )
Back on topic. Size discussion:
Let me describe to the best of my knowledge how the control system works.
Signal air is sourced from the compressor nipple and delivered to the wastegate actuator ( WGA ) to open the wastegate, thus stopping the turbo from overboosting. Control over how much and how fast the wastegate opens is by the boost control solenoid ( BCS ), the internal spring pressure of the WGA, and the exhaust gas pressure trying to open the WG .
The BCS works by venting the signal air to the vent line returning to the turbo inlet tube. It is metered air so it needs to go back in there. More air is vented by increasing the WGDC, less air by decreasing the WGDC. Increased WGDC = higher boost, as the signal air that the WGA sees is less, thus it doesnt open as far. decreasing the WGDC allows the WGA to see more signal and open further, thus less boost.
i am not sure of the size of the factory pills, but for this discussion, lets say the turbo pill is 54 and the BCS pill is 52. Without changing any table values, if we use a larger bcs pill, it will allow more air to be vented, thus allowing the WGA to stay closed longer . If we use a smaller BCS pill, it will reduce the air vented, allowing the WGA to open sooner.
So a larger bcs pill allows more boost and may reduce taper. why not just eliminate it ? some tuners do and adjust the tables accordingly.
On to the turbo pill. it passes air from the turbo to the WGA and BCS. If more air is available, the WGA will open more readily, reducing boost. if less air is available, the WGA will be open less readily. So use a smaller turbo pill to increase the boost.
Putting it together: if we went with a 56 turbo pill and 50 BCS pill, we would get more boost due to less air signal, and hold boost more due to venting more air through the bcs.
How does the internal spring pressure of the WG factor in ? Well the factory WG has a 8-12 psi spring in it. Meaning it can hold closed up to 12 psi. The degree of opening is then controlled by the BCS and air available and how much exhaust pressure is trying to open the WG. If we are running factory boost say 17 psi at 3200, the WG may be open 20 degrees, allowing some of the exhaust to bypass, thus slowing the compressor wheel. If we use a smaller turbo pill, the WGA sees less air and will open less, diverting less exhaust and allowing more boost.
Aftermarket WGA have a stronger spring to hold closed more. IIRC the Forge ones can be adjusted from 12 -23 depending on the spring in them ( shameless plug Mike@Forge is gonna owe me
) The downside is that boost will be at the spring pressure ie 18 psi and the BCS wont be able to control that very well since it can only vent so much air. So how does exhaust pressure have to do in the equation ? At higher rpms, the amount of exhaust flow is much higher, this forces the WG open introducing our friend taper. Remember the air being supplied to the WGA is helping to open it too. So to deal with this we can reduce the air supply to the WGA by increasing the WGDC at higher rpms ( Ralliart roms do this ) or by reducing the amount of air available ( smaller turbo pill ) This is not 100 % effective though as the exhaust continues to build and eventually boost falls off.
End of theory section
see further posts
Jun 8, 2007, 03:31 PM
#2
My MR: Stock engine, exhaust everything, except sway bar in rear. Seats dont count as power mods
i started with a 60 Turbo pill and 58 BCS pill, this allowed ~20 psi but tapered to 14-15 by 7000. this is with the stock wga and stock tables.
I then increased the wgdc to 100 and got 21 psi and taper of 15-16. After experimenting, i settled on a 56 BCS pill that allowed less taper 21 peak and 17 taper. This boost peak was at 3500
after a short period, the plastic factory DV in the VIIi developed cracks and boost dropped off to 12 psi. I replaced it with one from Forge, as they are adjustable and rebuildable and made from billet aluminum. i set it up with a 18 psi spring and was happy to see 23 peak and 18 taper. i reduced the wgdc to get back to 21 and 17 again.
I then tried a forge WGA with a 18 psi spring. No change in pills or tables. System overboosted going to 30+ on first roll on pull. I had set the Forge WGA actuator to the same setting as the factory one which is ~ 1/2 hole offset shorter. i attempted to reduce the boost via WGDC without switching pills and was able to get it down to 25 - 27 without much taper ( 24-25 )
not good for engine life, so I adjusted the pill size down to 56 turbo and 53 BCS, allowing for 23-25 with 21-22 taper.
Not satisfied, i adjusted the WGA to have 1 turn offset only about 1/16" . this was much better, allowing 22-21 peak and taper of 19-20. Safe but fast. a heck of a lot faster than the 8 psi the car was originally boosting when i bought her.
Tables and data to follow, I may have to digsome up off my old hardrive.
Podium is open gentlemen.....
PS the diagram of the pill above is not lined up right it should be wider at the bottom but I cant get the editor to do that properly
milburn
i started with a 60 Turbo pill and 58 BCS pill, this allowed ~20 psi but tapered to 14-15 by 7000. this is with the stock wga and stock tables.
I then increased the wgdc to 100 and got 21 psi and taper of 15-16. After experimenting, i settled on a 56 BCS pill that allowed less taper 21 peak and 17 taper. This boost peak was at 3500
after a short period, the plastic factory DV in the VIIi developed cracks and boost dropped off to 12 psi. I replaced it with one from Forge, as they are adjustable and rebuildable and made from billet aluminum. i set it up with a 18 psi spring and was happy to see 23 peak and 18 taper. i reduced the wgdc to get back to 21 and 17 again.
I then tried a forge WGA with a 18 psi spring. No change in pills or tables. System overboosted going to 30+ on first roll on pull. I had set the Forge WGA actuator to the same setting as the factory one which is ~ 1/2 hole offset shorter. i attempted to reduce the boost via WGDC without switching pills and was able to get it down to 25 - 27 without much taper ( 24-25 )
not good for engine life, so I adjusted the pill size down to 56 turbo and 53 BCS, allowing for 23-25 with 21-22 taper. Not satisfied, i adjusted the WGA to have 1 turn offset only about 1/16" . this was much better, allowing 22-21 peak and taper of 19-20. Safe but fast. a heck of a lot faster than the 8 psi the car was originally boosting when i bought her.
Tables and data to follow, I may have to digsome up off my old hardrive.
Podium is open gentlemen.....
PS the diagram of the pill above is not lined up right it should be wider at the bottom but I cant get the editor to do that properly
milburn
Last edited by wrcwannabe; Jun 8, 2007 at 03:33 PM.
Jun 9, 2007, 11:32 AM
#6
Spdracerut: Thanks for the link to the thread. Yes thats that you can do with the BCS pill removed and the stockpill at the turbo
Nice work. Other threads have mentioned that the #1 and # 3 WGDC tables are more than likely for cold operation or fall back in case of trouble.
I am in the process of trying my combination ( Forge WGA and Forge DV ) with the BCS restrictor removed and a 56 turbo pill, but despite almost zeroing the WGDC I still get 25 psi. So it looks like I need to re-adjust the WGA, and start from there.
I eventually want to go with my idea of a duplicate USDM BCS solenoid wired in parallel, using half the duty cycle values to start and ramping both up simultaneously.
Meanwhile on the error correction chart spdracerut is showing, he has low values inserted. I have not tried this. The way it was explained to me is that those numbers are what the ECU tries to add or subtract to the wgdc depending what the difference in actual load and desired load is toadjust the desired load. So for my next experiment i will try using back to back pulls with only those changes in place.
Please post more data !!
Nice work. Other threads have mentioned that the #1 and # 3 WGDC tables are more than likely for cold operation or fall back in case of trouble.
I am in the process of trying my combination ( Forge WGA and Forge DV ) with the BCS restrictor removed and a 56 turbo pill, but despite almost zeroing the WGDC I still get 25 psi. So it looks like I need to re-adjust the WGA, and start from there.
I eventually want to go with my idea of a duplicate USDM BCS solenoid wired in parallel, using half the duty cycle values to start and ramping both up simultaneously.
Meanwhile on the error correction chart spdracerut is showing, he has low values inserted. I have not tried this. The way it was explained to me is that those numbers are what the ECU tries to add or subtract to the wgdc depending what the difference in actual load and desired load is toadjust the desired load. So for my next experiment i will try using back to back pulls with only those changes in place.
Please post more data !!
Last edited by wrcwannabe; Jun 10, 2007 at 07:49 AM.
Jun 9, 2007, 08:13 PM
#7
Account Disabled
this morning I removed the pill at the solonoid and changed the boost corrections maps a bit, and turned the Wast. duty cycle down to 92.5 and after all day logging, i am able to get 21 psi peak holding 19-20 till redline, With ocasional 22 peak. No knock problems.
Trending Topics
Jun 10, 2007, 07:10 AM
#8
Link to MrFred thread and data he posted about logging in 4th vs 3rd. Note the difference in the boost profile, 4th gear hits peak at ~4000 3rd gear hits peak ~3500.
The reason Im pointing you guys there is four-fold:
1. mrfred logs with a 3 bar JDM EVO map sensor which is the right way to do boost logging. He can source these for you.
2. To illustrate that taper is sooner and larger in lower gears, whilst peak is hit sooner too. This is because the engine speed is increasing so rapidly and exhaust volume increase is so rapid that the control system cant keep up in lower gears. The wastegate opens sooner and farther. The ideal system will take into account what gear you are in and apply a different map. As far as i can figure out this will need a dual solenoid system and require WGDC of 50% in higher gears and closer to 75-100 in 1-2-3 gears. It may also require larger and smaller values in the TBEC table .
3. 4th gear pulls are inherently dangerous. An alternative is to log 2nd up a steep grade, putting more load on while keeping speeds down.
4. Have a friend do the logging while you drive
Milburn
The reason Im pointing you guys there is four-fold:
1. mrfred logs with a 3 bar JDM EVO map sensor which is the right way to do boost logging. He can source these for you.
2. To illustrate that taper is sooner and larger in lower gears, whilst peak is hit sooner too. This is because the engine speed is increasing so rapidly and exhaust volume increase is so rapid that the control system cant keep up in lower gears. The wastegate opens sooner and farther. The ideal system will take into account what gear you are in and apply a different map. As far as i can figure out this will need a dual solenoid system and require WGDC of 50% in higher gears and closer to 75-100 in 1-2-3 gears. It may also require larger and smaller values in the TBEC table .
3. 4th gear pulls are inherently dangerous. An alternative is to log 2nd up a steep grade, putting more load on while keeping speeds down.
4. Have a friend do the logging while you drive
Milburn
Jun 10, 2007, 07:13 AM
#9
Thanks for the reply
Milburn
Last edited by wrcwannabe; Jun 10, 2007 at 07:15 AM.
Jun 10, 2007, 07:41 AM
#10
Here are my current tables as of last night:
TBEC
-20 20
-17.5 17
-15 15
-12.5 12
-10 10
-7.5 8
-5 5
-2.5 2
0 0
2.5 -2
5 -5
7.5 -7.5
10 -10
12.5 -12
15 -16
17.5 -20
19.8 -20
WGDC table 2 starting at 2750
100
50
45
50
55
60
65
70
75
80
Setup Forge WGA 18 psi at 1 turn peload from slipping on Forge EVO 15 DV 23psi spring Turbo pill 56 BCS pill 52. Everything else is stock but fuel and timing maps have been modified to keep knocksum in check on 93 octane.
This is allowing me to peak 23 and hold 22 in 3rd, 22/ 21 in 2nd . I get several counts of 1-2 @WOT in second gear between 2500 - 3500 nothing anywhere else. Outside temps yesterday in Atlanta were 93F and 50% humidity.
The change in WGA and DV allows me to run larger pills closer to stock size. Using a stock WGA you are gonna need 58-60 at the turbo and will have some difficulty controlling boost peak and creep. A pill at the BCS will help control this.
I know i can get more power and faster spoolup with a TBE, but this is very quiet, and to me very fast, thus lowering my profile to the cops except the wing
And it takes the autox'ers by surprise 
I hope this helps. Please post any data you might have and questions so that other people can learn.
Here is one for the Logworks guys: what changes are you making to the protocol file for Logworks 3 to log 2 -byte load ?
milburn
TBEC
-20 20
-17.5 17
-15 15
-12.5 12
-10 10
-7.5 8
-5 5
-2.5 2
0 0
2.5 -2
5 -5
7.5 -7.5
10 -10
12.5 -12
15 -16
17.5 -20
19.8 -20
WGDC table 2 starting at 2750
100
50
45
50
55
60
65
70
75
80
Setup Forge WGA 18 psi at 1 turn peload from slipping on Forge EVO 15 DV 23psi spring Turbo pill 56 BCS pill 52. Everything else is stock but fuel and timing maps have been modified to keep knocksum in check on 93 octane.
This is allowing me to peak 23 and hold 22 in 3rd, 22/ 21 in 2nd . I get several counts of 1-2 @WOT in second gear between 2500 - 3500 nothing anywhere else. Outside temps yesterday in Atlanta were 93F and 50% humidity.
The change in WGA and DV allows me to run larger pills closer to stock size. Using a stock WGA you are gonna need 58-60 at the turbo and will have some difficulty controlling boost peak and creep. A pill at the BCS will help control this.
I know i can get more power and faster spoolup with a TBE, but this is very quiet, and to me very fast, thus lowering my profile to the cops except the wing
And it takes the autox'ers by surprise I hope this helps. Please post any data you might have and questions so that other people can learn.
Here is one for the Logworks guys: what changes are you making to the protocol file for Logworks 3 to log 2 -byte load ?
milburn
Last edited by wrcwannabe; Jun 10, 2007 at 07:47 AM.
Jun 11, 2007, 12:52 PM
#11
So on a third gear pull, it would spike the boost and overshoot the target load, then drop too far down and go below the target load, go back up and overshoot the target again, and then maybe hit the correct load. It explains why my dyno graph has these funky hills in it.
Jun 11, 2007, 02:40 PM
#12
Boost definitely will hold longer in a higher gear though as your comment suggests.
spdracenut mentioned reducing the correction factor. This has been my solution too. I run with zero upward correction.
Jun 11, 2007, 07:20 PM
#13
I greatly reduced the correction factor values because I found that the stock system was constantly 'hunting' and adjusting the boost to try to hit the target desired engine load. Even when the car was completely stock, the boost would cycle 2-3 times on a third gear pull before it settled down and 'hit' the target engine load. Bascially, it was a underdamped system and it was overcompenstating cause it to overshoot the target load.
So on a third gear pull, it would spike the boost and overshoot the target load, then drop too far down and go below the target load, go back up and overshoot the target again, and then maybe hit the correct load. It explains why my dyno graph has these funky hills in it.
So on a third gear pull, it would spike the boost and overshoot the target load, then drop too far down and go below the target load, go back up and overshoot the target again, and then maybe hit the correct load. It explains why my dyno graph has these funky hills in it.
Have you played with the WGA preload? There is a fine line there where you get very little boost creep and less taper. With your TBEC settings, that should make it rock solid. You might try making it shorter, 1/2 turn at a time to see.
Milburn
Jun 11, 2007, 07:32 PM
#14
I could actually hit peak boost at 3100 rpm (or perhaps sooner) in 4th if I wanted. I'd just have to increase the BWGDC from 36 to 40-43 at 3100 rpm. I choose not to do it though as I'm a bit worried about 25 psi at 3100 rpm.
Boost definitely will hold longer in a higher gear though as your comment suggests.
spdracenut mentioned reducing the correction factor. This has been my solution too. I run with zero upward correction.
Boost definitely will hold longer in a higher gear though as your comment suggests.
spdracenut mentioned reducing the correction factor. This has been my solution too. I run with zero upward correction.
Dang I got lots of things to try this weekend now. I think I'll revisit the TBEC tables and see if I can get the boost curve moved lower, but still keep the 1 psi taper. I'll post any data if I can get it done.
mr fred have you changed the protocol.xml in Logworks ? If so can you PM about it?
Thanks for posting guys !!!
milburn