Initial Details into How The Boost Based Tuning Works with AccessTUNER

[Christian.]

Here is some initial information about the boost-based tuning that I've been able to put together for ya'all. I will update this as time allows. Please let me know this can be made more clear or more easily understood.

Use Boost Based Dynamics (Boost Control)

When this box is checked, the ECU will function using several Boost Tables and the MAP sensor readings in order to control boost. This should provide what you are calling "PSI tuning" vs. the “Load Tuning” the factory implements. In other words, the ECU will take the result of the Boost Target table and compare it against the actual Boost measured by the MAP sensor. If Actual Boost is greater than target boost, it will reduce the WGDC (attempt to lower actual boost). The opposite is true if Actual Boost is less than Target; the ECU will then use the authority given to it within the Boost Dynamics table to increase WGDC.

The ECU will use the reported “Throttle Position” in the datalogs for looking up the Boost Target values. Since the ECU reports ~74 for WOT, the ECU appears to not be using the 4 final rows from 75-100. This loss in resolution should not hinder your ability to properly calibrate Boost Targets. Without including the last four rows, this ECU still has 15 x-axis RPM break points and 13 y-axis TPS break points for the psi-based Boost Targets table. This is more than sufficient considering that the GTR has a Boost Targets table that is 8x8 in resolution.

Depending on how your boost control system is mechanically set up, you may need to modify other tables within the Boost Tables folder in order to allow for appropriate boost control. We will go over this in greater detail below.

The following screen shots and logic descriptions explain how the closed-loop boost control system functions using this custom coding.


1st - The ECM logic will “look-up” what the Boost Targets have been calibrated to based on RPM and Throttle Position look-up values; achieving these Boost Targets is the ECM's primary goal for closed-loop boost control system.


2nd - The ECM logic will then “look-up” what the WG Duty Cycles have been calibrated to based on RPM and Throttle Position look-up values; the ECU will then drive the boost control solenoid in order to achieve the desired Boost Targets.

3rd - At very fast rates, the ECM take readings from the MAP sensor and measures the Delta Δ (or difference between) the desired Boost Targets and the actual measured Boost for the measured RPM and Throttle Position.


4th - The ECM logic will then “look-up” compensatory Boost Dynamics values that will modify the WGDC in order to achieve the Boost Targets for the corresponding RPM and Throttle Position.

The ECU logic will then cycle over again in this “closed-loop” operation.

The values in the Boost Dynamics table give the ECU the authority to modify the WGDC during over boost and under boost conditions. The values on the right hand side of this table give the ECU the authority to reduce WGDC during over boost conditions. The values on the left hand side of this table give the ECU the authority to increase the WGDC during under boost conditions. If you are getting significant boost oscillations, then you may need to fine tune the values in this table or you may need to recalibrate the WG Duty Cycles table. Generally speaking, it is easier to start with less WGDC than you need in order see how the turbo responds.

One easy way to help figure out proper WGDC values is to set the WG Duty Cycles tables to 25% and then run the car at WOT with the desired Boost Targets set. The ECU will then try to do everything it can within its authority (Boost Dynamics table settings) to figure out the proper WGDC values to run in order to hit the Boost Targets. Simply observe those values and begin to enter them into your WG Duty Cycles table until you see the ECU is controlling boost very consistently.

I've also attached a base calibration that you are welcome to start with.

2007-2008 MS3 = Stage1+SF 93 v107 FMIC BT vF.ptm
2009 MS3 = Stage1+SF 93 v107 FMIC BT vA.ptm
2006-2007 MS6 = Stage1+SF 93 v105 BT vA.ptm

XOXO,
Christian.

[Cataphract_40]

Originally Posted by Christian. This is more than sufficient considering that the GTR has a Boost Targets table that is 8x8 in resolution.

Nice! Always wanted a GT-R.

on a more serious note, I just wanted to make sure that this stuff is correct for our car's ATR and not just the GT-R.

As always...thanks for the info Christian, it will come in very handy

[speed23]

Thanks for helping us get started Christian. I dont mean to push our luck, but any chance that the ecu could do a final check of the boost compensation tables to determine boost, so that we could limit boost in 1st and 2nd for the ms3 guys?

[Lex]

It seems that the first and second steps should be reversed.

It makes sense that the ECU starts by command a certain WG duty cycle since it controls that directly.

Then, it uses the load dynamics table to achieve the target in the boost targets table.

[Christian.]

Originally Posted by speed23 Thanks for helping us get started Christian. I dont mean to push our luck, but any chance that the ecu could do a final check of the boost compensation tables to determine boost, so that we could limit boost in 1st and 2nd for the ms3 guys?

I have not checked if the existing boost comp. tables per gear are still functional.

Originally Posted by Lex It seems that the first and second steps should be reversed. It makes sense that the ECU starts by command a certain WG duty cycle since it controls that directly. Then, it uses the load dynamics table to achieve the target in the boost targets table.

I debated that, but I wanted to explain things in an easy to understand manner so I thought stating boost before WGDC was best. It would not matter if WGDC or Boost was first...this system is pretty damn fast with boost control and it operates in a closed-loop manner with MAP delta feedback.

Christian.

[m4tic]

how does this boost based switch affect ignition timing? since the ecu doesn't use full set timing if not necessary with achieved load. (or is that even correct?)

[Lex]

Originally Posted by Christian. I debated that, but I wanted to explain things in an easy to understand manner so I thought stating boost before WGDC was best. It would not matter if WGDC or Boost was first...this system is pretty damn fast with boost control and it operates in a closed-loop manner with MAP delta feedback. Christian.

If you've played around with this, how much do initial conditions matter? In other words if I start with 100% in the WGDC table, how much does the boost overshoot for example?

Also, say there is a MAP sensor failure, does the ECU revert to using the WGDC table only?

[djuosnteisn]

Originally Posted by bmorrisj how does this boost based switch affect ignition timing? since the ecu doesn't use full set timing if not necessary with achieved load. (or is that even correct?)

With constant boost, you'll have varying load. For instance 15psi in 100 degree weather would result in much lower load (air flow) than in 30 degree weather (and 100 degree weather would be quite a bit slower). The ign advance will be calculated as normally for that given rpm and load point.

[m4tic]

would the tune need to be modified occasionally for temperature swings while using the boost target switch?

[Jarods7920]

Would this table function correctly when using an external gate and a grimspeed solenoid given the correct adjustments are made?

Also with this checked, do we still have the ability to do boost by gear?

[manelscout4life]

Originally Posted by bmorrisj would the tune need to be modified occasionally for temperature swings while using the boost target switch?

I asked Christian about this and he told me the car should target the same boost targets even with changing weather conditions

[evidence]

Originally Posted by manelscout4life I asked Christian about this and he told me the car should target the same boost targets even with changing weather conditions

Point is though (this is the reason many people prefer the stock load system), if your running 20psi in the summer and pushing the limits, you might not want to be running this same boost in the winter. It would result in much higher loads and stress on the engine.

I suppose the only reason to run a boost setup like this is if you can't get your per gear set up properly, or don't want to bother with it? Configuring with the boost setup seems simpler. Losing unique configs for gears 1 and 2 make this even more a deal breaker for me also.

[Haltech]

Damn, i wont be able to play with this until thur, but i def want to modify one of my maps based off psi value. What should the load cap tables be set, to, 5.0 still?

[fjames]

Originally Posted by evidence Point is though (this is the reason many people prefer the stock load system), if your running 20psi in the summer and pushing the limits, you might not want to be running this same boost in the winter. It would result in much higher loads and stress on the engine. I suppose the only reason to run a boost setup like this is if you can't get your per gear set up properly, or don't want to bother with it? Configuring with the boost setup seems simpler. Losing unique configs for gears 1 and 2 make this even more a deal breaker for me also.

The same is true in reverse for load tuning though. If you tune for max, safe load in cold weather, you'll be in for some pain trying to push enough boost to meet that load target in the summer.

I'm trying the boost toggle today and loving it (ambient 28F.) I have enough experience to know what boost my car can run year round so no problems there. With my max boost at max summer setting, I get 20x load reported, so no problems there.

The huge thing for me is it's much, much smoother to drive. Christian or Trey mentioned in passing that this might happen, but didn't offer an explanation. On a pair of 4-5 shifts I got no, and just a smidge of spike, so that's interesting. Just preliminary, but so far it's looking very promising for a DD.

[manelscout4life]

Originally Posted by fjames The same is true in reverse for load tuning though. If you tune for max, safe load in cold weather, you'll be in for some pain trying to push enough boost to meet that load target in the summer. I'm trying the boost toggle today and loving it (ambient 28F.) I have enough experience to know what boost my car can run year round so no problems there. With my max boost at max summer setting, I get 20x load reported, so no problems there. The huge thing for me is it's much, much smoother to drive. Christian or Trey mentioned in passing that this might happen, but didn't offer an explanation. On a pair of 4-5 shifts I got no, and just a smidge of spike, so that's interesting. Just preliminary, but so far it's looking very promising for a DD.

On the map Christian worked on is hitting 18psi with an ambient of around 34 degrees, my calculated load is showing as 2.29 which I'm guess is 229. Is this too much load?

Heres my log:
http://www.mazdaspeedforum.org/forum/foru...-fmic-bt-e.csv

[fjames]

You can't really compare loads car to car because the MAF cal will be different. I just go by what the 2871 guys see, and what fully bolted guys see, compared to me. My conclusion is anything that starts with "2" is probably okay if boost is in line. In other words, if you were running 22 psi to get that load, you should be worried, but at 18 psi, you're in the safe zone for performance tuning (which is inherently unsafe.)

Your AFR is pretty erratic and you know your pressure is fucked. I wouldn't be doing anymore logging or aggressive driving until you get that fixed.

[manelscout4life]

Yea I know I'm waiting to get a fuel pump the tune is done for now

[Christian.]

Originally Posted by bmorrisj how does this boost based switch affect ignition timing? since the ecu doesn't use full set timing if not necessary with achieved load. (or is that even correct?)

This does not affect how the ECU look-ups and calculates ignition advance values. The ECU will continue to use RPM and Calculated Load for the main ignition advance look-up functions.

Originally Posted by Lex If you've played around with this, how much do initial conditions matter? In other words if I start with 100% in the WGDC table, how much does the boost overshoot for example? Also, say there is a MAP sensor failure, does the ECU revert to using the WGDC table only?

I have properly calibrated boost controls thus far. I've not intentionally thrown off WGDC in order to see what happens. My guess is that boost spikes and oscillations will be introduced if WGDC is set to high.

I have not created a failure scenario for the MAP sensor in order to see how the ECU will respond...although, I do believe that you should be able to zero out the Boost Dynamics table and then the ECU should operate with a open-loop boost control system and just run what is in the tables.

Originally Posted by bmorrisj would the tune need to be modified occasionally for temperature swings while using the boost target switch?

I've uploaded a base calibration that has reasonable WG Duty - IAT Comp. values to help protect for super cold over boost conditions and to lower WGDC with higher IAT values.

Originally Posted by Jarods7920 Would this table function correctly when using an external gate and a grimspeed solenoid given the correct adjustments are made? Also with this checked, do we still have the ability to do boost by gear?

Boost control with any configuration will work just fine as long as it is mechanically set up properly. I have not tested the boost by gear tables yet, so I don't have an answer for ya there.

Originally Posted by Haltech Damn, i wont be able to play with this until thur, but i def want to modify one of my maps based off psi value. What should the load cap tables be set, to, 5.0 still?

Yes, I would still set the Calc. Load Max. A-B to 5...or whatever value you need to in order to not run into that fueling limitation imposed by the ECU.

Originally Posted by fjames The huge thing for me is it's much, much smoother to drive. Christian or Trey mentioned in passing that this might happen, but didn't offer an explanation. On a pair of 4-5 shifts I got no, and just a smidge of spike, so that's interesting. Just preliminary, but so far it's looking very promising for a DD.

I may have some suggested WGDC setting to help avoid spikes after shifts. Please allow me some more time to experiment or you can IM me with some questions and I will go over what I am thinking with you there.

Originally Posted by fjames Your AFR is pretty erratic and you know your pressure is fucked. I wouldn't be doing anymore logging or aggressive driving until you get that fixed.

Agreed, please don't drive aggressively until your fueling issues are addressed. manelscout4life, I greatly appreciate you allowing me to use your vehicle to verify some things. Tuning the vehicle using a boost reference was pretty easy IMO and I will continue to update our support documentation as I learn more.

Christian.

[Ziggo]

Originally Posted by Christian. I've uploaded a base calibration that has reasonable WG Duty - IAT Comp. values to help protect for super cold over boost conditions and to lower WGDC with higher IAT values.

But the ecu is still going to target boost right? So load (and thus rod bending torque) is still going to vary with temperature.

I guess I just want a bit of caution thrown in here. When we first got started on this platform people "tuned" by installing MBCs and map clamps (ie. mechanical boost targeting) and I remember well all the blown engine threads when winter and cold temperatures rolled around.

[djuosnteisn]

I think a fair amount of those blown motors resulted from load cap fueling issues. Colder weather = higher calc loads.

Higher calc loads + calc load cap = lean afr's and propensity for deto...

[Ziggo]

Originally Posted by djuosnteisn I think a fair amount of those blown motors resulted from load cap fueling issues. Colder weather = higher calc loads. Higher calc loads + calc load cap = lean afr's and propensity for deto...

I think a fair amount were also from fuel pressure dropping due to stock pumps, but the common variable here is varying load. If load were constant neither of those things could have happened. Next up, heads needing to be replaced from blown stock injector seals eating the bore when temperatures fall and load goes up?

Varying load can introduce a thousand problems, most of them we probably have not discovered yet, thats probably why OEMs switched to using load.

[themytb]

just subbin yo

[FORZDA 1]

[quote]

Originally Posted by Christian. ..... Since the ECU reports ~74 for WOT, the ECU appears to not be using the 4 final rows from 75-100. .....

Are you SURE about this? My boost results prior to this update "seem" to indicate otherwise. Maybe it was just a fluke based on my Boost targets and the OEM Load Dynamics settings?

I've also attached a base calibration that you are welcome to start with.....

Must be a "3" map. I can't open it.....

[Christian.]

Originally Posted by FORZDA[quote Are you SURE about this? My boost results prior to this update "seem" to indicate otherwise. Maybe it was just a fluke based on my Boost targets and the OEM Load Dynamics settings? Must be a "3" map. I can't open it.....

Yes, I am pretty confident.

I've also updated additional base calibrations for the different MS platforms. I apologize about only showing love to the 2007-2008 MS3 owners

Originally Posted by Ziggo But the ecu is still going to target boost right? So load (and thus rod bending torque) is still going to vary with temperature.

The calculated load generated by tuning the system according to boost is going to occur, but that is also how almost every other OEM MFG handles ECM calibrations so I don't think it will be that big of an issue. Some other table changes may need to be made (as I have done with the WG Duty - IAT Comp. table settings), in order to allow for consistent performance. We will be sure to address that or to make suggestions once the data shows us to.

Christian.

[fjames]

Originally Posted by Christian. Yes, I am pretty confident. I've also updated additional base calibrations for the different MS platforms. I apologize about only showing love to the 2007-2008 MS3 owners Christian.

If you're saying that any of the 3D tables that use throttle position are subject to this, then that's not my experience.

If it were true all the time, then I wouldn't be able to run over a 1.6 load as this is the value at the 75% row in my TRL A&C tables (gear values are 1.84.) So, hypothesizing here, it may be a technically correct "fact" in isolation, but when added to a system, the result will vary according to the many other variables involved? Most particularly I'd guess would be the DBW table, where most have max throttle at say a 1.5 load or even lower.

I tested again today a bit (not logging, just eye-balling) and no to little spike. It's almost thrilling to see the boost simply climb and hold at maybe +.25 psi from target on a shift - this in 19F ambient. Looks like you did what I figured you'd do on the IAT table - I'm going to try that to see what happens, but I run Boost Dynamics at a 30% reduction both sides, and have max values at higher rpm in WGDC at maybe 40, so it may not apply to my setup.

I have a feeling some of these options for tuning boost response may be redundant - different path to the same result. Trying to figure now what the "best" path might be, or if it really matters.

Again and still, just loving the boost toggle.

[Haltech]

This thread is now stickied and should be a good resource for future reference.

[Christian.]

Originally Posted by fjames If you're saying that any of the 3D tables that use throttle position are subject to this, then that's not my experience.

No, just the pressure based boost calibration tables from what the data has shown me.

Christian.

[Jarods7920]

Fjames, have you checked to see if the boost by gear is still functioning?

[fjames]

Originally Posted by Jarods7920 Fjames, have you checked to see if the boost by gear is still functioning?

No, but it better not be lol. I did knock down boost comp 30% for 1st and 2nd, but haven't checked to see if it did anything. Both tuning methods have equal issues fundamentally, but the difference is the boost toggle seems to toggle OFF the factory load stuff - the stuff that hasn't been exposed in ATR I mean. All that perkiness they coded in is gone with the toggle.

I wouldn't want to start tuning with boost though - the reason it's easy for me is I'm so familiar with load tuning on this car. I can verify boost levels against load in previous logs. In other words, it would be pretty stupid to run more boost than you need at high rpm, and you won't know those limits until you fine tune load first. I think that makes sense Now thanks to Cobb, you can do both.

[FORZDA 1]

Originally Posted by fjames ...... run more boost than you need at high rpm, .....

WTF? Is there such a thing? LOL

At high rpms WOT, I'm thinking that boost is ALL that matters, until the head pops off, or the rods exit. The Load is a calculation of efficiency which does not matter when making max power. Heck, you can keep jacking up the Load targets until the motor blows and you'll not know what manifold pressure it took to blow unless you happen to be logging at the time.

IMO, Load tuning is great for emissions and if you want to consistently limit the power output, but "big power" is made with manifold pressure to get the air into and through the motor.

[fjames]

With your beaner3, sure, within reason, but with the stocker (I should have qualified my comment to apply to the stock turbo) there's so much heat it doesn't matter. This car likes to make power (by power, I mean hp and tq together) between 3.5 and 5K - tuning above that is in the "fine" tune area for sure to me. Of course, this is for a DD perspective. Most DD people aren't interested in using holes in the block as a tune indicator You could run meth, and just blow the turbo instead of the whole motor ... my comments are sans meth.

I prefer load tuning because I like using the ecu's brain power to manage boost. Trouble is, that brain has been corrupted by the factory in ways I don't like, that are still unexposed to us. So the boost toggle to me isn't "boost tuning" it's the "anti-factory load tune."

I don't know what you mean by limiting power. We're just moving air through a cylinder here, how you control that air is just a detail. If one guy tunes for 18psi boost at 6K and another guy tunes with load to give the same boost at same rpm, they're making the same power at 6K. If they did this in dead of winter, come summer the load tuned guy would be making more boost. Reverse the seasons and the result is reversed. That's a good thing to me, but the bugger is in the details. like always, and the factory tuning doesn't get those details right for me.

[FORZDA 1]

Originally Posted by fjames .....I don't know what you mean by limiting power. ..

No problem, just pokin' ya.

As for the Load based tuning, it certainly gives the tuner (OEMs especially) much finer control over the power output than the much simpler (but admittedly coarser) boost based tuning. In the old(er) days an EBC was the hot setup rather than fiddling with WG springs/preload.

[Dano]

Couldn't you blend boost tuning with the IAT comp tables to get the best of both worlds as far as keeping load in check in cold weather? ya know applying Christian's approach to your maps. I think this would be a great approach if it works as intended.

Of course you still lose gear based tuning unless the Boost comp tables do anything and I do like my gear based load tuning but think I would like gear based boost tuning better if we could get that to work. I think boost tuning would be both easier for the tuner and ECU. Trying to input load targets that match the natural rise in load through the RPM range is much harder to do than inputing target PSI. This naturally induces boost occiliations if you don't match the commanded load rise with RPM and PSI.

so many new things to play with

[fjames]

Originally Posted by Dano2010 Couldn't you blend boost tuning with the IAT comp tables to get the best of both worlds as far as keeping load in check in cold weather? ya know applying Christian's approach to your maps. I think this would be a great approach if it works as intended. Of course you still lose gear based tuning unless the Boost comp tables do anything and I do like my gear based load tuning but think I would like gear based boost tuning better if we could get that to work. I think boost tuning would be both easier for the tuner and ECU. Trying to input load targets that match the natural rise in load through the RPM range is much harder to do than inputing target PSI. This naturally induces boost occiliations if you don't match the commanded load rise with RPM and PSI. so many new things to play with

Yeah, that's a thought on the IAT table. Except I don't want to keep load in check in cold weather, I just mentioned it as something to think about for others.

I never thought it was trouble matching load through the rpm range, unless months of tedious, time consuming logging is trouble I never went for max load in the midrange though. I just kept increasing until the WGDC got out of line, or the increased boost produced diminishing results. I still think the culprit is still to be revealed to us, but working with what we've got, the boost toggle seems the way to go. I'm real curious to log, and should probably lay off hammering it until I do. Next week maybe.

Speaking of new stuff, I'm surprised you guys haven't tried setting the "BATvECT Comp % used" ignition table to 1.0. My car has never met mapped spark really, and now we know why - it knocks it down first thing according to the new help file. I set mine to .89 in the 6K column, because it's obvious that table is in tune with the stock ignition table where the 5.5K column is king - i.e. artificially high compared to others. If I was into spark tuning, that would be the first thing I did, so you're starting with half a clean slate with your map spark values (the other half is still dirty, with the "subtractive" table that follows.)

[fjames]

Originally Posted by FORZDA 1 No problem, just pokin' ya.

I'm honored to be worthy of a poke Would have appreciated some lube though.

[Dano]

on IGN

so you are saying you took the .70 value out of 6K and input .89...I was just working on my IGN advance and noticed in my logs that once I hit 6K I don't hit IGN targets anymore...ureka!

And a value of 1.0 effectively negates that whole table just like we do with ABS load tables setting to 2.50?

[fjames]

Originally Posted by Dano2010 on IGN so you are saying you took the .70 value out of 6K and input .89...I was just working on my IGN advance and noticed in my logs that once I hit 6K I don't hit IGN targets anymore...ureka!

Yes, that's what I did - haven't verified it yet.

Originally Posted by Dano2010 And a value of 1.0 effectively negates that whole table just like we do with ABS load tables setting to 2.50?

That's my hypothesis - you be the guinea pig

[Dano]

hahah

I have too much work to do with boost/load targeting to mess around too much with timing...there are plenty of other guys just itching to give it a go.

[08_ms3_gt]

Originally Posted by djuosnteisn I think a fair amount of those blown motors resulted from load cap fueling issues. Colder weather = higher calc loads. Higher calc loads + calc load cap = lean afr's and propensity for deto...

i've been experiencing this. first cold weather with fuel pump and downpipe --> running WOT gave me AFR's around 12.2 when commanded AFR is 11.5.

i may work on fracking around with it (Dano, you've done this?), but for now i flashed a 1.06 map in the hope that it'll keep my AFR's healthy.

[FORZDA 1]

Originally Posted by 08_ms3_gt i've been experiencing this. first cold weather with fuel pump and downpipe --> running WOT gave me AFR's around 12.2 when commanded AFR is 11.5.....

Sounds like your MAF cal is off just a bit.