Discussion regarding power limitations of the MS3
 

This discussion was started in the CPE forum following their announcement that they have found a way to extract more power from the DISI motor via reflashing the ECU and tuning using the SB.

I want to start a discussion about what the possible limitations of the DISI are and see if we can collectively determine what it is that limits power.

Before going any further, DJ and I have done some investigation of the injector PW. We have seen that at high RPM, as the injection window narrows, the car injects fuel all the way up to the spark event. A lot of this fuel is not atomized properly and doesn't effectively contribute to combustion. Read the discussion here:

http://www.mazdaspeedforums.org/foru...jectors-39696/

In order to deliver more fuel sooner and allow it more time to atomize, we found increasing pressure to be a solution:

http://www.mazdaspeedforums.org/foru...-sensor-42364/

CPE claim they have gained as much as 60whp under the curve by simply increasing fuel pressure but they say their discovery has nothing to do with fuel.

In reality, there isn't much to an engine. You feed it air, fuel, and spark and give it enough time to mix and you will make power.

We know fuel pressure increases improve combustion and we know the Cobalts were up against a wall until they upped their pressure.

IMO this is the missing ingredient but CPE claims it isn't the only one and fuel is not an issue. So be it. What else could the ECU be doing to limit power? If it's not fuel, we're left with it limiting air and spark. Even though we can't see the logic inside the ECU we can see the effects it has.

Air can be restricted through closing the TB, or adjusting VVT.
Spark can be seriously retarded.

These are all parameters we can log. Has anyone making upper 300hp monitored these parameters?

Please chime in with your observations. Below is a copy of the discussion removed from the CPE thread.

EDIT: I should add, there are 2 parameters regarding fuel that are important and we don't have in ATR - injection timing and fuel pressure.

I believe these are quite important but then again CPE claim this is far from what is needed. Timing is important because we can perhaps start injecting a little earlier and giving the fuel more time to mix.

well is there something we can rule out as number one? we need a BT car to test each variable on *cough* dustin *cough* and improve on it until we have all 3 working correctly together.

i know that with ATR we cannot adjust timing all the way. ATP contains tables that ATR does not and these tables are pertinent to seeing actual substantial changes in values. for ex there is a total timing table in ATP that basically determines the MAX timing the ECU will run. i think this works on the same concept as the TRL baro and MAX A/B tables where these tables need to be HIGHER then the individual gear TRL tables in order for the gear tables to work. ie the ecu takes the lowest of these 3 values. can we get a protuners opinion in here to see whats diff in ATR vs ATP?

Well, my argument has always been this car has been starved for fuel. Tweaking the solenoid unlocks more power in these cars with larger turbos and manifolds. If you look back on diesel tuning from 2003, they too hit an essential wall until fueling tables became unlocked.

However, from my own inquiries since owning this car, the DISI ECU is the most sophisticated ECU logic ever to come to America in a damn car. Ive often wondered why they chose to go with such an overkill system for this platform. There just may be a few other things that we have all overlooked. Who knows, eventually its going to be uncovered by someone else, its just a matter of time.

At the beginning of 2009, i was going to start playing with mixing and matching solenoid control systems for these pumps to scale the fuel pressure. Even after you maximize your pressure, you will eventually hit another wall somewhere else, but we had to start somewhere. Fuel internals werent the answer to unlocking the power in this car.

The only thing i cant understand about what Cpe has going on is why they have to wait for just a certain car. We know there is a large number of cars running their parts and piggy's on this forum who live local to them. So, is it a combination of larger fuel pumps, modified HPFP internals and a flash? Im just curious why any other car in their area cant show the results of their discovery if its such a simple and obvious tweak. If i had a potential money making discovery, i wouldnt be down playing it one bit. I would have the results so i can sell the shit out of it.

So, the only observation i can make to this is that they erased a checksum somewhere that is a failsafe in the ECU. Is the car going into limp mode when certain voltage values are maximized? Removing this failsafe allows this car to put the power down? I dont know, its almost a discussion im tired of going over each year at this point.

Every datalog I have seen that included timing always showed something reasonable. I have not seen any drop-outs.

Is there an absolute limit where load calculated will result in the TB closing? I haven't seen too many people log TB data after it was all figured out.

I also haven't seen much done or logged with VVT.

EDIT: Haltech, this is my question as well - is the car going into some sort of limp mode? But if it is, we would see it in our logs. There are only so many things the ECU can do to limit the power if it hits that value. So the question is - what is it limiting?

i did a buncha VVt testing and the car always targeted exactly what i put int he tables so it always worked but i obvi diddent see much of a diff at all. i think values closer to 20 near redline netted me about .1 more load then values near 0 at redline.

when i first started doing timing adjustments with christian on the beta maps he was having problems targeting what was in the table. this was wayyyyyyy before ATR came out so i guess they figured that out in the ATP but i have about 16-18 deg of timing at redline and i rarely ever see more then 12-14 on a datalog.

Interesting observation on the timing. I also always see around that amount - might have to do with the knock sensor going deaf beyond 5700RPM and the ECU protecting itself.

So there's no way to push timing beyond 12-14 with ATR? Timing plays an important role as RPMs climb.

Honestly, I thought the same thing. Although, when it first started, I offered great flexibility to them. I was, and still am, ready to drive up there when they are.

Unfortunately, the first scheduled weekend was ruined because of the snow. Then, Jason (Dadasracecar) Left for Vacation. He wont be back until the 4th of January. Other options as far as who to have tune it were discussed. But it seems when it comes to the standback, and the mazdaspeed3, Jason is the goto guy.

Chances are, even if I was local up there, it wouldnt have happend yet. They were close to cancelling our schedule the first time, before Jake and I suggested that we move forward.

Im just as antsy as anyone to find out what the deal is, and have it done to my car.

Also, Lex,

If you give me a set of parameters, I will be more than happy to log them for you, although, my maf housing is cutting it close, so you may see a cut or two ;)

-Sean

I have talked to an engine builder, he deals with the 3 valve 4.6 ford motors with a vvt actuator, and he is a firm believer in locking the camshaft in one position once power is increased (ie turbocharging). I do know that a lot of the Honda's do away with the Vtech once the engine is boosted. I do not know if this is important on the MZR, but once I build an engine I will explore this a little more.

Great thread.

Not sure what else to say other than I think it's time we all get out our dataloggers lol.


On the o2 sensor subject. Don't they measure the presence of oxygen, not fuel (e.g. they measure the lack of consumed oxygen)? So wouldn't they be accurate even if there was liquid fuel flying by?

Not a BT car here...but if I could help by doing some logs, then sure thing! Just let me know the testing regiment...

After a lil more though on o2 sensors...

I think what i typed applies only to narrow bad sensors.

Cause with a wideband, if what i wrote was true, how would the sensor ever discern the difference between stoich and rich, cause in both scenarios (theoretically) all the air is consumed.

Maybe it is the o2 sensor... I'll have to study up on how they work later.

The O2 sensor measures the amount of O2 present in the exhaust stream relative to ambient O2 present.

What I was suggesting was that oxygen continues to be consumed late in the stroke as it continues to mix with fuel and burn, but that late contribution does little for power.


Thanks Sean, you're tuning with ATR right? Have you played with upping timing above 5500 RPM?

I would try upping timing to 16 degrees for the load you're running close to redline.

Then I would log timing advance, MAF, throttle opening, boost, AFR, injector PW, on a 4th gear run up to redline.

I am right now tuning with ATR. I havent even TOUCHED the timing. I will take a look at it.

People want horsepower. Horsepower is a rpm phenomenon. It's obvious from ATR that the stock tuning kills everything above 6K, so it's a reasonable assumption that there's something(s) that we can't see as well. This is why I question their claim that it won't help stock turbo guys - the raw numbers might not be impressive, but some small percentage increase between 5.5 and 6.5K would be nice on the stocker.

At least at low rpm, it's obvious the timing control in ATR doesn't do much. Just monitor it realtime and it's all over the place, several deg higher than max in ATR. Has anyone ever logged say, 16 deg at 6K WOT? I never see more than 12ish, but my timing table is very conservative.

I use VVT at higher rpm to generate the same logged load with less boost and WG activity.

i will say that during my last dyno i did a couple logs and for some insane reason i saw 20 deg of timing on one of my runs near redline. lemmie see if i can dig that up.

On high hp engines, I think the ignition timing is directly tied to the injector closing since you get almost nothing in return for fuel injected after the flame front has been started.

I honestly think that power on DI engines is down low, the engine just has to be built strong enough to run it and the CDFP has to flow an ungodly amout of additional fuel... like double the fuel.

I am not in the 300HP+ crowd at the moment, but perhaps this data can help initiate some discussion...


Here is my dyno and the datalogs:

http://i77.photobucket.com/albums/j6...3DynoPlots.jpg


Pull 002 datalog:

http://i77.photobucket.com/albums/j6...o/100909-2.jpg

Pull 003 datalog:

http://i77.photobucket.com/albums/j6...o/100909-3.jpg

My timing is flat around 5-6° from 4000-5500 rpms and then increases to 14° at redline.



Here is my Ign map:

http://i77.photobucket.com/albums/j6...eOLNoKnock.jpg

This Ign map is based off the Stage2+93 v103 map. I pulled some timing around 0.3-0.5 load trying to remove the peak in the curve around the point where a lot of random PT knock seemed to occur. All the other differences from the stock map are from Cobb.


Here is the result of subtracting the stock map values from my map:
http://i77.photobucket.com/albums/j6...gesvsstock.jpg

You can see where Cobb pulled/added timing (neg = timing pulled, positive = added).

I wish someone here knew how they came to make these changes.

I am temped to add back some of that timing in the 4500-5500 RPM range as it seems reasonable that the power gain I have from 5500-6000 is due to the timing climbing.


One other concern of mine, is that my TRL values are currently targeting 1.95 load. However, when I add my FMIC in the next few days I imagine I will be targeting over 2.00. How does the ECU determine timing values for loads above 2.00?

What load values are people with BT and upper 300HP targeting?

i dont even know the targets in my map. im using a boost controller, so i dont need to mess with that.

for loads above 2.0 christian said you basically just stick to the 2.0 line. its a really crude way to go about it but thats what he said.

A lot of boosted Honda's do go this route BUT....there are still Vtek s2k's running 700hp so IMO the VVT will only help with small gains and has nothing to do with the ever illusive "Wall"

How 'bout this effort?

Race Roots 428whp 361tq - Mazda Forums

ATOMIZATION
 

I THINK THAT BEFORE WE ASSUME THAT PROPER ATOMIZATION IS NOT HAPPENING , WE HAVE OBVIOUSLY ACCEEDED THE FACTORY BOOST SETTINGS AND ARE PUMPING MORE AND FUEL INTO THE MOTOR I THINK THAT THE PROBLEM IS THAT THE STOCK IGNITION DOES NOT HAVE ENOUGH JUICE AND THERE IS NOT ENOUGH TIME FOR PROPER FLAME TRAVEL TO OCCUR

I'm really torn between a couple of different theories though.

Something has to be limiting air flow electronically. When I drove around hitting sub 500psi at WOT I hit AFR targets DEAD ON but almost no air was coming in and I could barely hit 10 pounds of boost.

So IF AFR is the number one target looked at by the ECU then it would make sense that air flow would be the first variable changed by the ecu to get a proper mixture. This seems to make very simple...perfect sense.
Target AFR is the desired result, Fuel Pressure is a constant, and air flow is the variable that the ecu can change...

Ok well that's great and would explain why more fuel pressure would = more power
BUT
CP-e says Fuel Pressure is not whats causing this wall...fuck lol

Maybe...just maybe... the these same tables that limit air flow to save your engine under low fuel pressure situations will only let more airflow in untill say 1750psi of fuel pressure...

Then(since untill now we couldn't run more pressure) things start to make sense. if the ecu thought it would never have to account for pressures above that than why would it have any need to allow more than X amount of air to mix...This also explains why BT's and ported heads/intake mani's aren't making power??? This would also account for a DEFINITIVE "wall"(ECU won't allow airflow past point X)

IDK just more crazy talk but...hypothetically if CP-e found these tables...and could allow more air in to compliment more fuel pressure...it would... A: WORK!!! and B: make their claims that fuel pressure was not the "wall"...true.
So their statement could read something like "We found a way to stop the ecu limiting airflow...so THEN we raised the fuel pressure..."

Hmmmmm....

O yea and on top of all that timing, and pulse width and 250 other unknown tables could also play a huge role lol.

And Kudos to RR on 428!!!

Good points Realgib3!!

just wanted to say thanku to all the people putting the time and money to figure out this platform

I think cld12pk2go makes an excellent point. The COBB tables are limited to a load to 2.0.

So let's assume you have a large MAF housing and are able to flow much more than the stock MAF and the MAF tables is calibrated properly. The ECU will calculate loads above 2.0 ... and then what? All of a sudden you are outside of the "tunable" area. That means you can no longer change fueling or spark etc for the higher loads.

At this point, you won't be able to tune above a load of 2.0. Does the ECU also try to save the motor in another way if you're above that? Is it just dumping fuel, is it closing the throttle?

I want to see a log of a car running above 2.0 load into the high RPMs and I want to see Throttle, Boost, Inj PW, MAF, AFR, Timing. It would also be interesting to see what the variable tumble and swirl butterflies are doing in the manifold.

Additional fueling plays a large role. RR made 420+ hp by adding fuel. So fuel pressure will be huge for hp since the torque curve won't fall flat. Torque however is what represents load. Notice they were at the 368wtq point and no higher. That means they were probably riding the top of the load curve in the ECU.

I agree with this 100%. It's part of what I was getting at in my last post...the tables all stop somewhere soooo what happens after that???

And I really can't wait to see what kinda power RR can make if they run say...2100psi AND the secondary fueling....future is starting to look pretty bright :)

small typo but Max Load is 3.0, right?

Also, when you run a larger MAF housing and calibrate the MAF tables accordingly... you need a higher load value to hit a certain tq target. So running larger MAF is gonna be hurting you in this regard im pretty sure. So.. let's just run really small MAF housings, thats the secret to mazdaspeed powa!! jk

Most tables run to 2.0 load on the axis which is around the "wall" torque AFAIK.

Also, load is directly based off g/s and not MAF voltage. So a larger MAF will let you measure more g/s and that will result in a higher calculated load.

Oh ok I see what you're saying not. I slightly misread your previous post. We can go to loads above 2.0, but above that we can't really control a lot of what's going on.

As far as the MAF tables/load.. all I can tell you is that if you add percent to the MAF tables, you need to increased requested load a good bit to get back to the same power level that you were at before.

Dammit! Lex beat me to it... I was thinking about what the VTCV.. butterfly thingies are doing... That would be one way for the ECU to control air flow. It would essentially cut max air flow in half...

But think about it...Whoosh had no VTCS and a ported mani..so does SUMark. VTCS isn't the answer.

I am not running the vtcs system either....

back to my first post on this topic...

I thought about the intake mani too but like it has been proven, they still don't make power even with a Pnp head...
So airflow(among other things) is limited SOMEHOW by the ecu. I don't know how, but somehow it is. I think Fuel pressure, and Load are major contributing factors to this limitation.

Sounds like something along the line, the G/s isn't being read correctly by the MAF or ECU once it reaches a certain point...add all the freaking fuel and spark you want, you get nothing more than what that sensor can tell the ECU to make...

it may not be something physical...and more than likely, it's due to an inexpensively developed ECU logic, developed as an OEM pkg, without any expectations or much room for modifications beyond factory spec.

redefine the logic of what the ECU can read as incoming air, and sky's the limit...

but of course, you should also up the fuel pressure at this point...lol...

Just curious, Has anyone tried a standAlone system to run and tune our engine?...

My honest opinion is that there is no limp mode and that the ECU doesn't "limit" power in any way. Throttle body likely stays open, etc etc. A couple of logs at high boost on a BT should reveal this.

The big breakthrough is making fuel available earlier and in higher volume and tuning for it in spite of what CPE trying to make it sound like there's more to it.

With the AP, as super mentioned, once at 2.0 load you probably just ride that row for any higher numbers and although this is not ideal for tuning I doubt it puts the car in some sort of "limp mode"

Think about it like this: Air, Fuel, Spark -- how can the ecu limit each of these individually?

Air -- Throttle (done), VTCS (done),
Fuel -- HPFP (done), Intank (done), fuel pressure (in progress)
Spark -- ??

hmmm good point indeed

^^ Now we're thinking. Top post mentions all of these. This is why it's not magic. We would SEE if any of these parameters went into limp mode or did anything crazy regardless of what anyone claims.

OK OK so now I'm a little excited...

Ok so for arguments sake, lets say timing is the issue...
and for arguments sake...lets say CP-e has NOT unlocked injector timing...

So what can be changed...right now???

Just a thought, but if we now have more fuel(yay!), and the timing of the injection process is a constant...but we need to get fuel in earlier in the ignition process...
Can't we change the timing of said ignition???
Everyone has been trying to up the timing up top to combat the loss of power...but isn't that just making it HARDER to get more fuel in before ignition???

So if we RETARD the timing up top shouldn't that give us MORE time to inject fuel prior to ignition????????

You just have to think systematically and if you break everything down in sections. You remove the blocks one by one until there isn't any blocks left. The community and vendors have been doing this one by one for the last 3 years. Remember when we were all complaining about the damn throttle plate..lol

We have to be systematic.

EDIT: Well, in Jan. I get my fuel pump, and I'll have my GT35. If we don't figure it out by then, I'll get some logs at redline for you guys. I don't have an AP, and will either have CPE's flashs or Versa tune, but I'm down to help if I can.

FYI, I do not believe that we have any airflow limitations due to the ECU.

If we did, we would see the MAF g/s fall off. This obviously doesn't happen now that the throttle blade is held open.

Obviously, spark isn't the issue either.

So it is quite simply a fuel related phenomenon.

Based upon the Cobalt's recent break through, our path forward is quite apparent.

There might be a few new twists involved in rectifying our situation due to specifics in our ECU and fuel rail hardware, but it will be resolved in the near future.

Here's to 2010!

Just babbling some generalities, and yeah, I know you can't compare dyno numbers, but I'm going to anyway. Like I said, general observations ...

12pk and RR make the same peak tq.

12pk and RR hold peak tq for the same time (~1000 rpm)

12pk and RR are above 300 tq for the same time (~2000 rpm although if you squint, maybe 12pk holds a little longer)

12pk holds above 250 tq for significantly longer than RR - about 750 rpm longer.

12pk and me log the same loads but he makes way more tq presumably.

I understand that the RR solution to turning the speed into a drag car is a good one and I'm not criticizing them, I just think the comparison (similarities) to a stock turbo screaming for mercy is interesting.

I don't think the calculated loads are very accurate. I make over 2.0 loads when temps are below 30F, with lower MAF g/s than when I am at ~1.8 load when it is around 60F.

IIRC my recent logs showed about 4.5v on MAF compared to 4.6v ish when warmer and my WGDC was going to 0%.

Temperatures have an effect on calculated load as it should but there's more to it than just oxygen mass. I would like to know what formula is used to derive the load parameter.

My highly inaccurate butt-o-meter indicates that the higher load does not equate to faster go in the above two situations.

I would also like to know how it is calculated...

I read on a post on a bmw forum by a bosch guy.

He stated that load is not just air, but rather the amount of fueling needed for the air that entered.

So it may shed a lil light on your observations 12pk.


Here's the quote:

I have been thinking of the load as

Tl = Q/n

Which should pretty much represent torque production.

I am not sure I understand why an injector constant would need to be added...

I'm pretty sure the constant is just used to "scale" the equation. Basically adjust the slope of the line for a given injector characteristic.

im really interested in this too,

pre maf cal, i was hitting loads around 180 at 3k (21psi) tappering to 155 at 6k (17.x psi)

now post maf cal, im hitting 20X at 3k (18.X psi) tappering to 16X at 6k (14-16psi)

its really a huge difference and I feel its really high load for a stg1 mscai 105 (i also have tmic, bpv,aem)

Well hopefully there will be some enlightenment in the near future

How much did your afr's & LTFT's change from pre to post maf cal? Would be best if you could give numbers like LTFT went from +12 to -1 or something like that.

That sounds like cold weather. I gain about .12 in the cold here (20-30F vs. 75 when calibrated.) I target 1.84 and hit about 1.96, and that carries through most of the range. In normal weather I'm spot on.

where in the PID datalogging in the AP do you find Timing? I am interested in knowing how much timing my car is running at WOT.

I think they call it spark advance

Were they discussing a direct injection system? I'm pretty sure injector size (as opposed to fuel pressure) is less important on a DI car than on port injected cars.

If the system can handle significantly increased pressure, we should be able to get a lot more fuel before ignition without changing the injection start time.

ON OTHER DISI MTRS WHAT PSI ARE THEY RUNNING TO GET THE LETS SAY 500 HP. AND IS IT GOING TO BE REASONABLE TO GET OUR DI FUEL SYSTEM UP TO THIS PRESSURE AND FLOW RATE, OR BE CHEAPER TO AD THE AUXILIARY FUEL SYSTEM. BECAUSE IN MY SIMPLE MIND WE NEED THE FUEL SHOT IN BEFORE THE SQUEEZE ON THE FUEL. AND WHAT KIND OF INJECTOR IS THAT GOING TO TAKE IF THEY EVEN MAKE THEM JUST MY .02 I HOPE ITS PROGRAMING BUT I DOUBT IT

Our stock injectors should be able to inject a lot more fuel if we increase the pressure and stay with DI. If we add port injection we should be able to use just about any injectors we want since it will have to be a custom setup anyway - it shouldn't take very big injectors since we'll still have the stock ones too.

what is considered a safe range for "spark advance" or timing?

A good question, which I haven't found much info on.

I can vouch that my timing curve is safe for my mod combination. It is actually in all likelihood conservative.

I am not looking to trailblaze; however, and would like to know what others have done successfully prior to tweaking.

IMO it's best to tune spark advance on a dyno, where you can get immediate feed back on each incremental change, not just butt dyno. Safe range is determined by detonation and EGT's. Typically you want to shoot for minimum advance for maximum torque.

I'm all talk though, cause i've yet to actually tune my timing on a dyno, still waiting for my hardware to be finalized.

Also, on a dyno, as you advance the timing, you'll eventually see a point where torque starts to fall off with increased timing, and egt's will go up quite a bit. This means you've gone too far.

well, so far, I believe people have tried to advance timing but didn't really see any results. Of course things are changing around here pretty quickly so we will see.

I picked up 10 HP per degree in timing advance. With no detonation and EGT's were still well within check. The problem I ran into, was once we tried to go past 3 degrees, the car threw a multiple misfire code.

Can you post your initial/final Ign maps and your mods....

Tks

Please post it up, i would like to see where you uped it.

Now i know the ATR that we use is somewhat limited in how we can adjust timing. what timing changes does the professional ATR allow that we cannot?

the new 500hp 911 Turbo's with DI uses 2000PSI fuel pumps...

so do they have better flowing injectors cause you would think if the Cly is in the intake stroke i shouldn't take a lot of pressure to over come Cly pressure if the injector open's up to let it flow. but on DJ,s post on injector pulse with i just looked to me it would just stay open longer the higher the revs? i assume the pressure was still good so that would tell me the injector is not able to flow the fuel in the short amount of time just my .02 btw you guys kick ass :wavey:

I use the SB, so this will not help you guys that run the AP very much. The day of this tuning session the ECU was running around 10* advanced at WOT. This same tune is also running a solid 11.8 AFR and about 1300* on the EGT (probe is in the DP) I looked through the Dash hawk logs I did a few days after this tune, and the highest recorded knock at WOT was .7 for less than a few tenths of a second.

that's crazy. I was pulling so much timing before my build to get rid of knock

So 3° above the stock map everywhere above 3.8v on the MAF.

That would mean that you were running ~4-7° more than me in the 4500-5500 RPM range and about the same from 6000-7000 RPM.

Very interesting.

So you are saying you picked up ~30hp with the 3° increase? Got any dyno plots?

I believe that in order to avoid codes with the SB, you have to incrementally advance timing such that the ECU does not see a sudden jump in crank speed. In other words you have to ramp up the timing, only 3 deg difference from cell to cell or less.

I just have the final dyno, I am sorry I completely forgot to ask for the baseline graph, but the car pulled 290hp and 285lbft on the first pull of the day, all the charts at zero.

EDIT: I have already posted the dyno chart in another thread.......let me dig up the other thread and I will post a link



LINK: http://www.mazdaspeedforums.org/foru...eed-3-a-34435/
POST 16

Double merge post FTL!! As you can see from the chart, we took the timing up incrementally, moving one degree of timing per cell. Once we got to 3* is when the car thew the code.

Wonder if you were pushing the spark advance into the fueling event, and that caused the misfires.

No audible misfire, the car did not stumble or nothing funny happened, it just threw the CEL.

Wonder if the ecu would interpret a spark during fueling as a misfire, even if ignition was still successful. I mean even with a piggy back the ecu would still know if it were spraying during a spark event. Just an idea.

HKS twin spark...

The HKS will not, at this point, cure the random misfire code, since it really is not a misfire. I think (and correct me if I am wrong) the ECU interprets how many spark events happen in a cam rotation, and advancing the timing to a point where the spark event happens outside of that window throws the CEL.

Misfires are determined by examining the rate at which each tooth of the crank sensor is accelerated during the power stroke. The SB retards timing by delaying the crank pulse. The ECU senses this discrepancy and throws a misfire code. The SB might also be doing this a little crudely - not sure how fast their processor is.

Ahhh, it is the crank position, not the cam position that makes the determination. Thanks for clearing that up for me.

I've been thinking a lot about the wording CPE used when describing the wall.

First, they said that there is a definitive wall in the ECU which would suggest that the ECU is limiting HP in some way, shape, or form. Immeditately everyone jumped on the fuel pressure thing, but their response was that the fuel pressure increase only became necessary once the wall was found. In my mind that means that the ecu started to lean out on the upgraded pump and CPE had to increase fuel pressure to keep the AFR's in line. So let's for a second assume that CPE was being truthful in what they were saying (and I have no reason to believe they weren't). That would mean that fueling wasn't the issue and that the airflow is somehow being limited.

Now why do I say airflow...because the car should still make power even if the spark was off target right? So then I started to think has anyone ever actually confirmed visually or at the TPS sensor directly that the throttle is not closing. I'm thinking out side the box for second but, the traction control still controls the throttle in the event of slippage right? Maybe it also limits the HP on the top end some how. It some how overrides the throttle, but the ECU never reports the closure through some programming mistake. Maybe mazda put ecu logic in to close the throttle after 6k should it detect over a certain g/s at the MAF and not report it to deter ecu programmers. Out side the box right?

I really think the wall was just the fact that all the tables goto 200 load

Maybe, but someone would have thought about that by now other than CPE. That is the first thing we came up with, I can imagine that was also the first thing Christian would have thought of too.

Any thoughts on this? I know I'm probably way off, but even so I just wanted to get some feedback.

The way I see it is that timing advance may give us some gains in the short term, but isn't it making it harder for us to get fuel in before ignition?

Well I remember that Dada said he did some radical stuff with the timing on Robbies build, but never truly elaborated. We need to do 2 years of logging in a couple of weeks..lol.

IDK it just makes sense that if we can't make the fuel come in sooner, then we should just make the spark come in later...

Retarding timing will result in less power just because the flame front won't start until the piston is too far down the stroke.

The throttle closure is possible - sure. Using a voltmeter attached to the TPS on a car with a big turbo on a dyno would prove/disprove this very quickly. I unfortunately don't have such a car.

The really ODD thing is how CPE brings up the fact that they broke the "wall" yet can't even locate the dyno sheets of the cars that did it ... strange if you ask me. But this is not about them.

Keep the ideas and discussion coming, some good stuff is coming out of the woodwork.

Well the tables end at 7000rpms too, and raceroots hasnt had an issue with that, going to 7500rpms? I thought christian said the ecu just interpolated values outside of the tables?

like i said on the first page, christian said for timing above 2.0 load, you just use the 2.0 line and the ECU basically uses this. he said it was very crude but its the way they do it.

FWIW on my last dyno when i hit 385tq i saw loads of around 2.6 and was running about the same psi (26). that was just a peak number and i always usually see around 1.9 load by redline with about 22psi. my BT will be done by end of jan so ill be able to record whatever you guys want. my tune at akuma should also open up some new doors because john knows his shit when tuning and i hopefully will be able to see what he did.

I have the full control flash from CP-E, and I verified what the throttle plate does in 1st, 2nd, and 3rd (with the Dash Hawk).... but never paid any attention what the throttle plate was doing at WOT after I put the turbo on and had it on the dyno. Hmmmm interesting...........

My guess is that the wall is strictly tuning from what CPE has said, since we cant change injector timing and rail pressures require hardware, the only thing left is spark timing.

On DI engines, RPM is the enemy, so for them to have enough fuel for 400+whp they must be making the power in the lower rpms. If this engine makes ~415hp @ the crank @ ~7000rpm, then it can also make ~830hp @ the crank @ 3500rpm.

If it were tuning then we would have to go to them. Jake specifically said, it was in the flashes, but that the SB was only necessary to raise fuel pressure. I know some would disagree but I really do not believe the timing is playing role here. That is much too simple. It is another system that you would not normally think of that is limiting the airflow. TCS, DSC, EGR, PCV, VTCS, emissions, or something….it is one of these damn systems causing the problem I know. I just cannot see CPE coming out and exaggerating the 'wall' flash for it to be something as simple as timing or fuel. If it were that simple they would have just said, "Hey, here is what we did with the standback." This is something bigger; we just need to start looking into the extra systems and one by one eliminating them. Bring it back to basics and then see how much power is made. The more I think about this, the more confident I am that it is something extraneous creating this supposed wall.

Can you post the data logs for runs with the 3° advanced timing? I just want to see where it actually is at each RPM/load break point.

Tks,

It makes me wonder if they are flashing it to run 35-45* of timing and pulling fuel so that it isnt in the injection event. They have said that they were pulling fuel before.

I am pretty sure this is the correct file. I recently bought a new computer and I am trying to work between the new and old computer. It looks like the max torque is right on with the max timing advance.

It would appear that you definitely have ~5° more timing advance than I do in the 4500-5500RPM range.

After I install my FMIC and digest the changes that it requires in my tune, I will probably make my next map adding 1° in those ranges and monitoring for awhile.

How much boost are you running? We pulled 2 psi (ended up at 15psi) and added the timing to achieve the end result. It worked on my setup, give it a try and monitor the knock and see where you end up.

21 tapering to 17 PSI.

http://i77.photobucket.com/albums/j6...o/100909-3.jpg

hey i found this thread on M6C

the cobb ap - Mazda6 / Atenza

check out post #5, i dunno what to believe

WTF. I don't think I believe that, but even at 72% we shouldn't be as limited as we are.