we know there are different air fuel/timing maps for acceleration and deceleration so what you say there makes sense.

which is probably why leading throttle seems to work so well as we have discussed previously and as charles hill and old dragger have brought up today in a different thread.

 

*off topic*

under what conditions would the computer not open the throttle plate fully?
i don't mean in the low rev zone, but up top.

 

That's what I want to know? If more air is going through the MAF then it was programed to understand will it pull the throttle. Actually we know it will not because it hasn't on the turbo installs. But there it is adding fuel and I was not asking it to. Wait!!! That's not right is it? Those are blow through systems. Does the MAF read the added density? It's called a Mass flow sensor for a reason. But if that were true you would not need to do anything with the turbo system or mine. It would know how much air is going in and fuel it.

So it is not true. If it could work at alll it would be with my system. It is just getting more mass and should fuel it. It should just think the air got denser because you drove below sealevel. But it has a baro sensor so it will not think that. Maybe it will think the engine grew in size. Maybe it found another rotor sitting next to it.

Really why would it not just keep fueling it when more air is going in. Because there is no map?? Anyone know for sure how this thing works? I'm going to find out by trying it. I'll just go slowly into boost and watch the A/F. I know once it maxes out the stock injectors that it is over. But lets find out.

Since you know how the map uses the injectors maybe you have some insight as to how it wil react.

 

You know what I hate, when I stomp on the pedal and my Throttle plate doesn't open all the way. That is one of the reasons are times are so high at drag strips because we have this power but the computer won't let us have it when we want it. I think it would be great to have a system like you guys are talking about. I have a haltech e11x on my RX-7 and when I get comfortable with it I was thinking about doing something with the RX-8 s system. The computers in the RX-8 is our biggest hurdle to get around.

 

exactly... snoochie's car isn't going past 67% throttle. i've been trying to figure out why, but no luck yet.

 

Correct me if I'm wrong here, but the MAF is just a temperature-sensitive resistor, calibrated for the temp ranges changes that would be seen on the stock intake at all flow levels. By putting boost in front of it, you've increased the temp of of the air, so I'm not sure how this would affect things. But even if you assume that the ECU sees it simply as more air, it won't necessarily change things.

The ECU looks up fuel based on a matrix of RPM vs MAF values, but it also looks at the other sensors to make sure that everything is reading correctly. The ECU will pick a cell based on RPM and MAF and then will try to trim A/F based on the o2 sensor and remain in that cell. If the o2 trim drifts too far, it may switch cells even if MAF and RPM stay the same. The o2 trim doesn't apply during WOT tho.

So my take is that at high MAF values, the corresponding fuel value isn't all that high, and the range of MAF values runs out far before you'd get to full boost. Therefore the ECU gets stuck in one cell at the extreme of the fueling map, and probably eventually decides that the MAF is screwed up because its values are too high.

I dunno actually tho. That's just speculation. I would think what you would want would be a second MAF sensor that has a large range of calibration, or tap into the stock MAF and make sure responds thru the whole range of boost. Then you would want to map out the fueling values for the MAF-RPM matrix and determine where the ECU goes to during boost. Then a piggyback could be setup so that when MAF goes outside of the stock ECU's fueling map, the piggyback starts modifying the duty cycle based on RPM and MAF minus the stock ECU's max MAF value.

I would think that for some portion of boost, the stock ECU will actually work fine, but this value may be quite low given that the ECU was designed to only see atmospheric pressures. I'm not sure how much flow the NA engine generates at WOT, but I'd guess its probably less than the WOT flow at say 3 or 4 psi. I think this makes sense because if you were the ECU programmer, you'd want to use all of the possible resolution available in the ECU, so you'd try to establish max and min MAF values that the stock engine would see, and base the borders of your fueling map around those. With a partial throttle condition and significant boost, I'd think that the ECU would again get stuck on a border cell and would try to add fuel based on the o2 reading lean, but this trim range is small compared to the range of values for MAF vs RPM, so it will only have a negligible effect.

 

He's not going to run boost through it. The maf will be before the supercharger. THe ecu has already proven that when you make the engine breathe better, such as with porting, that it runs alot richer. It would be nice if the added airflow across the maf from the supercharger pulling in more air has the same effect.

 

Richer in terms of A/F or just in terms of adding more fuel? My point was that I would assume that there is some max MAF value afterwhich the ECU won't know what to do, regardless of if the injectors are max'ed or not. Maybe I'm wrong and it goes into some sort of linear mode when the MAF value goes beyond the fuel map.

 

It wouldn't. Partially throttle only has a benefit at RPMs when the port size exceeds optimal. My take would be that's only low rpms, although I suppose if the secondary ports opened faster than optimal, it might also occur there.

 

[QUOTE=Horse]You know what I hate, when I stomp on the pedal and my Throttle plate doesn't open all the way. That is one of the reasons are times are so high at drag strips because we have this power but the computer won't let us have it when we want it. QUOTE]

I don't *think* that the throttle by wire is actually decreasing your performance. The only thing I could think is a negative, is that the maps for the throttle openings are almost certainly based on the assumption that the clutch is out and the tires are gripping. Therefore, there's no reason for the engine to rev faster than these conditions, which makes a lot of sense, unless you feel that throttle response (not rpms, as you can still get these to whatever you want) is essential for a good launch at the strip. I'm not a drag racer, so I don't know how important that is. But when I've launched at autox's, I've always felt that it was more clutch technique and delicate accelerator control that was important, not throttle response.

Port design is always a compromise. The problem is that we want to cram as much air into the engine as possible, but for any engine, there is a max amount of air that can flow into the engine at any given time. If we break down the intake portion of the cycle into discrete steps, its easier to see. As the apex seal first passes the port, intake begins. The chamber is expanding, creating a vacuum which draws air in. If we assume there are no obstructions in the intake, then the air will rush in at each step and fill the chamber completely to volume, until the pressure in the chamber equals the pressure in the intake. At some point the rotor will reach the max volume. After that point, all of the air that rushes into the chamber has to overcome the increasing compression of the chamber as it decreases in size.

The problem is that air has inertia, so if we simply make the ports larger, or open earlier, we won't necessarily gain anything. At any given point of the rotor's travel, there is an ideal port opening size. This gets more complicated when we consider that as rpm's change, the velocity of the air in the intake changes, so there is also an ideal port size for every rpm. This means that its not possible with just port sizes to increase power across the rpm range UNLESS a portion of the stock port shape is sub optimal for all rpms.

So the secondary and aux ports come into play, in that we can use valves to close them off, and only open them at higher rpms, effectively giving us two or more port sizes. So why don't we just leave them open all the time? Well if our port opening is too large for the available chamber volume at that point of the rotor's travel, the air will fill the chamber and start to backflow, stalling the intake stream.

To optimize the intake flow we may want to sacrifice how completely we fill the chamber at a given point in exchange for maintaining velocity. Therefore if the chamber is expanding slightly faster than the amount of air we can get thru the port, the intake stream is increasing in speed since we are generating more of a vacuum. I *think* that ideally we want the intake stream to reach max velocity at the point that the rotor chamber is most full. At all points after that, the velocity of the intake stream will be decreasing since the chamber volume is decrease, which increases the resistance to the flow. So for all points after the max volume, max flow velocity point, we want to get any extra air we can into the engine while the intake flow velocity exceeds the chamber pressure. Once the flow velocity matches the chamber pressure, the port should close, since it would be pointless to stay open. This is important to note, because it shows that delaying the port closing can result in losses of airflow. Likewise, advancing the timing of the opening can also result in losses by equalizing the pressure between the intake and the chamber too early, stalling flow velocity.

So if our engine was to only run at one RPM, we could actually use a port shape that would be ideal for every portion of the intake cycle. But since most of us can't run an engine like that, the port has to be a compromise. Now we've said that things like actuated ports are basically a free advantage because they let us get more air into the engine at rpms when its efficient, and close to prevent losses at rpms when its not.

The throttle works exactly the same way. Since we can't have 10 or 20 different sets of actuated ports for ideal flow throughout the power band, we can use the throttle to control flow into the intake. By *decreasing* the amount of air we allow into the engine, its possible to maximize intake velocity at certain RPMs. Of course, I'm assuming that the throttle control software was designed to optimize performance, and not something else like emissions or fuel economy.

Now here's where it gets interesting... Since the throttle-by-wire unit almost certainly has a map to tell it how to operate, it should be possible to port an engine aggressively, and then modify this map to increase power and drivability at low rpms! A new mapping could reduce throttle openings at low rpms to increase intake flow, compensating for the increased port size. Yeah, this is all just speculation, so I'd be interesting the thoughts of the more rotary-knowledgeable.

 

that's not what i meant. under what conditions, like a bad tps, throttle plate motor, signals to ecu, etc...under what conditions would it NOT open fully--besides the obvious malfunctions?

 

Ahh, I see what you're saying.. Does the ECU ever pull throttle as a protective measure? This is pretty interesting. Seems like it would make sense in 'limp home mode.' No sense letting someone go WOT if they have a sensor out.

I went and read the other thread on Snoochies car and the 10.5 A/F ratio. It really sounds like the MAF values the ECU is reading are either out of range of the MAF's expected operating range, or that the MAF range is out of bounds of the fuel table. It makes sense that it would go into limp-home in either case.

An interesting experiment might be to find the max MAF voltage on a stock car, then wire up a circuit to limit the MAF->ECU line to never exceed that max. If it would stay linear up to that point, the stock map looks like its ok (you're a/f didn't plunge in the lower rpms).

I saw that the greddy turbo kit is a draw-thru unit. They can't be using the stock MAF and running its signal unaltered to the ECU can they? The only thing I can think is that they alter the signal before it gets to the stock ECU, or they sustitute a different MAF.

 

HAHAHA...that's like an fcd for a 2nd gen but we've got the canzoomer on there anyway.

 

No The greddy is not a draw through, My system is a draw through. That was my point, in a blow through the wire is hotter and requires less voltage to keep it at the rquired temp. Therefore it would fuel incorectly. On the other hand mine being a draw through the nominal air temp is the same as if it were NA. Why would it not continue to fuel correctly with the added airflow?

Knowing there is a point where the map no longer has a program because they never figured it would go there. But the MAF is of a very large size for the amount of flow this engine will use. I'm sure that is because it came from a Ford parts bin.
Why can't the ECU just be extended? There really is no need for exotic manipulation of the electronics, if this is a true mass flow system then nothing should matter. I can see a spark timing mod being needed but the fuel should be taken into stride by the mass flow system.

Am I dreaming here? Why isn't this a possability? The MAF doesn't know it has a supercharger, unless you guys told it so.

 

Again, I can't say this as fact, but off that same Ford parts bin (or Aisine's or whoever's) you probably have a generic ECU that does interpolation between points of your fuel matrix. Since Ford makes all sorts of engines, it probably makes sense to design of the software so that the min and max rpm's are range variables, and provide some fixed number of subdivisions in that range. That way you always provide the highest possible resolution for a given range. So if you need an ECU that goes to 9000, why should half your slots be wasted when that same ECU goes in the Ranger? Some of the aftermarket ECU's do this also. And it also makes sense to do the same thing on A/F, since its the other axis of the matrix.

So the Mazda ECU is probably fully programmable to any a wide range of possible MAF values, but since they know that say 50% of those possible values will never occur in their application, the rx-8 designers excluded them from the interpeted range. This has the effect of simultaneously defining saftey conditions, ie boundries, at which the ECU should interpet the signal input as in error and fail into limp-home.

Further, even if the ECU had positions available for high MAF readings, why would they be populated? The whole point of a bunch of cells of discrete values is that optimal fueling is not a linear function of rpm. Otherwise we could just input the slope of the curve and be done with it; which as I understand it, was pretty much how carb's are/were tuned. Modern FI comes with the benefit these nonlinear fueling maps which, over an entire rpm range, are more optimal than carbs. So the ECU doesn't really know what values to except at abnormally high MAF values since the engine was never tuned for them. The only thing the MAF could have would be the last fueling value replicated for all high MAF values, or some sort of linear function extending from the last fueling value.

However, the second doesn't make much sense. If you expect the engine to run in that range, you would just reprogram the complete matrix across that range. Mazda would have put in fueling some range of MAF's outside normal, but probably not very many rows, since anything much out of expected is going to be more indicative of sensor problems. Ie. the guy programming the ECU didn't say, hmm, I'd better put in MAF values that are really high in case some dude slaps a blower on this thing! :-)

 

Is it so strange? This situation seriously wouldn't surprise me. Maybe the ECU decides instead of cutting fuel to one rotor (like an FCD), lets just run ultra pig rich and prevent any detonation or boost creep or whatever since the engine knows that some besides normal operation is going on.

Its sort of the same trade-off, except that with cutting fuel, there is a boost point at which it won't matter, since the turbine will still manage to creep. If you just run rich, you'll be safe up until the injectors get maxed. Its like they said, 'if you're going to error in the ECU, make sure you error rich, not lean.'

I'm not familiar with the canzoomer, but can it clamp the MAF to the ECU? Have you tried it? Might be worth a shot.

 

Resistor from the MAF to the ECU? Because it is closed loop at low RPM it should tune appropriately even with a weak resistor inline. At higher flow rates though, it might stay within a normal operating range. A possibility?

 

Lucky Ford guys can just flash their ECU's... http://www.diablosport.com

 

That's just because Ford spent about 10 minutes total tuning the car :-)

 

hey, if it works...

 

Are the fords using the same Denzo boards in their ECUs?

 

i wonder if there's ever gonna be a power programmer--like there is for the gm's

 

What do you know about the aux nozzle systems used by Vortec and Paxton??

 

nothing. i was talking aobut the tuner module that plugs into the obd2 port...i guess it's a portable "flasher", but i could be misunderstanding its purpose.

 

As far as I can tell, the only way to program the ECU is to upload a new flash to it. You can do this thru a programmer, like canscan with the CAN pass-thru programming function. But, the problem is that the ECU will not accept a program that doesn't meet checksums or possibly isn't encrypted. In order to really program the ECU, you would need either:

-a WDS, which is a portable tablet computer that Ford uses for dianostics, that can do scanning (like canscan) as well as control the ECU parameters in real-time.

-a Source Development Kit for the ECU / ECU's OS. Its my understanding that Hitachi supplies the OS for these ECU's, so Mazda either compiles the OS and their settings together and publishes this binary as a flash, or they simply publish their settings and the OS imports them. A source development kit would let you create your own binaries or setting bundles. My guess is that the OS on the ECU simply accepts new setting bundles rather than having to be completely reprogrammed. This is because if the ECU was to loose power while reprogramming it, it would be essentially unbootable might require special tools or chip extraction to fix. In any case, this method would be preferrable over the WDS.

My belief is that porting on the Renesis will never be very successful unless there is some modifcation to either the MAF signal into the ECU, or to the ECU programming itself. The only way modifiying a MAF signal will be successful is if it can map the full spectrum of signals from the MAF on the ported engine, to the spectrum of signals the MAF produced on the stock engine. I think this is crucial because if the MAF signal on the ported engine is passed unmodified to the ECU, the ECU see the higher levels eventually as out of normal ranges, and reverts to a safe map with very rich mixtures. Its also important to note that I think simply attenuating the MAF signal to make the engine think there is less air (and thus cut fuel, leaning the mix) is not sufficient. The ECU is still getting other signals, most importantly the rpm's of the engine. The reduced or evelated signal level at an rpm may be treated as out of normal ranges. So if the ECU thinks its flowing the same amount air at 3k when its really running at 7k, it will go fail-safe since it thinks one of the sensors is malfunctioning. Also, I believe there is not just one 'limp-home' mode/map in the ECU, but several. This would explain why the ECU sometimes pulls power under certain conditions. Like on a dyno, the ECU may start to see values that are very unexpected (like rear-wheel speed at 60mph and front wheel speed at 0 mph) and decide to switch more conservative. It may go to this map and see if conditions improve or decline, and then switch again. Basically values that are slightly out of normal may decrease performance only slightly, so it may not be obvious that the ECU suspects trouble. This is to mask any errors that may naturally occur between the control components. The ECU must correctly handle and recover from these 'errors' but it may also see sucessive 'errors' as reason to switch maps.