i think the drive by wire, and can network is the big deterrent for installing an ems in the rx8. so i started thinking about how to make everything work...and work properly.

well, unless you buy a $5k motec, you're about **** outta luck with the drive by wire, so let's just stop this before it even gets started.

so piggy back then...but you don't have the control of a true ems. or do you? how about you piggy-back all the input signals from the major sensors, but you "pirate" all the output signals?

the stock ecu will still see engine rpm's and hopefully control the drive by wire, along with communicating with all the other junk it talks to.

here's a crappy drawing i did to illustrate (poorly) this idea. i'll get into the other stuff later, but here's the basic idea:

 

excuse the pic...i'll get a better one.

the maf and map words are supposed to be lined up with the wires just to the right of them. in the pic, they're cutting across the crank position sensor line...don't worry about it--my bad.

let's pretend it's a haltech just for the hell of it. like most aftermarket ems's, it's a speed density system, so you don't even need the signal from the maf sensor-instead, wire in the map all by itself. inputs from the crank sensor, ect, iat, tps can be piggy-backed, so both the stock computer and haltech share them.
now, the coil signals and injectors will be totally controlled by the haltech. full ems benefits, without the drawbacks of taking the stock computer out. also, this may or may not be technically smog-legal (depending on where you live), but the stock computer could still control all of the emissions stuff, so at the very least, you could pass the sniffer.

actually, you might be able to pass the sniffer with a little heavier porting. if you port it to the point where there's overlap, forget the sniffer... the ecu runs the motor in closed loop at idle (not like previous generations), so you'll have a bit of a roughness there--which means misfire--which means HC's go up. with the stand alone, you can have it idle a tad richer--whatever it takes to smoothen it out, and have the smog pump dumping into the cat at idle to clean things up.

i really wonder if you could even get as radical as a small half-bridge and still pass the sniffer. yea, it's a long shot, but i wouldn't rule it out. earlier motors had a lot of overlap, so when you bp'd them, you got that lovable "brap brap brap" idle. if you start with zero overlap, then add the overlap with the bridge, there still wouldn't be as much overlap as one of the older motors, so it might end up idling like a big streetport. i dunno....it probably wouldn't pass, but it's an idea anyway.

back to the drawing...i drew single lines to the sensors for simplicity. if you don't know about the reference voltage and ground, don't bother wiring it up yourself.

i'll start on the quirks as soon as i can. if anyone notices anything i missed or got wrong, lemme know.

 

now the first problem i thought of is the secondary injectors... well actually, it was how do we make the crank signal work, but i'll deal with that later.

for those of you who don't know, the haltech sets the staging injection based off of load. it doesn't acount rpm. so what if you stomp on the gas before the secondary ports open? probably a nice fuel buildup in the manifold.

so how do we keep the secondaries from turning on before the ports are open? we rely on the stock ecu to trigger it! this is assuming (and i may be wrong) that the secondaries start firing when the ssv activates. if i'm wrong, scratch this idea and i'll think of something else.

anyway, the power wire for the injectors gets wired to the primaries like usual, but the power to the secondaries has to pass through a relay first, and that relay is only active when the secondary ports are open. no secondaries, no unwanted fuel puddles.

again, single wires for simplicity.

 

i'll stop here until i get some replies. maybe some of you guys can either find a better way, or maybe elaborate a little more. but the ball is rollin' mentally anyway.

i always think of **** like this when i can't sleep.

 

Where are you going to get the tps signal from? Since the engine does not always need the throttleplate open all the way due to rpm and engine air requirements at lower rpm's, just because your foot is planted to the floor does not mean that the throttle plate is open all the way. Depending on the rpm, it could be open only half way when you have it floored since it flows as much as the engine can handle. Are you going to try to integrate a tps onto the throttlebody anyways or will you try to mount it to the gas pedal inside the car? Think about that one for a while!

Personally I'd use a conventional throttlebody and run a cable to the gas pedal. I'd leave the stock throttlebody motor attached to the stock ecu tucked out of the way somewhere so it thinks it still has control.

Are you sure the Haltech won't allow staged injection based on rpm? Even my $150 Megasquirt gives you the option of either load based or rpm based staged injection.

I don't know about other states but in Texas if the car has OBDII, you do not need to pass a sniffer test. Just plug into the OBDII port and if there are no CEL's, you pass. No visual inspection or anything. It's simple here.

 

yea, thought about that. tps on the gas pedal. (doesn't it have one anyway?)



that's my contingency plan in case the drive by wire doesn't work.


yep...it's load based.

 

how are the secondary ports controled vac or electricaly?
how many injectors do we have?

 

All of the actuators except the auxillary ports are opened with vacuum however there is an electric solenoid inline with each one which opens to allow the vacuum to do it's job. So the answer to that one is both.

The automatic cars have 4 injectors, the standards have 6.

 

can you fire 2 pairs of injectors together? most engine managements can't handle more than one bank of staged injectors so you would need 4 of the injectors to be the same capacity and be fireing at the same time (duty cycle). so the renseis would need some injectors changed wouldn't it?

if the secondary ports have an electrical signal you could wire a circuit that allowed the injectors only to work when the valves where open, or the other way round and have the ports actuated when the staged injectors came on.

 

I've seen RG's megasquirt box for his 1stgen, and its a cool little device that might work for your purposes and its 1/5 the price of a Haltech box. Granted you have to build it yourself but its not a big deal. the MS can control everything your thinking of in your drawings sans throttle and traction control. It might be fun to see if we could get the stock ECU to handle those for us, and control the rest of the car with the MS.

 

Does the (now defunct, or so I've heard) Haltech offer auxiliary low volt or ground outputs based on rpm? Something like that would be handy if you were hoping to incorporate an ecu-controlled shift light, warning buzzer, wheelie bars, etc... If so, you could use that to trigger the secondary injector relay, and bypass that part of the stock ecu completely.

 

This tread makes me feel like a dope. I miss carburators!!

 

yea, you can fire (i think) up to 4 injectors per driver on the haltechs, and there's like 4/8 drivers.

your second question was already answered

 

hey, we may end up having to do this with snoochie's car. the emanage isn't taking enough fuel out.

 

Allow me to highjack this thread for a moment. No! it has nothing to do with airplanes. I'm wondering what you EFI junkies think of this for my system.

I want to leave everything stock intact. Then add two or more injectors before the SC. These would only activate when the map went positive. The piggyback would not change the way the engine runs until it gets into boost, then it would control fuel based on man press and RPM. A speed density system running along side the MAF. Both are independent though.

I know the secondarys would cause some concern. But they could be left slightly open at all times, preventing fuel build up. In my case the fuel build up would be minimized because it is so well mixed in the air that it would travel with it.

Rg what do think of the fuel working that way, can I not bother with the A/F below boost on a rotary? I mean it runs there now, will the sensors somehow know there is a SC on board? When it goes positive will the stock ECU try and do something? Not to leave you out Guitar man can you opine on the electronics of it.

 

I don't see an issue with running the fuel injectors in front of the blower so long as they come on higher up the rpm range where we know intake velocity is fast. My concern as you've mentioned is with fuel building up inside the runners of the closed off ports. On the other hand, would this really be an issue? If there is no air flowing through them, it merely acts as a spring and all the airflow goes around it. Wouldn't the fuel avoid these dead runners too?

If you aren't under boost, no need to touch the map there. If you leave any runners open when they aren't normally, the ecu will know something is wrong based on the a/f readings it takes. I am guessing here but I would watch out for this.

 

Why/when would you not want the vacuum hitting the secondaries? I would think that there would be some threshold of vacuum needed to start to open the ports, and that this would be matched to the RPM when they should begin to open?

So since the ECU controls when the vacuum can be applied to the secondaries, does it always activate the solenoid at a set RPM, or does it vary? Also, the solenoid is not positional, right? ie. its either on or off? So once its on, the opening amount of the secondaries is directly tied to the amount of vacuum?

Sorry for all my questions guys, just trying to learn more about this engine.

 

When it goes into boost there is plenty of airflow so no worry there.

I think there is a possability of some fuel build up behind the butterflys but I'm not sure 'cause I don't know how far away it is from airflow. I'm not affraid of cracking the butterflys. I don't think it will change a thing.
Next, why would there be more airflow if I cracked the secondarys? Isn't the airflow controled by the main TB?
Just because they are open doesn't mean the engine will want more air. OK let's say it does, the driver automaticly lifts on the gas. He isn't aware of anything niether is the engine.

The engine only wants to eat so much depending on the TB opening. I take that back, the engine will always eat the same volume of air no matter what. The TB only controls the density. Mass flow.

Question is will the ECU know anything. It does not have a MAP sensor so doesn't read manifold pressure. What will the MAF think when a lot more air is going in??
Is it programed to shut down when it thinks a sensor is bad. Go into a limp home mode? Will it try and shut the throttle down??

These are the important things I have to know. Then how to trick the thing. Did someone say Webers. We can't sell it with a Motec, it costs more then the blower.
We need a simple solution. Cheap, cheap, cheap. We want to sell to the masses.

 

Good point. I didn't think that one through too well when I typed it. That's what I get for looking at 2 different sites at the same time.

 

Doesn't the ECU get gas-pedal-position (GPP sensor? dunno what its actually called) and control the duty cycle of the wave to the servo on the throttlebody? How is this different than a TPS? I also thought that the throttle pedal used two potentiometers, each referenced at a different voltage, and then averaged the two signals to cancel noise. Once you have the mean signal, isn't that basically the same as what a conventional TPS would provide?

The thottlebody actuator is a sevro isn't it? Or is it stepper? If its servo, then I would think there would be an actual TPS on the throttlebody for feedback to the servo controller.

The question then is, how does the ECU control the throttlebody from this signal? My thought would be that the ECU would have a map of RPM to TPS values, and a map of GPP to TPS values, so when you mash the pedal, the ECU would see that the GPP->TPS value is higher than the current TPS value, so it would increase duty cycle until TPS reads the same as the RPM->TPS value. This would mean you would actually need 2 maps for RPM->TPS, one for acceleration and decel, so that a given RPM->TPS value will actually result in the RPMs rising or falling, which then gives a new RPM->TPS value, and the cycle continues until TPS equals the GPP->TPS value. My take was that whole point behind drive by wire was that there is one ideal throttle opening per RPM, so while you want the engine to rev up as fast as possible, you don't want the throttle to open faster than the RPMs can increase.

 

along with the piggy-back idea:

what if you pirated 2 of the secondaries with an additional injector controller? i'm sure you could get it tuned about where mazda had it originally, and still give the extra gas needed above boost.

 

The idea behind this excersize is to get the homoginizing effect of the axial flow blower. It is my beleif that this is the rotary weak point. It is the area of the combustion chamber that creates a varying amount of A/F mixtures scatered along the way. You must tune to the leanest point, meaning you are rich everywhere else. That is why you have to put in so much fuel.

If I can make the mixture better homoginized to the point where it resists breaking down then you could increase the efficeincy of the engine. It is with this teory that I would like to do some dyno work with a four port engine. AT guys don't attack me, I'm proposing a research project here, just to prove or disprove my theorys. An in car installation would not meet my needs.

My SC is the only one that can utilize pre blower injection. Bonus is a great reduction in intake temp. Keeping the boosted temps below the critical point for rotarys. This would allow more boost without the IC. I have some proof of this using V8 engines but never was given a buget to get lab clinical data.

I think it is very important to try this with a rotary. Are you listning Mazda?
I've posted this paper before but some of you may not have seen it. It discribes my experiance with injecting fuel into the SC.

 

i get what you're saying.

i'm not diggin' the e-manage on snoochie's car, but maybe it's something i'm doing wrong. could you use the extra injector driver from that?

 

great reading, btw.

now i wanna get an axial flow for the fd and try that out!!

 

That is one of my questions, can I use the E manage w/ the injector harness for what I want. I sent them an E mail but did not get a response.
I think the piggyback will have to do some extra work in order to help the stock unit understand something else is going on but it is ok.