Tienvu45

Hey guys, quick intro. I'm rotary swapping my Miata with an N/A S4 engine to be a streetcar/cruiser. I've rebuilt the engine and done a mild street port to it. I am going to run a standalone ECU and custom wiring. I kind of want to only run two injectors for the sake of simplicity and to not bother with staged injections. I know i will probably miss out on some power potential, but I'm not going to be chasing big numbers. I was thinking of running only two 1000cc injectors, but don't know if it should run them in the primary or secondary injector ports. Does it even matter? I also kind of want to do dual throttle bodies and do know which injector location would be better. Any insight would be nice, thanks in advance.

mr2peak

Primary. All the rotary specific ECU settings will expect idle injection at the primary port location.

I have never done this myself, YMMV

Pete_89T2

iTrader: (4)

What ECU are you considering? If you're going to be running pretty much any modern standalone ECU, it really makes little sense to run only 2 injectors, especially if you've already done a mild street port on it. The "extra simplicity" isn't worth it, and tuning the secondaries isn't all that difficult on most ECUs, plus if you're using the OEM intake mani and throttle body, it's designed for staged fuel injection. And if you eventually want to switch to an ITB intake setup in the future as you alluded to, you'll need those secondary injectors, so you may as well plan & wire for them now. But to answer your question, a set of ID1050s in the primary FI holes will suffice for what you're asking in an NA application, but you're going to take a hit in performance potential.

j9fd3s

iTrader: (3)

Primary, you want the injectors are close to the chamber as possible.

Mazda has an SAE paper that touches on this, i'm not sure which one it is though

TeamRX8

iTrader: (2)

it can work well enough on a race-specific engine, but it will act up worse than a poorly done bridgeport on the street

I almost did it on a 4-injector Renesis LIM, but thought better and bought a later model 2-injector LIM instead.

there’s a post on here a few or more years back by Elliott at Turbosource on one they tried to tune and had a lot of issues with.

and you’ll either need the NLA early-style ID injectors or long extension nose cups on the current ID injectors. The reason is the old style had a longer pintle with injector wiring connector further out toward the fuel rail. It will fit into the machined iron PRI injector hole no problem.

The newer style; which started with the ID1300, has a much shorter stainless steel pintle with the connector much closer to the injection end and the wiring connector on the body hits the PRI injector hole machined in the iron before the end seal will seat at the bottom.
.

Tienvu45

I am going to use a Megasquirt DIY ECU. I used it on the Miata before i started the swap and am very familiar with the wiring and features. After reading what you said, I probably will try to wire it up for 4 injectors just in case I change in the future. Do you know how much of a performance differences I would have? running like 4 500cc vs 2 1050ccs? My thought process is that modern injector technology is good enough where I could run larger injectors and idle much better than if I were to use older big injectors. Unless you're saying the injector location is what the performance differences would be from.

j9fd3s

iTrader: (3)

its a good idea to put the wiring in, future proof is a good thing. Mazda was able to get 2x680cc injectors to work with an ECU that is almost as sophisticated as a kitchen timer, i don't see why 2x1050's wouldn't be fine with an actual computer

the 84-85 GSL-SE had a staged throttle setup (its like a baby FC), but only 2 injectors, its not even sequential injection and it works fine. 680cc injectors too.
the 86-88 FC doubles the ECU speed to 4bit, and goes to sequential injection and 4x460cc injectors (on the NA)
the 89-91 gets an 8 bit ecu, which helps a lot, its still a noticeable improvement to run something faster
the Rx8 goes to a 32bit ecu (i think), and the early ones are 6 injector, so three stages, they later decided that was silly and went back to two stages

in the race cars they played with injector location, its in the R26B SAE paper

Tienvu45

I can see how running only 2 injectors will make it less streetable, but this is something I am going to have to see how bad it is. At the end of the day it wont be a serious race car or anything so I can live with some drivability issues. I bet i will be fine, but who knows. Ill run the 2x1050cc but have wiring for the other two in place, in case i decide to change. Appreciate the help man

mr2peak

Why spend so much on 1050s?. For the price of two 1050, you can get four normal injectors and the performance will be better for the money spent. The benefits of the 1050 are injection accuracy and fuel volume, but the injection accuracy doesn't matter as much if it's all going in a single port, and you can get the same volume with 4 smaller injectors..

Why bother?

Tienvu45

You do have a point, running multiple would allow for more even air fuel mixture. I think possibly running only primaries would cause the air fuel mixture to be not so even/mixed as it enters with the intake cycle.

dguy

iTrader: (8)

If its for simplicities sake, just run primaries in the iron. If you're going to run an IDA style, run your two injectors post throttle plate at a 45ish angle on the outside radius of the manifold.

gxl90rx7

iTrader: (1)

ive been running two injectors for years now on a 350whp single turbo 13BREW. i was frustrated with getting staging to work correctly on my haltech ps1k, so was like **** it, blocked the secondary injector holes with freeze plugs and threw some 2200cc in the primaries, good for 400whp or so. works ok, but some downsides i have noticed:

not 100% sure, but power seems to flatten about 6krpm, may be because primary 2x afr and secondary 0x afr runners dont mix until in the rotor chamber, so maybe not ideal. no problem at low rpm, still got 25-28mpg highway

not sure if related, but I also warped my apex seals one time at track day. i found out my TPS was unplugged, which caused it to dump a lot of fuel off throttle, normally have decel fuel cut enabled, i think maybe the extra off-throttle fuel warped the seals. had done 20+ track days on this setup, this was the only time it warped the seals. having cold fuel spray on primary side of the seal, with nothing on secondary side probably not very good for seal warpage under extreme conditions

last time i tore engine apart, one side of the rotor faces had more carbon buildup, the injector side was cleaner

if i did it again, i would use 4x 1000cc injectors in primary and secondary, no staging. it will still idle around 12.5AFR at 1.0-1.1ms pw on pump 93, just like my 2200cc primaries, dont have to **** with staging, and should have more even fuel distribution in the chamber

mr2peak

You shouldn't run 4 on a single stage with the OEM throttle body. The secondary throttles don't open at low throttle and idle, it will just pool and drip into the chamber. Throw a single throttle body on there and that issue will be mostly resolved.. next issue will be injection phasing, secondaries are further from the irons. Mazda used injection staging for a reason, and it wasn't a government regulation reason..

peejay

I went for years with just a pair of GSL-SE injectors running out to 100% duty cycle on stock port, street port, and bridge port engines. They seem to be good for about 210-220hp at the crank if your computer can run duty cycles that high.

I currently have two ID1050Xs and I am not really thrilled with them. They seem to heat soak pretty badly. A tune that is rich at low pulsewidths becomes painfully lean after a couple hours on the highway. My pulsewidths are in the 2ms range, give or take, and starting out at 12.5:1 had the tune running as lean as 16:1 under cruise after a while.

I threw out the spacers and used flat O-rings to mate them up with the injector diffusers. I feel the diffusers are super important for good idle and drivability. I am using a Racing Beat Holley intake manifold and went to the trouble of adding an air passage for the air bleeds.

They are still nowhere near as bad as the RC1000s I used to have. One injector flowed 850cc/min and the other flowed 750cc/min, at the GSL-SE rail's 2.5 bar fuel pressure.

peejay

They do flow a certain amount of air. Granted, it's still not a great idea, but not for that reason. But it's not as bad an idea as:

I did that on a spare T2 intake manifold. Used a 75mm LS1 throttle body on it. It drove like complete garbage at low throttle openings and didn't make any more power.

It was, sort of. Very Large Injectors were not really available in 1985, neither was computer processing and control precise enough to control them finely enough to make a catalytic converter happy to meet emissions standards. The largest commonly available injectors at the time were the ones used in the GSL-SE, and were for a very long time. The only production injectors larger were the ones from the 12A Turbo at 800cc/min.

mr2peak

So we should be running 4 injectors and spend the time to figure out the staging?

TeamRX8

iTrader: (2)

with the latest current ecus you should only need to plug in the basic information and then it stages them automatically based on demand and any other relative parameters that are set, could just depend on the brand/model though.
.

peejay

Or you can just run two injectors.

When I first did my Turbo II engine, I tried running all four injectors, and found that the staging was almost impossible to get set up without knowing precisely what the injector constants were. For example if you have a deadtime of 1ms and a duty cycle of 4ms, then it will power the injector for 5ms. When it switches to staging it will split the difference between the two injectors, so deadtime of 1ms and duty cycle of 2ms, to power the injector for 3ms. However if your actual deadtime is .8ms, then it "should" be splitting a duty cycle of 4.2ms to 2.1ms with a deadtime of .8 = powering the injector for 2.9ms. So in this example it will be injecting too much fuel and it may bog.

It gets more complicated when you factor that deadtime is not the only thing to worry about, but also the precise injector output, which is not always linear at lower duty cycles, power demands changing the voltage drop in the circuits, the different pressure pulses (water hammer) in the fuel rails when it goes from 2 injectors to 4 injectors opening, especially if you've eliminated the pulsation damper because engine fires are bad....

dguy

iTrader: (8)

These are issues that are no longer relevant. Injector data is either supplied or can be divined and profiled by more than a few services.

This thread seems to be morphing into a fair amount of what ifs rather than sticking to the original posters initial constraints of twin injectors and ITBs on an NA street car.