RX744CSP

iTrader: (4)

I have a S5 motor using a hybrid of S5 and S4 stuff for the engine management; specifically an S5 air flow meter, an S4 ECM, a S4 TPS and all other sensors are stock S4/S5 common stuff in their stock locations. I am using an MSD fuel pump with the stock pressure regulator and injectors. The thing runs super rich, to the point where it will foul plugs very quickly and whatever it is affects both rotors equally. When the plugs are removed, there is liquid fuel on them, both on the leading and trailing plugs. I get no signal to the secondary injectors under 3500 RPM, which is normal.

I'm going to jump deeply into this thing in the next few days, ohming all the sensors and harness, checking the fuel pressure etc. but was wondering if there is anything else I should look for.

Also, if for some reason the primary injectors were connected backwards (rear rotor wires which is light green with black stripe plugged into the front rotor injector which is supposed to be light green with no stripe and vice versa) could this create a problem?

jjcobm

iTrader: (18)

As far as I know, the S4 and S5 AFM's are completely different. Don't know why you opted to use all S4 electronics but managed to get an S5 AFM on there. If using an S4 chassis as the base, you typically use the stock s4 wiring harness, an S4 ECU, and ALL S4 sensors. Doesn't matter what engine you are using as the engine is just the block, the sensors are what are interpreting what's going on in the engine. Having a "bad or disconnected" AFM will usually make the car run rich.

RX744CSP

iTrader: (4)

I'm using the S5 AFM mostly due to space concerns, the engine is not in an RX7. The S5 AFM with an S4 ECM is a pretty well documented swap and going over the two FSMs I have the resistance values are almost exactly the same, the main difference is the S5 does not have the fuel pump switch.

I've gotten all the specs for everything together and plan to jump into this with both feet tomorrow night. I'll be ohming the AFM as well.

RX744CSP

iTrader: (4)

Okay, tonight's results: all the sensors fell within normal range including the MAF when compared to early 'flap type' specs. Fuel pressure = ~39 PSI, it varies with throttle opening as it should. Pulled the injectors, wired them to the logs, no leaks after 10 minutes at ~40 PSI. Pinched the return line to spike the pressure to ~ 80 PSI, still no leaks. No raging vacuum leaks.

*Allegedly* the late S4 injector has the same impedance as the S5 injector but with a different shape plug. That's real hard to confirm, so far it seems that NA cars had one type for 1986, then '86 and '87 had the same injector, then '89 up all had the same.

I believe my ECU is from a 1987. I'd have to snake under the dash to see for sure, but IIRC it's a N352 piece.

I'm running out of idears.

jjcobm

iTrader: (18)

It's not allegedly, the injectors ARE different for the 87.5-88 years.

http://mazdatrix.com/c-6.htm

I can't see anything else wrong besides those injectors if you have been double checking everything else. It seems like you have gone through the whole nine yards. What reading does ohming out your injectors bring?

RX744CSP

iTrader: (4)

According to my '88 and '90 FSM's, they both require 12-16 ohms. All mine hit around 13 but they are the later S5 'oval plug' type, not the earlier 'square plug' type.

Rotary Resurrection has a page that covers injector swaps, it seems that all the ECU's from 86-up need to 'see' 13 ohms, give or take, at the injectors. The 86- 87 1/2 cars used a resistor pack to achieve this. Again according to that page, low impedance injectors can be used with all ECU's by using the resistor pack and high impedance injectors can be used with any of the 86-up ECU's as long as you remove the resistor pack, if it has one.

So I'm still stumped. More wire chasing and head scratching tonight, maybe I'll have an epiphany.

Oh, forgot to mention something: I removed the 6PI actuators, sleeves etc completely from this motor. I rev the snot out of it so low end torque is not really important. on the first motor, the sleeves were in place but i turned them and locked them in place so the ports would be open all the time.

RotaryEvolution

iTrader: (30)

then you're leaving out an essential piece of info here. so what in the hell are you trying to run with all the stock electronics? gonna assume it's a renesis in which case the variometric efficiencies of the engines are too different that it will not run properly with S4/S5 electronics.

jjcobm

iTrader: (18)

Have you tried using the stock S4 AFM?

What is the condition of your fuel injectors? If they are old and clogged, with the MSD fuel pump, the system will be seeing higher pressures (I am assuming thats why you have the MSD pump?) than stock. Pair this with leaky/clogged injectors and the fact that your are now pushing MORE fuel through the injectors using the STOCK ecu, you will run rich not only due to the engine seeing more fuel, but your injectors may be LEAKING fuel like mad.

As far as the High/Low impedance injector issue. I am not certain how Mazda designed their low impedance system. Typically low impedance injectors use a PEAK and HOLD driver. Since the low impedance injectors require MORE current to open then High impedence injectors, it will deliver a SPIKE of current. Then the ECU will lower the current once the injectors are open to "hold" them open. When they need to be closed the ECU cuts the current. High impedance just gets straight burst of current the whole time the injector needs to be open.

Now for the resistor pack. I find it hard to believe that Mazda just made a box, and put 4 resistors in there for the low impedance injectors. I would think they would of made some sort of peak and hold driver for the injectors, but again, they have screwed up a few things before and I also believe some other car manufacturers did use just simple resistors when going low impedance.

The only other thing I can see making the car run rich is a mis-routed vacuum line. That would do the trick (or a vacuum leak). Always will. All this with the assumption every other sensor is in spec.

Engine compression numbers good? Timing good?

RX744CSP

iTrader: (4)

Not a Renesis, man I wish it was. It's an S5 long block with S5 intake manifold, custom headers and the S4 electronics all in a 1974 Jensen Healey that's a dedicated hillclimb and AX car. FI is a must for this app, some hillclimbs can cover ~1K feet in altitude change in 2 minutes. Keeping a carburetor happy under those conditions is nearly impossible.

I sat down with a 1990 and 1988 FSM, compared part #'s etc for everything I could think of, researched my azz off, then built from there. The wiring harness is an S5 with all the electronic OMP stuff gone spliced to an S4 ECU and ECU plug. Trivia: all the other color codes are the same. When picking the ECU, I went through that same FSM and found that the 1988 and 1990 injectors speced the same resistance, so I went hunting for a late S4 ECU. I did another pinout tonight, everything matches the stock setup as far as position of wires in the connector and resistance values for the sensors.

The MSD pump is there because it fit the area I needed it to and the max pressure specs are ~100 PSI, very close to the stock pump's specs. Last night when I pinched off the return line, it hit about 80 and stayed there, about where the stock pump should be. The injectors still didn't leak.

Teach me here, jjcobm: are you saying the fuel pressure could be too high and the ECU can't shorten the pulse width to make up for it, or if the pressure's too high it's forcing the injectors to stay open too long? BTW, I re-ohmed the injectors, one of them was a bit flaky and I couldn't get a constant resistance reading from it, even though it showed no short to ground. I checked my spares, they all tested nice and smooth so I subbed one of those for the (possible) weirdo.

I have sorta considered Megasquirt etc but I see no reason this can't be done with the stock ECU, not to mention it's time consuming as hell to build and test a Megasquirt, time I just don't have at the moment. BTW, I did get the variable resistor, altitude sensor, etc from the S4 and all that is in place. Pretty much the only thing missing is the wiring for the rat's nest solenoids and the various transmission switches for emissions stuff.

But in spite of all this testing the damn thing just runs pig rich. I have a second ECU, it does the same with it. I can't find a reason for this to occur.

I put everything back together again tonight, looking carefully for vacuum leaks etc. I didn't see anything, but it's a bit late to get out there with a can of carb cleaner to check for vac leaks the olde tyme way. Tomorrow I jump back into it again.

misterstyx69

iTrader: (142)

what harness is running this setup?and are you using a resistor pack?
a N326 and Lo's, is the Norm for a S4 N/A Wired engine.(with resistor pack on the harness).

jjcobm

iTrader: (18)

You have one very crazy setup, I am getting headaches thinking about it. Typically people just run the S4 engine harness since the s4/s5 are different. But you did mention you checked all the pining on the harness and re-pinned as necesarry. You get all the correct signals from the sensors. Then the ECU should be happy but it's not I really can't emphesize you need to be dead certain those pin-outs are proper if you are using a s5 harness into an s4 ECU. I don't have any experience working with different series harnesses, but from my experience working with EMS and ECU's one wrong pin and it will not run properly.

You are on the right idea. Since you have more pressure in the fuel system, that will mean you will have more amount of fuel flowing through. As you increase pressure, more fuel will spray for the same "x" amount of seconds of a lower fuel pressure. Maybe try swapping the secondary injectors into the primaries and see how it runs. Could be a bad injector in there somewhere. Still, the higher fuel pressure shouldn't make the car "pig rich" to the point it fouls plugs quickly, just a richer mixture than normal unless you are running some crazy amount of pressure.


First get this setup running. If you go straight megasquirt but can't get the stock ECU to run, it will be more headaches.

You checked the timing? Good luck and hope you find the problem soon. This sounds like a headache

RX744CSP

iTrader: (4)

I used the S5 harness since it fit the injectors. I did discover during that period that the S4 MAP sensor has 4 wires, the S5 has 3. So I added the extra wire and ran it to the correct pin on the ECU. I switched to an S4 TPS. I probably could have used the S5 piece, but decided not to mess with all that. All the other sensors (coolant temp, IAT, etc) are the same on both setups.

misterstyx69, from what I understand the resistors are needed ONLY if using low impedance (2 ohm) injectors. If using the high impedance (12-16 ohm) injectors the resistors are not needed.

I'm digging back into it this morning, a buddy suggested that I see if the secondary injectors are firing all the time instead of over 3500 only. I guess the RPM signal for that is the same one that runs the tach, and it works correctly. I'm also going to reconfirm the 5V reference signal to each sensor.

On fuel pressure: every FPR I ever saw increased fuel pressure with vacuum drop, i.e. if you whomp the throttle, the vacuum drops, this bumps the fuel pressure up to richen the intake charge, very similar to an accelerator pump on a carburetor. The FSMs are at best vague on this, they just say to check the fuel pressure and it should be within specified parameters. Last night I started wondering if I should put the orange vacuum solenoid from the 'rat's nest' back into the FPR vacuum source. My understanding of the operation of that thing is that it is only there to keep the engine from being hard to restart when hot. But I could be wrong; I've been wrong before.

On ignition timing: I was checking that the other night and my timing light took a dump. If it ain't one damn thing... So it's close AFAIK, but I need to bum a TL from someone and double check.

jjcobm

iTrader: (18)

I am not sure on the FPR vacuum line, but do know on TII engines, when not hooked up, causes a lean mixture and engine to go kaboom. I always have that line hooked up on my N/A. Have you tried to connect the FPR vacuum line to the UIM? You shouldn't need to have it go to a solenoid, I have had the emissions equipment removed and routed it straight to the UIM and it worked fine.

RX744CSP

iTrader: (4)

Still battling this sonofabitch. I verified the AFM is working by unplugging it while the engine was running, it nearly died and straightened out when I plugged it back in. I have two sets of plugs and I do believe they are both fouled beyond the regeneration capability of glass beading and carb cleaner. No one has plugs, I'll get some on Monday and try, try again.

jjcobm

iTrader: (18)

Did you try swapping the secondary injectors into the primary?

RX744CSP

iTrader: (4)

No, I didn't. They show to be the same P/N etc and all read the same impedance, is there something I missed?

RotaryEvolution

iTrader: (30)

sorry i misread what i quoted, sounded like it said it's not an RX7 engine.

make sure no one has taken the blind cap off the AFM and monkeyed with it, it is an air bypass screw that allows air to bypass the AFM flapper/cone to adjust the idle richness/leanness.

jjcobm

iTrader: (18)

There is always the possibility of the two primary injectors being completely bad or clogged. Just a suggestion of swapping the secondaries into the primaries and re-test. Unless you have tested them before and are 100% certain they work there is always that possibility.