MazdaManiac

iTrader: (3)

I late 2005, I installed the Greddy turbo/manifold and parts of the harness (the O2 dongle) into my custom system. I did this for testing purposes and to address some of the problems people have been having with the system.

There are three MAJOR problems with the Greddy system as it is shipped (with the "blue" e-manage):

1. The air pump runs for an unspecified amount of time after the ignition is turned off
2. The turbo is run on the verge of surge and is often over-speeding
3. The fueling scheme is inconsistent and causes the PCM to go into wild LTFT spasms

Problem #1 is, for most people, seemingly just an annoyance. However, running the pump when it is not needed drains the battery, wears and over-heats the pump and is functionally pointless. In fact, this problem actually keeps the entire PCM and EFI system running. In my case, the pump won't shut off at all if the car is put into the run position and then turned off without starting the engine! This is bad.
The fix is fortunately very simple.
The cause is the O2 fooler dongle that is included in kits released after the problem with the air pump outlet hose positioning in the downpipe was realized after complaints to Greddy. On the older kits, a CEL would be thrown because the ECU wouldn't see appropriately lean mixtures during air pump operation due to the late mixing of the air pump output (the factory setup has the air pump feeding directly into the manifold; the Greddy system puts this air in the downpipe right at the sensor). The dongle that was added sends a lean signal to the front O2 sensor during normal pump operation. It does this by hijacking the O2 sensor inputs to the PCM and applying a resistor across them when the air pump is turned on. The problem is that this dongle acts as a conduit to ground for the air pump relay when the motor is turned off because of the way it is wired. Fixing this requires that a diode be placed inline with the "trigger" wire to the dongle from the air pump relay wire on the Greddy harness.
This dongle has four wires - two light blue that are attached to the O2 sensor signal wires, a red wire that is hooked up to switched power and a green wire that goes to the air pump output of the PCM.
A diode (I always use 1N4007 or equivalent because they are cheap, readily available and can handle a fair amount of current) is spliced into this green wire with the cathode pointed to the PCM and the anode pointed to the dongle. This allows the air pump signal to pull the dongle to ground but not the opposite that would occur otherwise.
Simply locate the green wire for the air pump dongle and splice the diode in with the "banded" connection on the harness side. Do not confuse this green wire with the similar wire that goes to the temperature sensor dongle which has four wires as well - white, black, yellow and green.

Problem #2 is a little more complex and troubling, but the solution is equally simple.
Because Greddy chose to acquire the wastegate signal from the manifold rather than the turbo outlet, the turbo is constantly trying to produce boost. Nominally in all other systems I have either built or worked with, the wastegate signal is plumbed to the outlet of the turbo so that it will open when there is no need for boost.
Because the wastegate doesn't open at part throttle, the turbo spools up to very high boost levels - possibly beyond 20 PSI. This boost has nowhere to go as the kit is shipped and ends up stalling the compressor. Those with blow-off valves installed find that the valve often opens, causing a rich condition and stumbling as well as idle problems and stalling.
Unchecked, this over-speed condition will eventually destroy the turbo's bearings or worse. It also causes the PCM to retune for the lost air on BOV-equipped vehicles. Furthermore, it may damage the throttle body (some have experienced this, IIRC) and puts a lot of stress on the rest of the cold side of the system.
The fix is to simply install a vacuum nipple in the outlet pipe of the turbo and plumb a line to the wastegate. You can use an 1/8" nipple with an 1/8" NPT fitting tapped into the pipe. This can be done in about an hour with simple tools and a drill and the parts are available at Home Depot and the like.
The drawback is that the net maximum boost in the intake manifold will be about 2 PSI lower on cars not equipped with boost controllers. Those with boost controllers can program in the desired target boost and the net benefit is that the controller will have finer control over the boost since the cracking point will be effectively lower due to the earlier availability of pressure. This means equally fast spool without the surge and less over-shoot. This fix is absolutely essential to the survival of the turbo itself.
It also has the side benefit of making the part throttle/off-idle drivability much easier because the throttle isn't always fighting boost. Now you are not forced to spin your wheels every time you leave a stop light (though you still can if you want to).

Problem #3 is the result of a lack of R&D on Greddy's part. The fueling scheme on the E-Manage Blue relies too heavily on a mish-mosh of fuel adding MAPS that tend to interact with each other in a negative way and are further interfered with by the PCM.
The solution is the one that MadDog (Tim) and I had postulated and Tim worked out in painstaking detail over the last month or so.
It is the simplest of all three of these, though it takes a careful eye and requires the Support Tool and cable for the E-Manage - just cut two wires and load Tims most recent MAP.
The two wires are on the smallest connector on the PCM side of the harness (the middle of 5) and are wired to the E-Manage harness via two wires each - LB/R & G/R for one and W/R & P/R for the other. These are the signal wires from the PCM to the primary 2 injectors on each of the two rotors. The E-Manage connections should stay in place on the solid side of the harness adapter. Just cut the two yellow wires to which they are shrink-wrapped at the PCM side of the adapter. These are the top-right two wires as seen looking at the connector from the back towards the PCM itself. Refer to the wiring diagram if you are unsure which of these wires are the correct ones.
This small connector is only used on "high-power" RX-8s for controlling two of the unique features of that engine - the two additional injectors and the tertiary intake port activator (APVM). Therefore, it only has 5 wires going into it and the rest of the connector is unused. However, the GReddy adapter has all of the wires connected across the plug and socket for some reason.
The new fueling MAP that Tim worked out pushes ALL of the additional fueling requirements under boost off onto the primary 2 injectors where it should have been in the first place. These injectors are totally under-utilized in the normally aspirated condition of the Renesis and are placed completely under control of the E-Manage by severing these two wires.
The net result is almost no LTFT issues and a completely smooth transition into boost-fueling and consistent A/F ratios.

The combination of all the fixes yields a far superior driving experience and greater reliability and life from the GReddy components as well as the affected Renesis parts.

For those that have the newer kits shipped with the GReddy e-manage Ultimate, or have acquired the Ultimate upgrade kit, fixes #1 and #2 still apply.
The fueling scheme is entirely different, though it poses a different set of difficulties, depending on when you bought the kit.
GReddy is constantly changing the tuning on these systems and it is difficult to know what to recommend without first seeing how the injectors are wired.

MadDog

Great write-up, Jeff. Thanks for working on these issues for the benefit of the community.

I can't wait to try the solution to Problem #2. My only complaint with this system has been the almost unavoidable boosting on take-off. Its like I'm trying to race from every single stoplight in town! Its fun, but its takes too much effort not to peel-out every time.

I really hope your solution works equally well for me! Thanks again, Jeff. You are an asset to our community.

Petrus

Awesome info and DIY´s! This should be in the Hall-of-fame! (If there were one

rkostolni

Jeff should get a trophy. A big gold one. Thanks for your help. I'll definitely be installing the fitting for the wastegate before I put my new turbo on.

two rotors

Thank you for an excellent write up.Two minor questions:
- did you plumb your nipple into the discharge pipe on the compressor case(as on a TII for example ,or did you go into the pipe mated to the compressor discharge flange?

-is it absolutely essential to sever the P2 injectors,or are there still problems with the transition to boost and back.

MadDog

I ran it for a while during my trials with all the fuel pushed to the P2's without severing them. The only issue is that when both the emanage and the PCM are sending firing pulses to the injectors (over 6kRPM), the resulting signal can look a little ragged. Since the injectors get a ragged signal, it can make the AFR a little rough, too. Generally, this is not a problem since the injectors are delievering, on average, the specified fuel. But, in the interest of consistency and control, there's just no reason not to cut them loose from the PCM. The result is a very flat AFR curve. Plus, you should sleep a little better knowing that you are in total control of the fuel that you are trying add.

Before running this strategy, there was a pronounced tendancy for the AFR to run very lean when going from vacuum to boost. I saw AFRs of near 16. I never had detonation, but this made for some very disconcerting (and herky-jerky) moments on the freeway, where boost comes very easily! Now, I have no issues with AFR when going into boost. It goes gradually richer as I transition into boost without the annoying lean spikes that used to be there before a) you really got into boost or b) got nervous and laid into the throttle just to richen things up.

mikefrombarrie

Great info, but some pics would help out.

MazdaManiac

iTrader: (3)

I did it in the bolted-on right angle pipe that Greddy includes. It is easier to reach with the turbo installed.
In theory, you could do it anywhere before the throttle, but the closer to the turbo, the better.

See Tim's answer above.
Considering that it is a 30 second mod, I'd recommend it.

adrian-1

Does thing mean having to retune after a certain amount of time?
Also, what is the max possible boost attainable with this new fuel scheme?

MazdaManiac

iTrader: (3)

This seems to completely eliminate the PCM's tendency to tune out the changes, so after you are tuned it will stay that way.

Maximum safe boost at redline with the OEM injectors is about 7 PSI. You could probably go higher, but You would have to check your A/Fs.
With a set of the secondary injectors from an automatic tranny-equipped RX-8 in the primary 2 positions (like Tim has done) you should be able to do 9 PSI easily and maybe as high as 11.
It all comes down to your particular situation and any flow adjustments that may be in place. 1 PSI on my car is not 1 PSI on yours and 80% duty cycle on your car may be a totally different amount of fuel in mine.

Broke_Apex_Seal

where were you on my car jeff

Japan8

One question... so in this setup you are able to hold boost to redline, Jeff?

Ciao

Give Da Man a medal !!!

You should get a commish from Greddy Bro !!

Ciao

Dave

MrWigggles

I agree completely.

sammytcl

iTrader: (1)

Does this fix apply to greddy setup with interceptor-x also?

Thanks.

rkostolni

The new fitting for the wastegate control line does apply to those with the Interceptor. The other two do not.

MazdaManiac

iTrader: (3)

You broke it too quick!

brillo

Nice job Jeff & Tim, this info is great for those who don't have the money to upgrade the ECU and want to work with what they have.

its sad Greddy couldn't figure this stuff out, but it always seems to take a few blow motors in all cars to figure out the limits of aftermarket turbo kits.

I've been following all this closely as I've been doing some ECU tuning of my own NA. I think your solutions should provide stability and protection to the engines running less than 7psi.

This should be a sticky

rotarygod

This deserves to be a sticky!

BaronVonBigmeat

Sooo...does Greddy know about these issues/fixes yet?

MadDog

I haven't looked into this thoroughly yet, so this is half-baked. But, if you have a boost controller that totally removes the signal from the wastegate until you reach a certain boost, wouldn't this negate the fix to problem #2?

Nicky

YEAP

rkostolni

Not necessarily. If you tap the WG and pressure sensor for the boost controller at the manifold it will hold the WG shut until Xpsi is established in the manifold. But if you tap it directly after the compressor, or really anywhere before the throttle body, then it will open the wastegate at Xpsi in the compressor. The difference under full throttle will be minimal, 1psi or so. But at part throttle the throttle plate can act as a significant restriction, as its intended to. So the difference could be significant.

There's no need to make 20psi at the compressor to squeeze 7psi into the manifold because your at 30% throttle.

MadDog

^Assuming you have the pressure gage at the same place you are taking the pressure line to the boost controller. But, you should have the pressure gage after the throttle body - not on the turbo outlet. Otherwise, you, and your EMS, would have no idea what's really going into the engine.

rkostolni

I'm saying that your ems pressure sensor and boost gauge should be tapped after the TB, but the boost controller pressure sensor and wastegate should be before the TB.