I have a street ported S5 TII motor with stock turbo, stock ECU, greddy FMIC, RB 3" exhaust, stock injectors, walbro fuel pump wired for 12v, aeromotive FPR, emissions removed, RX7.com FCD, koyo radiator. no ac, no ps, electric fan

Innovate LC-1 wideband o2, XD-1 digital gauge, new sensor, recently replaced and re calibrated.

The car starts and idles fine. When cold the AFR is around 12 at idle, 12 to 10 when driving, the car pulls strong and drives like I would expect it to.

As soon as it gets warm however, the smallest amount of load makes the AFR climb fast, 13, 14, 15, 16.... It will go over 14 just pulling out of the driveway. It appears to happen at medium or high throttle, but I'm trying to avoid high throttle because I don't want to go into boost with a lean AFR. If I lightly feather the throttle I can slowly build speed and stay under 12 or 13. But if the load increases, like trying to drive uphill, the AFR goes lean again.

I have a new FCD from rx7.com, but the problem happens with and without it installed.

I have an aeromotive FPR set at ~38 psi fuel at 0psi air. It appears to fall 1:1 with vacuum and rise 1:1 with boost.

I ran a new positive 12g wire from the fuel circuit open relay (under the dash), directly to the pump, according to the S5 wiring diagrams, so that the aftermarket pump always gets 12volts when it is on.

I have tested all of the sensors [at room temperature] (AFM including built in temp sensor, Pressure sensor, water temp sensor, intake air sensor, and TPS), according to the fuel and emissions section of the FSM. They all appear correct.

My question is: What condition would cause the engine to lean out under load only when warm?

Thanks if you can help. Feel free to ask questions if there's more information I can give. I've been trying to solve this problem for over a month and can't drive my car. I've been tirelessly studying the FSM and wiring diagrams and reading any relevant posts I can find on the forum.

Additionally, I don't know if this is related. My water pump only has one sensor on the back. My water temp. gauge in the dash is not working. It worked fine before the engine swap. Do some cars have a separate temp. sensor for the gauge and ECU? The 3 NA FCs I've owned, S4 and S5 have all only had one sensor on the water pump. I'm pretty sure this is a turbo water pump because it has the coolant line for the turbo near the bottom radiator hose.

 

Here's my theory:

Some (or one) of your fuel-injector grommets may not be sealing well as the rubber heats up. Perhaps they're slight warped and when they get warm they flatten or "squish", causing a small but potent vac. leak directly at the injectors. Especially if you're using stock and higher-mileage injectors will hardened rubber grommets/boots. Think about how hard these little leaks could be 'sucking' under load--mixing right with the fuel.

To test: Buy a brand new canister of carb/brake-cleaner... Drive your car for 10-15 minutes, or until it's fully hot and you're experiencing the problem. Park and let the engine idle. Pop your hood and stick the nozzle of the carb-cleaner as close to your primary and secondary injectors as possible and spray the CRAP out of them. If your idle changes, then you've found your leak and the reason for your lean-condition.

 

I don't know about the first part, but the stock temp sender for the coolant gauge is on the rear iron on the pass side, next to the oil pressure sender. Check the connection to it.

 

The idle mixture will be rich until the engine water temp is up to normal. Say 11afr when cold idle then rises to 13afr or so when the engine temp is up to normal. That is just an example.

It would have helped as to what the afr was when the engine was cold at idle, then what the afr was when the engine temp is hot at idle. It's most likely as I described above. Similar. That would be a normal condition.

Maybe your aftermarket FPR? Zero pressure the fuel pressure should be approx 37psi and then add a psi for each psi of boost. At idle the fuel pressure should be around 28-32psi (vacuum being pulled at idle).

I've no real idea why the afr are that low (15-16 under LOAD). IF the load is under 3800rpm, then It can't be secondary injectors.

Crusing along at 65mph steady will result in a afr in the 14's if the 02 sensor is connected up to the ECU. Normal. Let off the throttle just a hair while at 65mph (front fuel injector cuts off) and the afr will actually get a bit rich, like a full point richer, til you put the pressure back on the pedal to maintain 65 mph. Let off the throttle a lot and the afr will go into the 22afr or more range (both fuel injectors cut off by the ECU).

Drive along as above and get on it so the boost is now a few psi (2-3psi), and the afr should go to the lower range. Like maybe 12 or so, give or take. Depends on *stuff*.

It wouldn't hurt a thing if you got something to give you more fuel under boost. Your going to need to do that. I still don't quite get why you'd need anymore fuel for low load conditions.

You might try something that sounds retarded. Put the original stock afm/filter assy on the car again. Then go for a ride and see if things are much different. IF better as far as afr, then since the afm controls the amount of fuel delivered, maybe you just need some more fuel (safc. RTEK2.0, aftermarket whatever management).

 

Thanks for the replies so far.

I got more testing done, checked the various temperature sensors when hot. They all look good still.

I took the filter off the AFM and checked it while slowly pushing it in...

there are many dead spots on the way in, the resistance just drops to zero
and the futher in it gets, the more the readings jump dramatically, they'll go as low as (I think) 60 ohms and then up to 460 then back to 120.. etc.. all over the place.

I'm going to try to get my hands on another s5 turbo afm this week and try swapping it out.

re: safc, rtek, etc.. I agree. Once I get the car actually driving, before I try anything crazy I fully intend to at least get an SAFC or other means of adjusting the fuel considering the mods. I hoped getting that immediately wouldn't be necessary for low and off boost driving.

 

there are many dead spots on the way in, the resistance just drops to zero
and the futher in it gets, the more the readings jump dramatically, they'll go as low as (I think) 60 ohms and then up to 460 then back to 120.. etc.. all over the place.

************************************************** ************************************************** *****************************

That's all normal. Don't look at resistance. Look at the voltage output. I believe it goes from high voltage i.e. 5vdc down to ??? half volt dc when pushed all the way in.

Try disconnecting the 02 sensor and go for a ride and see if the afr is different or not. Some of what your seeing might be normal if the 02 is connected up and the ECU goes into closed loop. Takes but a moment to find out. I'd definetly do this, because I think some of what your seeing is Normal, especially for a engine going into closed loop.

The water temperature sensor for the GAUGE, is next to the oil pressure sensor. Has a yellow/???? wire on it. Yellow/white maybe.

The water thermo sensor for the ECU, should be putting out approx a half volt dc when the engine is fully warmed up. Maybe 4.5 to 5vdc. Anything close to those figures is as normal as apple pie. Even if the wire falls off the sensor, the ECU defaults to 180* i.e. a hot engine.

 

Just an FYI, I very recently had been dealing with a very, very similar problem like what you're currently experiencing. I also got stuck looking down various rabbit-holes, and pursuing unlikely and complicated electrical/ECU/sensor-related problems.

My lean condition was really only during primary injector operation (it was masked by the additional secondary injector fuel), and I was seeing AFR spikes between 16-19 during light/mid cruising and initial throttle. Bogging and hesitation were more pronounced under heavier load (I.E. climbing a hill).

I partially solved my problem about a week ago when I (again, grasping for straws), changed my fuel-filter and ALSO changed my primary fuel-injector mixing-plates/diffusers (things that fit into the center-iron injector "sockets"). The N/A and Turbo mixing-plates/diffusers are of a slightly different design, and I wanted to see if I noticed a difference. After putting everything back together, I noticed that my car was running much better and I simply attributed it to replacing the old fuel-filter. However, there was still the lean-condition registering on the wideband, and it appeared that the "main" problem was still there--just slightly better and not as pronounced.

In frustration I took my primary injectors to get cleaned. I got them back with new grommets/o-rings, installed them and my car ran like CRAP. Terrible. After some thinking, I determined that the new grommets were much more flexible, and not quite as sturdy/thick as the old OEM grommets (which my secondaries were still using). I swapped grommets (back to the old one's) and seated them firmly and securely into the injector sockets and then pushed my injectors into them--to make sure there was no chance of pinching/binding/deformation after bolting the fuel-rails down.

Car ran much, much better. However, after realizing that my problem all along had been sneaky primary injector leaks I did the carb-cleaner test and found that even after careful installation, there was still a vac. leak right at the socket.

I tried several different variations using the new injectors grommets and the old (top) O-rings and boots (which were thicker, putting more pressure down on the lower seal when the rail was bolted down), and I was finally able to come up with a combination that allowed the best possible sealing. It was actually kind of shocking how difficult it was to obtain a seal that wouldn't allow a noticeable vac. leak (translated into hesitations, bogging and lean spikes), and I actually was unable to even use the brand-new grommets/boots that came with my freshly cleaned injectors because they didn't provide a tight enough seal.

Part of the reason for my difficult sealing is that I'm using injector-bleeders/sockets from a N/A center-iron which does not have the little-plastic sleeve/retainer that's sealed with an o-ring at the top and bottom. The N/A one is made entirely of plastic and does not allow "deep" sealing like the OEM turbo socket-retainer/thingy's.

Either way, just wanted to give you some background as to where my insight came from--and to provide a little moral support.

 

I checked the voltage when running, that looked stable. Maybe the AFM is OK then.

I tried driving with and without the ECU O2 sensor line connected, no difference.

 

Thanks for the advice, I'll search for vacuum leaks again and pay extra attention to the injector areas.

But would a vaccuum leak cause it to lean out like crazy when hot but run super rich when cold?

This is why I'm looking hard at the sensors, it seams like there is a distinct change when the car gets to a certain temperature. If I drive around while it's cold it stays rich at all throttle and pulls hard without leaning out. I run out of road.

As soon as it is hot, that changes immediately. And like I said, I can't even get out of the driveway before I see it. I don't have enough road to get the RPMs up slowly by feathering the throttle, to test high RPMs, once the problem is happening. I have tried getting it onto the road and going WOT to see if throttle amount makes a difference, it still goes lean and gets to about 16 right before hitting boost. At which point I let off. I'm not brave enough to go into boost at 16 AFR. I've been working on this car since January 2007 and STILL can't drive it home. This is the 3rd engine (previous two were blown or blew in the driveway somehow).

My fuel pressure is good, I have a gauge on the FPR in the bay, I stuck a camera in the bay with a light to watch it while test driving it. I previously had a fuel pressure problem before getting this far and that was noticeably different. The problem I had in the past was a torn fuel hose *inside* the tank where the walbro connects to the hard line.

It doesn't seam like a physical problem because while the car is cold it runs great. Tell me if I'm wrong.

I also can only troubleshoot it on the weekends so I have until next weekend to come up with ideas.

 

Think about this: These cars run rich when they're cold, and then as soon as they reach a certain temp. (thermosensor?) then the ECU cuts back on the fuel and runs a "normal" fuel-curve. I wonder if this additional "cold" fuel is simply masking the problem--and as soon as the normal/warm fuel-curve starts the problem becomes evident, even though it was basically there the whole time.

When I was dealing with my recent primary injector vac. leaks, it took a couple minutes before the lean condition would start registering on my wideband and the hesitations/bogging started.

How long (in minutes) before the problem becomes noticeable after a completely cold start?

Also, do you have an aftermarket boost/vac. gauge or water-temp gauge? Just little tools that might help get more data. Not sure important, but sometime all the info you can get will help.

Also monitor your wideband at light/mid-throttle (0psi) and take the RPM's from 2500-4500 RPM's to see what happens to the AFR's as the secondaries kick in--and during their operation. You should be running pretty frikkin' rich when they kick in, especially with that Walbro. Again, just a little more information to work with to isolate the issue.