Hey- the last post I did like this yesterday messed up, and didn't have any content! Damn server busy every 10 seconds!!!!!

Anyways, a lot of people here have had problems with their S-AFC's running properly. They would run rough, and run horrible when cold!
I tried a bunch of things and got mixed results, but it still ran like *** when cold. Anyways, here's what I did.

- I added a 2nd, heavier ground from the engine (under intake manifold) to a seperate bolt on the firewall

-I ran a heavy gauge wire from that firewall location up to the MAF and pressure sensor., I grounded the brown wires w/black stripe on both sensors together, and to the firewall wire.

-I atttached a heavy ground to the wire from the ECU with the double grounds on it for the S-AFC, (about 3 inches away from those) to a chassis bolt near the ECU.

I did all these at the same time, so I'm not sure if any one could have solved it, but my S-AFC now works perfectly, even right after a cold start! The car seems to run a lot more well mannered now.

It's still a little funky during closed loop though, (not sure why) so I just unplugged the o2 sensor cable under the hood- it runs teriffic now, and I havn't noticed any loss in economy yet.. though logically I'm sure it's gone down a tad.
If I feel like it I am just going to install a multi prong pushbutton in the console to turn the o2 sensor on/AFC off for highway cruising economy.

Hope this helps somebody! (It sure did for me!!)

 

b u m p

 

*Electronics Idiot!*
When you say you grounded the brown wire w/ white stripes, do you mean you actually connected another wire to the brown wire, then attached it to the firewall?

Thanks

 

how would this help with fuel economy??does'nt the car run on the rich side at lower rpm anyway's[stock],so fuel economy would be better with a s-afc at crusing..
hmm
david

 

Umm, it's the brown w/black stripes
Anyways, ya, ground always means a wire to the chassis or sometimes, battery negative. I ran a wire from those, to the chassis.

The car will run rich at lower rpms, but only if it's not in closed loop. (idle and more than 1/3 throttle)
The ECU goes into closed loop at light cruise, using feedback from the o2 sensor. For some reason, the ECU doesn't like the AFC signal while under closed loop. It runs a little rough.
So, with the o2 sensor disconnected, it will run richer than it would in closed loop. I can try to compensate with the AFC by leaning it out a lot at LO throttle, but it still won't be the same as the feedback system.

 

sweet finally!

this is excellent, everyone should just do these grounds anyways, the FC's are notorious for bad grounds.

 

Im thier. Thnx bambam. Also after u install this make sure to reset the ecu. This helped one persons problem.

 

Thanks!

 

VON !!!

What do you mean reset the ECU???

I just installed my S-AFC yesterday -- This morning at cold/start and low rpm's it ran rough. I will investigate grounding issues.

LOUD THANKS TO BAMBAM7 -- for clarifing installation issues. [I have searched fourms and found no concice information]

BAM BAM -- QUESTION: How did you come to the conclustion of "Lean the midrange down (from 2000-5000) by about -7 percent" stated in your other posting??
(Tom on the West Coast - We spoke about S-AFC and stereo installation)

I thought that I would respond to this thread- But I am going to direct that question in the other thread.

Thanks TJ

 

umm, stupid question over here too: What does S-AFC mean? (Im not completely ignorant, that just happens to be one of the acronyms I never heard about) ...what is it and what does it do?

...oh, and also, I took my car to the dealership recently and they took a look at it and the diagnosis mentions "at times ECU is not grounding" and also mentions "ECU has grounds where it should, except at one terminal" ....Now is there anywhere I could look up some well documented info on this? With pictures and instructions on how to check the grounds or add ground to the ECU?

...Has anyone made any photographed documentation on how to do all these grounds? (from the ECU to the engine to everything else that would need it)

...Just wondering.

 

It doesn't really man anything - it's what A'PEXi called their fuel computer.

If you REALLY want an answer, it stands for "Simple/Super Air Flow Converter"...



-Ted

 

The stock software oon the ECU commands a very rich mixture through the midrange...
Through my own testing (butt dyno vs gtech vs o2 sensor output voltage) I have definately noticed that leaning the midrange gives you the best increase.

The rotary runs best at around 13:1 through the midrange, steadily going to about 14:1 at redline.

I have mine at about -9 through the midrange, a small spike up to -5 at 6000, then -12 from 7-8000.
This keeps the most consistant ratio curve.
I tune it by flooring the gas at any rpm, and ALSO stepping on the brake, enough to KEEP it at the RPM, (whith the brake and gas floored) then I look at my ratio, and adjust it to what I want.
IMO, works as well as a dyno... at least only if you know what mixture your car needs.

Also- the stock programming will run richer and richer the more you open your airflow up... the more modded your engine is the more you will need to lean it out.

You may gain 20 hp in the midrange- but you will only get about 10 max at redline. Makes for a much more useable tourque curve through.

 

I would like to try this, but how (with my multimeter) can I confirm that I am grounding the right wires? Mine will rev/drive fine under full throttle, but slow reving/ steady throttle runs real rough, it also idles crappy/rough untill fully warmed up.

 

That's exactly how my car was running.
The biggie for me was the grounding of the pressure sensor and the MAF sensor.
The sensor ones you want to ground are both brown with a black stripe. There are two on the MAF (one for the actual MAF and one for the temp sensor in it) and one on the pressure sensor.
They are at the plugs right at the sensors.

I know these did it because a few days after, it suddenly started running like crap again. One of my splices their had broken. Fixed that- and it's fine.

Of course- also do the ECU and the engine-firewall one.

No need for a multimeter- but if you want to know...
set if for volts... put the neg probe on the chassis and the pos probe on the wire... if you see voltage- it's the wrong one!!! (more than .01 volts...)

 

I'd like to say that the comment of "steadily going to 14:1 at redline" is a little off. Thats way to lean for any rotary. it actually needs to drop to 12:1 at redline, if not richer. 14:1 is near stoic and is too lean except at idle or light cruising in closed loop it might get near 14:1, but not at redline. That's a engine replacement in the making if your hitting 14:1 at redline.

Tim Benton

 

Interesting.....my n/a FC just made peak hp at about 14.25 on the dyno at around 7,500. CJ

 

I don't know why (someome else might), but Mazda's TSB on the boost sensor specifically said NOT to ground its earth wire to the chassis, but to take it to one of the bolts on the thermostat housing. Like I said, dunno why, but Mazda knows best!

 

If you run any turbo engine at these mixtures, it will be dead very soon!
Do a search, ther's plenty of threads telling you the best mixtures at different loads. Do it soon!

 

Ha ha, more crazy "Japlish"! Apexi's original fuel controller was called AFC (air fuel converter), it's new one is the S-AFC (super air flow converter!). It's an intercepter type fuel controller. It is wired in between the airflow meter and the ECU, intercepting the AFM's signal and modifying it according to your tuning. Essentially it lies to the ECU, making it think more or less air is entering the engine than actually is, so you can tune the fuel mixtures to suit your cars modifications.

 

About BROWN AND BLACK WIRES. See the attached jpg and look where I marked the brown/black wires. They all join together with solder splices and enter the ECU at pin 2C. Think now. If the worker who made the harness had a bad day, and made one or more bad solder joints (cold solder joints) for the brown/black wires, then you can see how even though the *factory ground* may be good for the harness, it will do no good for that unit that has the cold solder joint. Thats why sometimes it takes more than an additional wire added to the boost sensor to correct a problem. I distinctly remember Six Rotors writing and saying that in addition to the extra ground at the boost sensor, he had to add an additional one to the tps. That corrected his problem. Its the multiple solder joints on this so called brown/black wire that causes problems. That said I've a 86 that has no problem, none, nadda. But I've a 87 n/a that was a bitch. No, BITCH. Because of bad solder joints. And I don't mean the factory grounds, I mean the multiple solder joints in the harness.

 

RE: 13.1 -14.1 mixtures

Hell ya- These are mixtures for tuned-to-the-edge N/A's made for power.
I wouldn't dream of runnning a turbo at 14:1 at 7000 rpms!!!!

That's oddd about grounding to thermostat housing....
It doesn't make any sense to me... The engine is grounded to the chassis anyways- and usually, if the ground is bad, the engine runs like crap, and gets 3800 rpm hesitation!
My system worked perfectly- you can't argue with results!

 

I just grounded out the afm and press sensor, with no changes I havent tried the ecu yet cuz I dont currently have my tools to get the plate outta the way.

 

Yeah, mine runs the same way. Under light throttle the car has a slight hessitation. I can see my Haltuner light zipping back and forth. So I know she is in closed loop mode.

It is SOOO bad when cold. But as soon as I give it some gas the ecu picks a mixture to run and it smooths out. Murder she wrote. Already did the Boost sensor ground. Will do others. Thanks

James

 

Ya- I disconnected my o2 sensor- so it's always in open loop- it eliminated all light throttle roughness/hesitation.
Funny thing is- I just plugged it back in yesterday- and it runs prefectly fine- even in closed loop!!