The Central Hub for DSM Community and Information

For 1990-1999 Mitsubishi Eclipse, Eagle Talon, Plymouth Laser, and Galant VR-4 Owners. This is where the DSM platform history is documented and archived. Log in to help us in our mission, and to remove most ads from the browsing experience.

SAFC/O2 Problem

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

MyBlack94GST

15+ Year Contributor
1,352
3
Nov 27, 2003
Richmond/Berkeley, California
Hey folks, just put in some 550cc injectors and an SAFC. Problem is, I just put in a new O2 sensor as well, and it won't cycle at idle unless I lean it out to -38% or lower at the 1000rpm mark. This obviously creates problems when I try to start from a stop at idle (car stumbles). Did they give me injectors that were way oversized? What's going on?
Thanks in advance.
 
I don't know if this will help you but it takes a little while for the o2 sensor to warm up before it gives accurate readings.
 
Check the injectors to see if they are tight enough and not leaking. I had a similer problem and it felt better when i leaned it out. Idled real bad, took it to the dyno and one of the injectors was leaking when reving. (only when getting on it)
 
I have a Walbro 190 fuel pump, I should buy a fuel pressure gauge eh?

Update: After a couple days of playing with it, I have determined that:

1) The (brand new Bosch) O2 sensor does not cycle at idle, but does in gear. At idle it stays pegged rich... like .91V.
2) My Low fuel trim updates anyway... I'm at 91.2% and leaned it out 2 more % after the trim changed to this number.

I removed my lower honeycomb and backed the screw out flush the other day.
My injectors are not leaking AFAIK. I'm leaning towards a fuel pressure problem, but would like to hear from people who had and solved a similar problem. Thanks for all your help, guys. :D
 
Hi Mahesh,

A few things come to mind here.

Check the FPR vacuum hose for leaks. Is your boost gage plugged into the FRP vacuum line? If that line is leaking badly it could raise your idle fuel pressure by as much as 10 psi, making 550's flow like 640's. Coincidentally, the AFC correction for 650cc injectors is about -38%. Maybe try running a new hose straight from the nipple on the intake manifold to the FPR, see if that helps.

Also check for any error codes on the logger. If your getting an Airflow Sensor error the ECU will run the car at about .90 volts on the O2 sensor until it gets a good airflow signal. Are the O2's cycling at idle?

What are the airflow readings at idle? With your slightly modified 1G MAF the SAFC should show around 28 to 35 Hz. while the logger should show about 35 to 42 Hz if all is working properly.

Let me know how it goes. I work in downtown S.F. so I could easily hop over the bridge and meet you somewhere for a tuning session.
 
OK! I bypassed the stock, tee'd for boost gauge, FPR line with a new piece of vacuum hose, and no changes. I'm not throwing any codes, and the readings were as follows:

Idling at 750-800rpm
SAFC Karman = 27/28 Hz
Logger MAF = 25.1 Hz
Logger O2 = .93V

My low fuel trim is at 81.2% now after a couple days of driving.

Dave, I'll be working in Mountain View (by Shoreline) starting Monday, so you won't have to hop over any bridges (you live there right?), but yeah, I need help :cry:


pneumo said:
Hi Mahesh,

A few things come to mind here.

Check the FPR vacuum hose for leaks. Is your boost gage plugged into the FRP vacuum line? If that line is leaking badly it could raise your idle fuel pressure by as much as 10 psi, making 550's flow like 640's. Coincidentally, the AFC correction for 650cc injectors is about -38%. Maybe try running a new hose straight from the nipple on the intake manifold to the FPR, see if that helps.

Also check for any error codes on the logger. If your getting an Airflow Sensor error the ECU will run the car at about .90 volts on the O2 sensor until it gets a good airflow signal. Are the O2's cycling at idle?

What are the airflow readings at idle? With your slightly modified 1G MAF the SAFC should show around 28 to 35 Hz. while the logger should show about 35 to 42 Hz if all is working properly.

Let me know how it goes. I work in downtown S.F. so I could easily hop over the bridge and meet you somewhere for a tuning session.
 
Yes, I'm in Mtn View, very close to Shoreline.

Looking at your Hz readings, they do look about right for the ammount of correction you're using on the AFC. So we know the MAS, AFC, and logger are working correctly, more or less.

Have you changed the caps in the ECU, or do you know when they were changed?
 
Our Mitsu supplier has steered us away from using Bosch O2 sensors on DSM's. They seem to be very flaky on our cars. They have caused numerous problems. I'm not bashing bosch here, just speaking from personal experience, as well as what I've seen on other DSM's running Bosch O2 sensors.

With Bosch O2 Sensors I've seen all sorts of running rich or running lean what's my problem questions....

Good Luck.
 
I'm not at -38% anymore... leaning it out that far doesn't seem to make a difference anymore at idle (i.e. it just stays at the top of the voltage range no matter what).
I'm at -14% correction now at idle, this is what has given me the 81.2% Low Fuel Trim after a couple ECU resets. It was at ~91% and -12% initially and I leaned it out a couple % on the SAFC (to -14%), but it seems to have done no good. Do the Hz readings still look correct?

I don't know when the caps were last changed on the ECU, and I haven't changed them myself. Thing is, everything was great (cycling, 100% fuel trims) before this new setup, so I doubt it's an ECU problem.

As far as the Bosch sensor goes, if worse comes to worst I will try to clean off the old sensor with a blowtorch or something, and reuse it. The reason I replaced the old one was because it was doing what this one is doing... staying at one high voltage. That reminds me - I had to lengthen the wiring on the O2 sensor to reach, because I put in a 2G O2 housing. I used stock wiring from the old O2 sensor, twist+solder. Could this be causing my problems?

And Dave, I'm moving in to my South Bay place on Saturday... so I'll be bugging you for help soon thereafter :D Thanks man.
 
MyBlack94GST said:
I'm not at -38% anymore... leaning it out that far doesn't seem to make a difference anymore at idle (i.e. it just stays at the top of the voltage range no matter what).
I'm at -14% correction now at idle, this is what has given me the 81.2% Low Fuel Trim after a couple ECU resets. It was at ~91% and -12% initially and I leaned it out a couple % on the SAFC (to -14%), but it seems to have done no good. Do the Hz readings still look correct?

Try adjusting the AFC at idle while watching the Hz on the logger to see if it changes. 27Hz on the SAFC and 25 Hz on the logger with -38% correction doesn't seem right. The logger should show less Hz. It sounds like the ECU decided that the airflow sensor was bad, ignored it, and ran the car off default maps/ limp mode, which makes it idle very rich. I've seen this when the Hz readings get around 16Hz and below, but your car is a '94, while mine is a '90, so they may have different limits where limp mode starts. Also, limp mode isn't permanent. As soon as the MAF gives good readings the ECU will use them and go into closed loop mode.

Maybe your fuel trims are behaving differently because the injectors and SAFC correction are putting the ECU into limp mode quicker so the ECU doesn't bother to try to update the trims?

MyBlack94GST said:
...Thing is, everything was great (cycling, 100% fuel trims) before this new setup, so I doubt it's an ECU problem.

As far as the Bosch sensor goes.... The reason I replaced the old one was because it was doing what this one is doing... staying at one high voltage.

It's also possible the O2 sensor was reading correctly, but something else is going wrong.

MyBlack94GST said:
That reminds me - I had to lengthen the wiring on the O2 sensor to reach, because I put in a 2G O2 housing. I used stock wiring from the old O2 sensor, twist+solder. Could this be causing my problems?

Not unless the wires are touching each other where you spliced them together. Did you wrap each wire individually to insulate it? There's plenty of heat and vibration in the engine bay.

I can think of several little things that might cause your cars problems. Maybe the two black ground wires for the SAFC were spliced in at the wrong place? Or the settings on the SAFC are off somewhere? Or the ground strap is missing/ bad on the downpipe? Or you have a boost/intake leak?

MyBlack94GST said:
And Dave, I'm moving in to my South Bay place on Saturday... so I'll be bugging you for help soon thereafter :D Thanks man.

OK my curiosity is piqued. I gotta see this car!
 
I have all honeycombs in and idle at 800 w/35hz. This is w/680cc inj and SAFC set at -16 at 1K.
 
For some reason I checked all this stuff today, LOL...

*The numbers above, 27/25 etc were at -14% correction, not -38%.

The logger seems to respond (Hz) to changes on the SAFC, i.e. more correction = more airflow.

I cleaned my ground strap today and tested it for continuity with a testlight while it was on the downpipe (it checked out OK).

I insulated the lengthened O2 sensor wires separately, they're not even touching.

I assume I wired in the SAFC correctly with respect to the grounds, whatever one was supposed to be closer to the connector is (brown wire vs. black?), and it seems to be working, just for kicks I took out 50% fuel at 1000 RPM and it acts like it should (stutterbog)...

And same problem! I reset the ECU and find that I can still cause the O2 sensor to cycle at idle if I take out a LOT of fuel and blip the throttle to persuade it.

The only thing I haven't checked for is a boost leak. This will happen shortly.

I love cars!!!!!
 
How does it run now with the AFC at -14%? That should be very close to the right setting for your setup if everything was working right.
 
Crappy, seems like the low fuel trim always ends up at 81.2% no matter what I do...

Finally got internet at work!
 
Does the low trim drop to 81.2 just from sitting at idle? Or does it end up there after driving around for a while?

What are your SAFC settings, hi, lo, and throttle points? I just wanted to check them to see if maybe it's running rich at other ranges, which will drop the lo trim, too. With your modded MAF the low settings should look like a curve. For example, if it's at -14% at 1k rpm, then 2k should be even lower, perhaps -18%. Then 3k should be slightly lower than that, maybe -20%, then 4k and up might be something around -21%.
 
Add Value - Be Respectful - No Trolling - No Misinformation - Participate Often!
Support Vendors who Support the DSM Community

Build Thread Updates

Latest Classifieds

Back
Top