Mitsubishi Eclipse 3G Club banner

1 - 9 of 9 Posts

New daddy
3,352 Posts
Discussion Starter #1
As a preface read this.

Basically what this person is doing is attaching the throttle position wire from the SAFC to a 0 to 5V MAP sensor. I know we do not have an MAP sensor, but 3 bar map sensors with output voltages of 0 to 5v are pretty cheap. Also, wiring one up and attaching it to the manifold shouldn't be too hard. Would this work on a boosted 3g?

New daddy
3,352 Posts
Discussion Starter #2
Taking a look at the FSM, and we do have a MAP, but it is only a 1 bar map. However, I have no idea what the ECU uses it for, because our cars use a MAF. Is anyone that is boosting tripping a P1400 code (Manifold Pressure Differential Pressure Sensor Malfunction)?

The criteria for a P1400 to trip is:
Coolant temparture > 113 F
Intake Air temp > 41 F
Volumetric Efficiency is between 30% and 55%
Sensor output is greater than 4.6 V for 2 seconds (4.6V = 17 psi absolute, or 2.3 psi on your boost gauge)


Coolant temp > 113 F
Intake air temp > 41 F
Volumetric Efficiency is lower than 30%
Sensor voltage is higher than 4.2V for more than 2 seconds. (4.2V=16 PSI absolute or 1.3 psi on your boost gauge)

New daddy
3,352 Posts
Discussion Starter #4 (Edited)
Come on, someone has to have a comment on this. I think I would work pretty good. If the low throttle setting was set for just around 0 psi, you could lean out the larger injectors. Then set the high throttle setting for high boost. Then let the SAFC average out the fuel corrections between the low and high throttle settings.

I think this would also help out when using a boost controller. This would supply the correct amount of fuel during high boost part throttle conditions. Such as going up a hill, which has been known to blow up engines that are running boost controllers.

Also, a hand powered pressure/vacuum pump could be used to calibrate the SAFC throttle position. You could then figure out what manifold vacuum/presure = what throttle position % on the SAFC.

For example:

Pressure \ Reading Voltage \ Throttle %
Idle Vac \ 1.4v \ 26
Zero Vac/Boost \ 2.5v \ 65
7 psi \ 3.45v \ 90
9 psi \ 3.7v \ 100*

Not the Face!
6,616 Posts
:) Been there, done that:

Date: Fri, 26 Apr 2002 03:03:53 -0700 (PDT)
From: Eric S. Lai
To: SoCal-DSM list <[email protected]>
Subject: [SoCal-DSM] S-AFC hack install and initial tuning.

So I did the S-AFC hack tonight. Now instead of throttle position as
input to my S-AFC, manifold pressure is coming in.

3 Bar MAP sensor from RRE reads 0-5V just like throttle. Just need a 5V
power source, a boost source, and then use the output from the MAP sensor
as the input to the S-AFC's throttle line--the GREY wire--rather than the
throttle position line from the ECU. You've all seen the "fool the S-AFC"

+5V can be grabbed from the MAP sensor on the intake manifold.

Still haven't figured out quite where to mount the sensor though.

The pictures tell all you need to know about the installation: Hacked S-AFC.jpg MAP sensor.jpg MAP sensor connector.jpg Manifold MAP sensor.jpg MAP-to-throttle.jpg


T28/28 @ 1.22
Denso 150lph hardwired 14V
550cc injectors
unmolested MAF

Manifold pressure versus reported throttle percentage:

atmosphere: approx 19.5% throttle and 1.5V.

-42 cmHg: 0% throttle.

17-18 psi: approx 65%-70% throttle.

S-AFC throttle position settings:

Before: Lo 31%, Hi 80%

After: Lo: 5%, Hi 88%

S-AFC correction settings:

Low settings before: -12% across the board.

Low settings after: unchanged.

High settings before:

1k: -18%
2k: -19%
3k: -16%
4k: -16%
4.5k: -15%
5k: -13%
6k: -12%
7k: -13%

High settings after: -10% across the board.


Tuning is of course different. You first figure out what your max boost
will be, then add in a bit for margin. This becomes your HI throttle

Low throttle point I'm not sure yet. Obviously, make it less than
whatever is read at atmosphere, in my case and probably all cases, that's
lower than 19%. I set it quite low, not sure yet as to why, but
part-throttle operation doesn't seem to run rich even so.

The S-AFC interpolates the correction curve between the low settings and
high settings based on throttle position between the lo and hi throttle
points. My max boost comes on at around 65-70 percent.

To get .94V across the rev range, I first zero'ed out the S-AFC high map
and set the hi throttle point to 80%. I then pulled down the curve to
about .96V at full throttle. I then started moving the throttle point
higher to get a safety margin. After some tweaking and a bunch of runs I
get consistant .92-.94V across the rev range in all gears.


Well, with such a fast-spooling turbo I never had THAT much of a problem
with bogging when I stomped the throttle. Even so, I can tell that I'm
having less lag. In standard form, tip-in full throttle and you get an
over-rich situation with the S-AFC--turbo isn't spooled. Throttle
position doesn't directly correspond to airflow with a turbocharged car.
The effect is more pronounced if you have a lot of lag.

I suggest that some of you big bad turbo boys try this out. Kind of funny
to never see full throttle on your S-AFC, haha.

Will keep you updated on the tuning. This is just preliminary. Got the
typical crappy 17 degrees at top of third, 840 degrees, .94V, so I'm happy
for now. Don't know what's gonna happen to LTFT yet--low throttle
probably needs tweaking.

Thanks Mike for the sensor and discussion!

Off to bed, see y'all tomorrow.

Not the Face!
6,616 Posts
Keep in mind that the above post is on a 2G GS-T and the values specified are not a guideline for how you should be programming your own S-AFC on your own car.

The sensor used is a GM 3-bar that RRE carries. It's a lot cheaper than the more accurate ones that you'll see out there going for near three-digit numbers and does the job well enough for this particular application.

New daddy
3,352 Posts
Discussion Starter #8
eslai said:
Keep in mind that the above post is on a 2G GS-T and the values specified are not a guideline for how you should be programming your own S-AFC on your own car.
Agreed very much so, but it is nice to know pressure vs voltage output of the sensor, with out using a vacuum/presure hand pump to figure it out.

Not the Face!
6,616 Posts
Even so, you will need to calibrate your own MAP sensor versus your own S-AFC, there's no guarantees for this sort of thing unless you have a spec sheet in front of you, and if you use a different kind of MAP sensor, definitely, you'll want to figure it out on your own.

In the end I was happy with the mod, but my car wasn't so burly that it was all that necessary. This was shortly before I went to the EMS.
1 - 9 of 9 Posts