ARC2 Tuning Tips

by Jeff Lucius

The ARC2-GP kit is manufactured by Split Second and GT PRO Performance Tuning for use on the Mitsubishi 3000GT VR4 and the Dodge Stealth R/T Twin Turbo. While the ARC2 can be used with the stock fuel system, it is specifically designed to be used with larger-than-stock fuel injectors and to replace the Karman-vortex type mass air sensor (MAS) that Mitsubishi uses. Split Second also manufactures and sells ARC2 controllers for other MAS's and to replace carbureted systems. The ARC2-GP kit consists of a Ford 80-mm hot-wire MAS (see 2-arc2mas80mm.htm) or a Granatelli 87-mm hot-wire MAS (see 2-arc2mas87mm.htm), a K&N 900 cfm 5" x 9" tapered-cone air filter (part number RE-0870), a wiring harness, the Air/Fuel Ratio Calibrator (ARC2 control unit) that is designed to output a Karman MAS frequency signal (square wave), the Split Second ARM1 A/F meter, and instruction sheets. The tips below will help you get the best performance out of your engine using the ARC2.


I recommend starting a record keeping system that lets you note the day and weather conditions, the 4 knob settings, any mods done, and the results of your changes (ARM1 lights showing rich or lean, engine performance, poor idle, etc.). I have a 8.5x11 sheet of paper in the glove box next to the ARC2 with just this info on it. Write down the knob settings before you start to adjust them.

ARC2 & ARM1 in glovebox

Here's how the ARC2 works. The ARC2 MAS has replaced the stock MAS and its air temp and pressure sensors. The standard ARC2 MAS is the Ford "F8LF" MAS used in 1998 4.6-L engines using 19 or 24 lb/hr injectors such as found in the Mustang GT and Cobra, Expedition, and F-series trucks. A larger 87-mm MAS, built by Granatelli, is also offered. The ARC2 hot-wire type MAS replaces the stock Karman-vortex type MAS and uses a control unit to convert the analog voltage signal (representing mass air flow) to the square-wave voltage signal (Karman frequency counts that represent volume air flow) that the Mitsubishi factory engine control unit wants to see. The ARC2 control unit sends fixed or constant signals to the ECM in place of the stock MAS's air temperature and barometric pressure variable signals. This permits easier calibration and translation of the ARC2 MAS signal, and allows consistant manipulation of that signal by the factory ECM.

When I was using the 80-mm Ford MAS, my ARC2 control unit produced an air temperature signal of 1.99 v, corresponding to 95F, and a barometric pressure signal of 3.22 v, corresponding to 0.80 bar. After switching to the 87-mm Granatelli MAS, the air temp signal changed to 2.01 v (94F) and the baro signal to 3.24 v (0.81 bar). Because the hot-wire MAS's do not produce any air temp or baro signals, that is, the ARC2 control unit produces these signals, I am not sure why there was any change. When I talked to Split Second in March of 2000, Mark told me that air temp should have been set at 68F (2.74 kOhm) and barometric pressure set to 1.00 bar (4.0 v). I do not know why my ARC2 unit sends different signals. A possible consequence of a higher air temp (94F vs 68F) is that the ECU will use a slightly lower timing advance. Considering that air temp in the plenum for an engine with stock intercoolers and turbos is likely near or above 95F, this should be of little practical effect as far as timing is concerned. However, the air temp and baro pressure values, if different than mine, may affect the settings of your ARC2 compared to mine.

With the factory MAS, the ECM uses the Karman volume air flow signal along with the air temperature and barometric pressure information to calculate the mass air flow. With the ARC2 MAS, the mass air flow is measured directly by the "hot-wire" method. In a hot-wire MAS, a wire or film structure is electrically heated and incorporated into a Wheatstone Bridge, which is balanced for no air flow. Air flow through the MAS removes heat from the element. The mass of the air flow is proportional to the heat lost, which unbalances the bridge producing a voltage differential to an amplifier, which in turn varies the resistance in the bridge in an effort to maintain a constant temperature of the heated element. The output voltage of the amplifier is proportional, in a slightly non-linear manner, to the mass air flow. For more information on how the MAS and other engine sensors are used to control fuel injection, take a look at my web page 2-fuelinjection.htm.]

The 4 dials on the control unit (LOW, MID, HIGH, and ACCEL) adjust the new MAS air flow signal that the ECM sees. There are 2 versions of the ARC2. I have the earlier version. The second version is nicer and the preferred one. The biggest difference is in the last knob to the right, the ACCEL adjustment. The ARC2 "classic" has the "0" position straight down at "6-o'clock". Version 2 has the "0" position straight up or at "12-o'clock". There also may be a difference in the way the MID adjustment is handled. Be sure you know which version you have. The adjustments are basically the same, but the dial positions may be different than what I describe for my "classic" version.

Before you start adjusting dials, you need to know how the ARM1 A/F meter works as it is probably your only "tuning" tool if you do not have a datalogger and EGT gauges. The ARM1 monitors, conditions, and then displays the electrical voltages sent by an oxygen (O2) sensor. For the turbocharged engines, there is an O2 sensor for each cylinder bank in the exhaust housing just past each turbocharger. I at first connected the ARM1 to the ECM wire for the rear cylinder bank O2 sensor. A better setup would be to install a switch to allow the ARM1 to monitor either O2 sensor. Later, I installed a second ARM1 meter so that I can monitor both O2 sensors at the same time. ARM1 installation instructions: 2-arm1instal.htm.

At cold engine start-up the ARM1 will read rich (blue lights) then probably switch to lean (red lights) and slowly move toward the rich end. As the engine warms the lights will quickly start to cycle (green and yellow lights). This is absolutely normal, as the ECM is using the O2 sensor voltage in a closed-loop feedback to increase a little then decrease a little the fuel injector pulse width in an attempt to achieve the ideal 14.7:1 mixing ratio. As the engine RPM slowly increase, the cycling rate of the lights will also increase. Compare at idle and at steady 3000 RPM with the car parked. If the engine RPM increase rapidly (car stationary) or the car is accelerating, the A/F will indicate a rich condition (blue lights). When the throttle is closed rapidly, the A/F will show lean (red lights or no lights) because the injectors are being shut off by the ECM. At low-load cruising and light acceleration, the lights should cycle. For more details by Split Second on the ARM1 click here.

Animated ARM1

The ARC2 works by changing the MAS airflow signal sent to the ECM. In a general sense, turning a knob clockwise "richens" the A/F mixture (look at your ARM1). When I went from 360 to 550 cc/min injectors I had to turn the HIGH down or counterclockwise (LOW, MID and ACCEL basically stayed the same), which tells the ECM that there is less air. This is because the 550s squirt more fuel than the stock 360s do, so we tell the ECM there is less air and it actuates the injectors less time.

Split Second has some instructions for the initial setup at this link. This Split Second web page also shows how the different dials are supposed to affect the gain of the input signal (from the MAS to the ECM). While the ARC2 on this is not the same model as we use, the way the adjustment knobs work is the same for all ARC2 models.

LOW shifts the the entire range of voltage up or down the same amount. However, a 0.5 volt shift at 1 volt represents a 50% change; but a 0.5 v shift at 4 volts represents only a 12.5% change. So LOW affects low voltage levels (or low air flow levels) more than it affects high air flow levels, at least percentage-wise.

The effect of MID is nonlinear, meaning it changes the voltage most in the mid-range of voltages (mid air flow) and almost none at the lowest and highest voltage levels.

HIGH shifts the entire voltage range by the same percentage. So in that sense it affects all flow levels the exact same.

So now we can see why LOW and MID do not change much regardless of the injector size used.

LOW affects mostly idle and low engine load (low air flow levels). The compensation for larger injectors does not change much here.

MID is a nonlinear adjustment. It's most likely usefulness is in compensating for the nonlinear characteristics of the Ford MAS; or if the Ford MAS is linear, in creating a nonlinear response to replicate the known nonlinearity of the Mitsu Karman MAS (see 2-mas_liter-per-hz.htm); or some combination of both.

HIGH is what is important when changing injector sizes. For example, when we go from factory 360 cc/min injectors to 550 cc/min injectors, we must decrease air flow response at all levels by 35% to fool the ECU into thinking that 360 injectors are still being used (in a manner of speaking).

Here is what is important to remember when adjusting the ARC2.

1. The effects of the adjustments (of turning a knob) is not exactly documented. In addition, each knob affects the adjustment effect of each other knob.

2. The LOW, MID, and HIGH do not represent low, mid and high air flow levels; especially not in the sense that they would on a S-AFCII or SAFR.

3. Because of installation peculiarities, the non-linear effects of the ARC2, the natural nonlinear response of the Ford and or Karman MAS, and the known compensating ability of the ECU in closed-loop mode, adjusting the ARC2 is an iterative process. In addition, the ARC2 (and even the S-AFCII and others) may need to be adjusted if driving habits change. This means we may want to lean the A/F a little when cruising on the highway and to richen the A/F when drag racing at high boost to reduce detonation. The HIGH adjustment does this for us on the ARC2.

Now let's get on to some practical adjustment.

The ARM1 must cycle at idle (similar to what the animated GIF image of the ARM1 above shows). LOW, MID, and ACCEL affect this. Do not adjust the HIGH. For me, moving ACCEL past 1 click always ended up too rich, and 0 was too lean. I have ACCEL at 1 click for either 360 or 550 cc/min injectors. I found that the MID needs to stay at about -7 or -8 clicks regardless of whether I had 360 or 550 injectors installed. The LOW is probably the only knob you want to mess for idle with once MID is set correctly.

Note that the Mitsubishi ECM has the ability to compensate up to a point for fuel mixtures in closed loop mode using "fuel trims". These values can be seen using a datalogger. So if you set the LOW at a certain spot and the car idles OK to you, if the mixture isn't right where the ECM wants it, the ECM will change injector "on time", essentially changing your ARC2 setting. What this means is that you will have a trial and error period as you get a feel for how the ECM is compensating for your ARC2 adjustments. It might be helpful to remove power from the ECU (disconnect the negative battery terminal for 20 seconds or so) to reset the fuel trim values before you start making adjustments to the ARC2. If you have a 1991-1993 3000GT/Stealth (with OBDI connectors) and really want to tune the car, get the TMO datalogger, DTI PocketLOGGER, MirageCorp Datalogger, or MMCd Datalogger. For 1994-1995 models MirageCorp has a datalogger. OBDII dataloggers can be used for 1996-1999 models.

The ARM1 should cycle at cruise and very light load. MID and some HIGH affect this. Again, I found that the MID must remain at -7 or -8 clicks for my model. Do not adjust the LOW or ACCEL. If either MID or HIGH is adjusted too high (clockwise) then the ARM1 will always show blue lights. Adjust the HIGH back until the lights cycle when cruising. It is OK though if they kind of "hang out" in the blue area as this is slightly rich, but gas mileage may suffer a little. If the lights "hang out" in the red area it could be bad because the engine is running lean, but fuel mileage may be better. Remember that the ECM will learn your ARC2 settings and try to adjust the fuel trims till a 14.7:1 ratio is achieved during closed-loop operation (cruising).

The ARM1 must be into the blue lights (rich) at any medium to heavy load (accelerating on flat ground or steady speed uphill). HIGH and some MID affect this. Again, I found that the MID must remain at -7 or -8 clicks for my model. Do not adjust the LOW or ACCEL. Now you will experience the limitation of the ARM1 and the stock O2 sensors. The blue lights just tell us that the mixture is rich (and safe) but not whether it is 13:1 (good rich) or 11:1 (too rich). Best power for turbocharged cars is often achieved with A/F in the 13.2:1 to 11.5:1 range (~12.5 is a good target). You will have to tune by feel. If the HIGH is too far clockwise (too rich) the engine will bog. If HIGH is too far counterclockwise (lean) then the ARM1 will either cycle or (very bad) go to the red lights. This is where an EGT gauge can be very helpful. I notice instant changes in exhaust gas temps as I adjust the HIGH even though the ARM1 shows constant blue lights. Fortunately you do not have to worry about the ECM, as it cannot compensate while in open-loop mode using fuel trims. It doesn't use the O2 sensors and instead uses internal fuel and timing maps. But that is of course why you have the ARC2!

The ECM maintains values that are used to compensate fuel injector activation duration during closed-loop mode (when O2 sensor feedback is used to maintain a 14.7:1 air/fuel ratio). These values are called fuel trims. The fuel trims function a little differently for OBDI (1991-1993) and "hybrid" (1994-1995), and OBDII (1996+) models. For 1991-1995 models (OBDI and "hybrid") there are fuel trims for low engine load (such as idle), moderate engine load (such as lower rpm cruising), and higher engine load (such as higher rpm cruising and light acceleration). The ECM actually selects Low, Medium, and High loads based on air flow "Hz" from the Karman sensor. For OBDI and hybrid models, the ECM does not use these values to compensate the internal fuel maps used for open-loop mode (when the O2 sensors are not used to determine A/F ratio). The ECM is constantly monitoring closed loop A/F ratios and will quickly (within seconds) adjust fuel trims to maintain the desired air-fuel mixture. For OBDII models there are short term and long term trims, and the long term fuel trim does affect the open-loop values determined from the fuel maps. For OBDI models the fuel trims are reset to 100% whenever backup power is removed from the ECM, such as when the negative battery is cable is removed from the battery for longer than 10 seconds, but not when the ignition switch is turned off. After a reset, the ECM will immediately begin to adjust the fuel trims. Fuel trims are not updated when coolant is below 169F.

For OBDI models (well at least for my 1992 Stealth TT) fuel trim values can be as low as 80% and as high as 140% when using the TMO datalogger. Fuel trim ranges may be different using other dataloggers. A value less than 100% means the ECM is reducing injector activation time (reducing fuel) in order to achieve the desired 14.7:1 A/F ratio. A value higher than 100% means the ECM is adding fuel. If you are datalogging when adjusting the ARC2 control unit during idle, you should see an almost immediate change in the low fuel trim value. For example, turning the LOW knob clockwise richens the air-fuel mixture and the Low Fuel Trim should reduce.

The situation I have encountered using the ARC2 with 550 cc/min injectors is that one position of the LOW knob will have the Low Fuel Trim near the maximum, for example, and adjusting the knob 1 click clockwise will send the Low Fuel Trim down to the minimum. I have tried adjusting fuel pressure a few psi one way or the other to change effective injector size a few percent to give some "fine tuning" control, but to no effect. Currently the engine runs great, starts easily, and idles well. Both O2 sensors cycle for closed loop operation (indicating the A/F is near 14.7), and O2 readings are 0.94 to 0.96 volts during WOT. However, the Low Fuel Trim is near 85% (suggesting that I have the ARC2 LOW adjusted too high) and the Mid and High Fuel Trims are near 80% (suggesting that the ARC2 MID and HIGH also are adjusted too high). While theoretically all fuel trims should be somewhat close to 100%, practically it appears that the engine can run well (and pass emissions tests) with one or more trims near the range limits. However, the fuel mixture is definitely rich; EGT at WOT is barely higher than 800F and the tail pipes are black on the inside; and I still get knock counts in the mid to high teens. I will be adding an A'PEXi SAFC-II and WI to solve these problems.

Here are settings that have worked for me here in CO (5500 ft ASL) and with the particular air temperature and barometric pressure constants my ARC2 unit sends the ECM.

For 380 cc/min injectors (cleaned and balanced stock) and 15G turbos and Ford 80-mm MAS.
LOW   +4%   (+2 clks or "1:00")
MID   -16%   (-8 clks or "8:00")
HIGH   +12%   (+6 clks or "3:00")
ACCEL   +2%   (+1 clk) ---> 1st model!

For Denso 550 cc/min injectors and 15G turbos and Granatelli 87-mm MAS.
LOW   +4%   (+2 clks or "1:00")
MID   -14%   (-7 clks or "8:30")
HIGH   -2%   (-1 clk or "11:30")
ACCEL   +2%   (+1 clk) ---> 1st model!

The LOW and MID will vary from car to car (and depend on what air temp and baro numbers the ARC2 sends the ECU), but for a particular car the LOW and MID will stay pretty much the same regardless of injector size, once "tuned in". The HIGH changes with injectors size. With 720 cc/min injectors, I have read that the HIGH should be set near -6 clicks (-12% or "9:00"). The ACCEL setting depends on the ARC2 version you have. Later versions use ACCEL to fine tune the LOW setting.

If you have comments about this web page or have additional tips you would like me to include here, then please send me an email to jlucius at stealth316 dot com.

Back Home Forward

Except for the small gif and jpg images, the content, images, photographs, text, and multimedia displayed are Copyright ©2000-2006 by Jeff Lucius and K2 Software. All rights reserved. No part, section, image, photo, article, or whole of this site may be reposted or redisplayed without permission of the author.
Page last updated May 2, 2006.