F/IC Installation Notes: Honda Civic 06-07

[Blackmagic88]

When the FIC is plugged in (using harness from boomslang) the MAF sensor doesn't recieve voltage. When the factory ecu is plugged in without the FIC the MAF works fine. Would the problem be rooted in the software of the FIC or in the PnP harness?

Yes, either of those things could cause the stock ECU to throw a code for "low MAF voltage" or something like that.

It will be easier to check your software configuration than to check the wiring:
Open the FIC software.
Open a FIC calibration file.
Connect to the FIC with your laptop; the software may detect a "cal file mismatch,"  be sure to click the [Get From FIC] button.
Open the Gauges window.
Turn the key on (don't start the car yet), look for the 'MAF in'  Volts. If this is connected to the stock MAF sensor, it should be about 0.4-0.8V when the engine is not running. Compare this to the 'MAF out' Volts. They should be the same number.


Please check that and let us know how it goes.

I'm trying to figure out what i need to do to help narrow down the issue. Do i need to do the update on the fi/c (version 2.01) even if it has nothing to do with the MAF problem? I dont have any calibration files to even open, where can i get one? 

[Lukish]

Guys from AEM,

If you have done successfully two Civics why don't you give a clear message to the community as to how to approach the FIC tune on these engines? If there is no base map for let's say stock injectors or you can't post it here for some reason, give us the right direction in which to move:

- Should (or can) this engine be tuned by adjusting the MAF map, O2 map, or both? Or should we not tamper with either to avoid disturbing the stock ECU's close loop and use instead the ability to change injector flow through the MAP settings?
- If O2 map is the way to go, give us the approximate values for a typical low-boost application to try and fine tune.
- If using larger than stock injectors and changing MAP values to tune is the correct procedure, just say so and suggest an approximate approach for injector size you used on your cars.

There is always just one option that actually works, why explore dozens of useless variants in search of the right one when you apparently have an answer ready?

As far as I get it the only base map available for Civic is the one that comes with FIC from TSI turbo-kit makers, but it is wrong in many ways:

The MAF settings in the map at low value are set at 0, which makes the ECU flash a low MAF voltage CEL and go into limp-mode with revs limited to 3.5K. These need to be changed to something sensible to stop ECU seeing MAF malfunction. The rest of the values in this map are also screwed, using the scanner I have set them to the actual values that the MAF is outputting for ECU.

The same is true for O2 map, the values are wrong for low load (I have not checked yet the high load) and the actual signal from stock lambda needs to be entered to make the ECU happy again. I am attaching the calibration file which has these settings corrected - changes only made for low load, so don't use it for anything more than to compare with your TSI cal file.

With this approach by TSI (i.e. using the MAF and O2 voltage settings instead of percent or offset) this means that I need to actaully recreate the entire stock ECU signal input from the sensors to make it work correctly before trying to modify the signals. This is a very extensive piece of work for such simple a task as adding fuel into NA conversion for turbo. The simle approach would be to keep the O2 and MAF maps in percent or offset seetings without changes at all (at low load) and correct the fuel injector flow through MAP load map.

What I am looking for is for someone from AEM to confirm which approach is the right one with respect to Civic's ECU to save everybody's time from useless research and make FIC tuning quick and easy instead of a pain. I hope what I am saying makes sense.

Regards all,
L.

[JR]

Guys from AEM,

If you have done successfully two Civics why don't you give a clear message to the community as to how to approach the FIC tune on these engines? If there is no base map for let's say stock injectors or you can't post it here for some reason, give us the right direction in which to move:

- Should (or can) this engine be tuned by adjusting the MAF map, O2 map, or both? Or should we not tamper with either to avoid disturbing the stock ECU's close loop and use instead the ability to change injector flow through the MAP settings?
- If O2 map is the way to go, give us the approximate values for a typical low-boost application to try and fine tune.
- If using larger than stock injectors and changing MAP values to tune is the correct procedure, just say so and suggest an approximate approach for injector size you used on your cars.

There is always just one option that actually works, why explore dozens of useless variants in search of the right one when you apparently have an answer ready?

As far as I get it the only base map available for Civic is the one that comes with FIC from TSI turbo-kit makers, but it is wrong in many ways:

The MAF settings in the map at low value are set at 0, which makes the ECU flash a low MAF voltage CEL and go into limp-mode with revs limited to 3.5K. These need to be changed to something sensible to stop ECU seeing MAF malfunction. The rest of the values in this map are also screwed, using the scanner I have set them to the actual values that the MAF is outputting for ECU.

The same is true for O2 map, the values are wrong for low load (I have not checked yet the high load) and the actual signal from stock lambda needs to be entered to make the ECU happy again. I am attaching the calibration file which has these settings corrected - changes only made for low load, so don't use it for anything more than to compare with your TSI cal file.

With this approach by TSI (i.e. using the MAF and O2 voltage settings instead of percent or offset) this means that I need to actaully recreate the entire stock ECU signal input from the sensors to make it work correctly before trying to modify the signals. This is a very extensive piece of work for such simple a task as adding fuel into NA conversion for turbo. The simle approach would be to keep the O2 and MAF maps in percent or offset seetings without changes at all (at low load) and correct the fuel injector flow through MAP load map.

What I am looking for is for someone from AEM to confirm which approach is the right one with respect to Civic's ECU to save everybody's time from useless research and make FIC tuning quick and easy instead of a pain. I hope what I am saying makes sense.

Regards all,
L.

I was all set to answer your post line by line, telling you that there is no one “right” way to tune it and that the map supplied to you by TSI, and all it’s problems is between you and TSI but that wont help you nor anyone else so let me do something else here.

First, there is no single way to correctly tune an engine. You can do many different things and end up at the same good result. You can also do a poor job of the “correct” way and still have a poorly tuned engine. But despite this, I have found that there are ways that are more likely to meet with success than others.

1. Don’t try to tune by tweaking the MAF sensor. The factory ECU uses the MAF as a trim device and any great difference between the signal value and it’s expected signal value will be ignored and will throw a code. Just determine a max MAF value for a N/A engine and clamp the MAF output table (in voltage mode) to this for each RPM.
I detailed how to do this on the tC in another thread:
http://forum.aempower.com/forum/index.php/topic,19962.msg110255.html#msg110255

Also, never try to tune with both the MAF and MAP sensor fuel tables at the same time. It's should be one or the other, not both. See here ==> http://forum.aempower.com/forum/index.php/topic,19962.msg110249.html#msg110249 for more discussion

2. Don’t try to fight the factory O2 feedback, roll with it. If the factory O2 feedback is trying to hit 14.7 AFR at partial throttle let it. Figure out when it goes out of feedback and only tune it in those places. They all go out of feedback at WOT and high RPM, your job is to figure out where it transitions.

3. If you are boosting a N/A car (everyone is) then you have to figure out the MAF and MAP clamps before you begin tuning. You MUST clamp both of those signals.

4. The ideal way to start tuning a car is in it’s totally stock form. Install the FIC with a zero map (all base maps are by definition zero change maps). It should start and run exactly as stock. Now put in your bigger injectors and change your fuel map to globally reduce the pulsewidth based on the size ratio ((old/new)-1) example new = 440cc, old = 280cc ((280/440)-1) = .636-1 = -36.36 or a 36.4% reduction in pulsewidth is required to get the same fuel amount. Filling the fuel map with -36% everywhere should get the car running the same as it did stock but with bigger injectors. Now you are ready to turbo.

I’m sorry I’m just giving you basic tuning tips rather than a tuned map. When selling the F/IC to turbo kit builders we rely on the fact that the turbo kit builders are going to map the FIC and deliver it in a complete form. AEM has no control over the maps supplied by an individual tuner for use with their kit. (I have not seen the TSI map, I am only going on your descriptions). I don’t know what the TSI kit consists of, the injectors, the fuel upgrades, the manifolds or anything. AEM simply confirms that the F/IC can be used on a specific platform and that a tune can be created but the kit specific tunes are up to the kit manufacturers. But if you post specific problems we will try to help you with them.

[Alam]

My question is pretty simple: For an all motor set up with cams, CAI and exhaust. If I send a STEADY 2.2V (across the whole rpm and load band) signal to the ECU I'll have my car running always thinking that the 14.6 AFR is been kept and been so, it (ECU) doesn't have to change nothing on my fuel parameters?

[Mura]

Hi everybody

We are tuning at the moment R18 TSI turbo kit and have found few strange issues.

On a dyno during appliance of load before power-run at 1800-2000 rpm our scan-tool shows knock and knock-retard about 9-12 degrees. It happens exactly at the moment of deactivation R18 vtec ( transit from one valve mode to power mode). BUT !  No any audible sounds. Absolutely silent!  We've tried to hear simply by ear and by knock-box with earphones.

Yes, we can define that issue as false knock and advance timing up to desirable power, but one hour before there was not that problem, ECU did not hear knock and did not retard timing. The issue looks like floating nature and we'd like definitely to find a reason before let the car go. Difference between "no knock" and "with knock retard" timings is about 8 degrees. Same power curves.

Also we have weak misfires during cruising on dyno.

One of my suggestion is that boomslang harness does not use shielded wires for CKP and CMP.

Any ideas or expirience ?

Thanks
Dmitry

[Mura]

My question is pretty simple: For an all motor set up with cams, CAI and exhaust. If I send a STEADY 2.2V (across the whole rpm and load band) signal to the ECU I'll have my car running always thinking that the 14.6 AFR is been kept and been so, it (ECU) doesn't have to change nothing on my fuel parameters?

From my personal experience with piggybacks on last hondas , 2.2 volt around all map means only that an ECU wll be trying to follow factory's TARGET map. I.E 14.5-14.7 on cruise and 12.5-12.8 on load.    Just because a conversion table (O2 table-target lambda) that was done on idle does not work properly during load conditions.

[SB]

Interesting situation, Dmitry.
How is the FIC's Ignition table configured at that load and RPM point? If you feel it is safe for your particular setup, can you try setting the FIC's ignition table to 0 during that transition? This is not a great explanation but the FIC may be 'buffering' the crank or cam signals when the ignition table is configured to retard timing, causing a perceived knock due to misalignment. When the table is set to zero there is no buffer, the input signal is sent straight through.


I agree that in theory it would be best to shield the cam and crank signal wires, but in practice I'm not aware of any noise issues that were caused by unshielded wires. It is possible that this is the problem in your case but I would look elsewhere as well.

[Mura]

Interesting situation, Dmitry.
How is the FIC's Ignition table configured at that load and RPM point? If you feel it is safe for your particular setup, can you try setting the FIC's ignition table to 0 during that transition? This is not a great explanation but the FIC may be 'buffering' the crank or cam signals when the ignition table is configured to retard timing, causing a perceived knock due to misalignment. When the table is set to zero there is no buffer, the input signal is sent straight through.


I agree that in theory it would be best to shield the cam and crank signal wires, but in practice I'm not aware of any noise issues that were caused by unshielded wires. It is possible that this is the problem in your case but I would look elsewhere as well.

Yes, SB, I've tried to set timing at low rpm and load conditions (1800rpm/80-110kpa) following my understanding what timing should be in this area and from safe side. And , sure, timing here is retarded enough.  Agreed with you that most probable reason could be in CKP/CMP sensors interaction.   

BTW , I've tuned few Euro FN2 CTRs with JRSC RSX-Race kit. Euro civic version has different MAF comparably to USDM and JR PnP kit does not work here. I used piggyback and ONLY CKP correction for timing. Everything worked well and all those civics run as well up to date.

Have you tried FIC without CMP correction ?

Anyway,  thanks, will try your suggestions tomorow, it's night here already )))

[SB]

I'm not 100% certain how the OEM ECU approaches variable cam timing, but I can guarantee that our EMS carefully measures the relationship between the crank and cam signals, and looks for different relationships depending on the desired cam advance.


When you perform the RPM calibration in the FIC software, it measures the relationship between the two signals so it knows how much each signal should be delayed at certain RPMs. The FIC was designed with variable cam timing in mind so it should not interfere with the relationship that the stock ECU is expecting.


It should be possible to intercept the crank signal only, but this will change the relationship between the crank and cam signals which may cause less camshaft advance (best case scenario) or fault codes (worst case scenario). Use caution if you attempt this, changing the cam timing will change the fuel and ignition requirements of the engine.

[sass]

To SB
As so as I know all your Hondas (2006-2007) calibrated by Z104. And you hasn't problem with jerking on low RPM (1400-2800)
In my case on TSX k24a2 compressor installed Z110 (software ver. 3.04) and it works perfectly and I haven't promlem with jerking on low RPM. AEM recommend me to install Z104, but I can't install it. May be Dmitry has same problem? Because he also have installed Z110, which does not recommended for Honda last generation.
I'll glad to listen your comments about this.
Thanks.

[Mura]

Does anybody know how EGR and Knock control are working together on 8th civics ?


2 sass  Thank you for idea, but I tried today to reflash with Z104.  Unfortunately there was same result as with Z110 -  jerking  on low rpm/low load.

[Maew-Rx]

Hey Guys! I have some question about wiring of O21+ in wiring diagram of civic 06.
if i use OEM narrow band O2 sensor, Do i need to intercept signal wire by resistor before FIC tab signal like instruction manual of FIC?
and What's different about install and not install resistor ?

Thankx.

[fantola]

Hey Guys! I have some question about wiring of O21+ in wiring diagram of civic 06.
if i use OEM narrow band O2 sensor, Do i need to intercept signal wire by resistor before FIC tab signal like instruction manual of FIC?
and What's different about install and not install resistor ?

Thankx.

Just tap it without resistor like illustrated in the first page.
You only need to add a resistor when the 0-1V swing type O2 sensor equipped.

[gohybrid]

I was reading about SB's test using the voltage mode to cause the ECU to work toward a different target AFR.  I have one question about doing it this way:  So, you increase the voltage to feed the ECU a leaner value than what is measured to cause the ECU to add fuel.  That much makes sense.  But say you run at the same RPM, same load indefinitely.  The voltage output from the F/IC should never change, so wouldn't the ECU continue to add fuel even after the target AFR is reached?  Doesn't this then over time cause the system to be progressively richer?  It seems to me that in offset or percentage mode, the input and output values are still related, so the ECU could add fuel until the altered signal came into agreement with what it was shooting for.

*edit*

thinking about this a little harder, if i log the stock AFRs and duplicate those with the voltage map, then i get the ECU to stop screwing with the trims.... the only downside i see to that then is that the system is no longer adaptive and might need to be retuned for significant variances in environmental conditions.

do i have this all right so far?

[SB]

The modern 4- or 5-wire 'wideband-type' factory O2 sensors can take a while to understand. Unlike the older 1- or 2-wire 'narrowband' sensors which will generate a different voltage at different AFR's, most modern sensors are measuring AFR based on the amount of current flowing through the sensor element. If you use a multimeter to measure the voltage at the sensor it will remain constant no matter what the AFR the sensor is measuring. So unfortunately the FIC is not able to measure the 4- or 5-wire factory O2 sensor and deduce an AFR reading from it, because the voltage is always going to be about 3V. If you want to know the AFR you will need to install a separate wideband O2 sensor such as the AEM 30-4100 gauge-type controller.

Here's where it gets tricky: when the FIC's O2 channel is tapped directly to the stock ECU's O2 signal wire (without a resistor), and the O2 table is configured in Voltage mode, you are actually using the FIC to change the current that the stock ECU is measuring. Since the current is different, the stock ECU will measure a different AFR than what the sensor is trying to send. The FIC is not generating the signal, it is only altering it a little. The stock ECU is still able to see "too rich" and "too lean" for its closed-loop trims, but the FIC essentially shifts the AFR target that is considered "just right."


Hope this helps,
SB