AEM FIC installation tips?

[Andy Blyth]

I am in the process of installing an AEM FIC to my 1993 J-Spec TT6. I'm having a few troubles getting the engine to start reliably. I have reason to believe this may be due to the crank and cam sensor wiring.

 

I started by following the wiring suggestion in the latest FIC manual, that is to cut each of the +ve sides of the sensors and wire them into the MAGI+ terminals of the FIC. The MAGO+ terminals were then connected to the stock ECU and the MAGI- terminal was tapped into the -ve side of the sensors. This did not work well. The engine would only fire up maybe 1 in 10 tries and could not hold an idle. The smell present indicated fuelling was working.

 

Attempt two was to revert to the wiring suggested in an earlier version of the FIC manual: both sensor wires cut and conected to MAGI+ and MAGI-, MAGO+ and MAGO- were connected to the stock ECU. This greatly improved the running, engine fired on 2nd or 3rd attempt. Once the engine was started it idled fine and revs smoothly. However the ECU kept reporting fault code 13 (crank sensor).

 

After a thorough search of the AEM forums, a suggestion in general for Toyotas was to wire a 2k2 resistor in series with each of the MAGI- lines. I did this to all three sensors which had the result of reducing the number of occurrences of fault codes generated but did not eliminate them. The faults are now only reported at engine start and at low idle speeds.

 

My next planned steps are to re-do some of the harness modifications I have made: replacing the 12" of ordinary hook-up wire with correctly shielded (shields terminated) pairs for the three sensors and experimenting with having the 2k2 resistors in parallel with the sensors instead of in series (another suggestion from the AEM forums).

 

Other relevant information includes:

• I have the O2 sensor tapped as suggested in the installation manual (with a 1k series resistor) however I don't belive this signal is necessary now and will be removing it.
  
• Hooking a scope up to the three sensors shows that the before and after traces for the two cam sensors are very similar. The crank sensor waveforms are different. When stock, I see a sawtooth waveform with rounded peaks. Coming out from the FIC is almost a square wave but with ramps instead of level voltages. Before adding the series resistances, the peak to peak voltages of the signals were higher with the FIC than with the stock ECU. This indicates to me that the FIC inputs have a higher input impedance.
  
• When the engine is running, the logger function of the FIC is indicating that everything is working as expected. I still have a saved log if it would help.
  
• One problem I encountered quite early on was pushed-back pins in the connector. The Molex Micro-Fit used here are quite fragile so I have fixed this using copious amounts of epoxy. I'm pretty sure that this has fixed the issue but I'm not prepared to rule this out entirely. I can make an entirely new harness for the FIC if necessary.
  
• The engine starts and runs normally with the FIC bypassed.
  
• I started with slave firmware z110.hex. I have reverted to z104.hex as the supra does not (I believe) have multiple injections per cycle.
  

 

Is anyone able to suggest anything to try? Or point out anything that I'm doing obviously wrong? How about an idiot-proof guide to reliably connecting an AEM FIC to the 2JZ-GTE?

 

Thanks for any help

 

Andy

[Wez]

Ryan.G is your man for this, he has done a few of these now

[Ryan.G]

Hi Andy

 

I wire the fic in as you have on attempt number 2

 

Only had a problem with error code 13 once and it was due to the molex pin on the connector not being pushed in far enough. (worse bit of the FIC)

 

How have you wired the sensor ground and pwrgrounds for the fic?

 

Sensor gnd should be to E2

PwrGnd to E1

 

No point running new sheilded wire btw! but you are correct with the wave form differences after the fic though.

 

Ryan

[Andy Blyth]

Hi Ryan, thanks for your reply.

 

Sensor ground is connected to E2 (pin 65) power ground 1 and 2 are connected to E01 (pin 80) and E02 (pin 79). I reckon this should be fine as all the power grounds are connected together (with E1) 2 feet into the harness anyway.

 

I modified the harness to put the 2k2 resistors in parallel with all three sensors a few days ago and what a difference! No error codes now. I've done about 7 cold starts now and a couple of dozen warm starts and not a single problem starting. It's just like stock. I've mounted the resistors on the back of the connector inside the FIC to keep the harness tidy.

 

I guess if you've had no trouble and don't need any shunt resistors then my sensors must be right on the edge of tolerance.

 

Have you ever needed to alter the closed-loop mapping of a TT when using a piggyback? I can't imagine the ECU being closed loop when boosting so I have removed the tap on the lambda signal.

 

Now to design a boost controller and get mapping!