749 in a 928 - MAP troubles

bshaw

New member
Hi All,

I hope you don't mind the intrusion, but I know there are some computer wizzards hanging around here and I'm pretty well stuck. I'm putting a '749 in my Porsche 928 4.7 liter V8. It's NA now but some positive manifld pressure is following the ECU conversion real close.

I have the '749 installed in the 928. I've replaced the injector drivers with Fairchild HUFA76429P3 drivers since I'm going to be running 8 low impedance injectors. I've wired in the resistors on the ECU connector for peak & hold operation, and used a Turbo Sunbird wiring diagram in hooking the '749 to the Porsche. I'm running SyTy code with a 1580 BBZB cal, cyl select set to 8, errors 31 and 32 disabled, trans set to manual. I'm connecting OK with Andy Whitaker's FreeScan and have battery voltage, TPS, coolant temp, and mass air temp reading out OK. That's the good news.

Bad news is I have no MAP, IAC, or knock signals. I've ohmed out all the wiring and have continuity between the ECU connectors and all the appropriate sensor connectors. I can see the MAP voltage on the ECU F15 pin at 2.3V nominally, going down with vacuum and up with pressure (2 bar MAP), but Freescan reads .9 bar and doesn't change. IAC has 11.9V across the 2 coils, no change (no pulses) when I turn the system on or try to reset the IAC position. Both IAC coils read about 52 ohms. I can see a knock signal with my scope at the ECU on F9 when I whack the sensor, but FreeScan doesn't bump the knock counts so I assume the ECU doesn't see the knocks. I swapped my 2nd '749 that I use on the bench and get the same results so I don't think it's a bad ECU.

In the garage, installed in the car, I read the correct MAP voltage at the sensor, at the ECU connector, and at the A/D chip (U5, pin 4). Freescan shows 2.3V, about 1 bar, and doesn't change as MAP voltage changes.

In the house on my ECM test rig I can see the correct MAP voltage at the ECU connector, and at the A/D chip, and Freescan displays the *correct* MAP voltage and pressure. Same ECU, same version of Freescan but works in the house but not in the car.

New data I gathered tonight... It seems to read correctly one time when it starts up. If I turn the key off, then short the MAP signal to power or ground it always powers up with the correct reading. After that if I change the MAP voltage it stays at the power-up reading. Wierd. Anyone have any thoughts that might help?

http://p-928.home.comcast.net/749.html
 

BillC

New member
Re: 749 in a 928 - MAP troubles

Sounds like an interesting project.

I would recommend that you try running new wires from the sensors to the ECM. At least, run temporary wires (over the fender and around into the door, into whereever the ECM is installed). That would at least let you isolate the wiring. It's possible there are some oddities in the 928 wiring that have been overlooked (like splices or connections to other components).
 

E-Rue

New member
do a key on and key off pin check at the ECU, i bet you have a bad ground or something silly throwing you for a loop.

e
 

bshaw

New member
What I found out yesterday is that he ECU does not read all the sensors if the engine is not running. MAP gets read once at startup for Baro and not again until the engine cranks. In the house I'm creating a fake RPM signal so teh ECU thinks the engine is running.

I now have spark and injectors pulsing. I had to send the plenum back to the welder to shorten it up a couple of inches. If I had the plenum back from the welder I think I'd be making some carbon monoxide :)

Thank for the replies guys.
 

bshaw

New member
I'm making some progress here. Got the plenum back and everything installed again. I've been chasing a problem with injector pulse width. I found the sense jumpers wrong on my test bench. Once I fixed that it started working properly. I can clearly see the fixed pulse and the variable hold portions of the injector cycle, and can vary the pulse width as expected.

Out in the car I can now see the pulse portion drop nicely to ground, the hold portion of the cycle seems to have a problem. It's only dropping a volt or so, not enough to hold the injector open. When I pull 6 of the injectors off suddenly the 'hold' portion is dropping 2-3 volts and it starts to try to run on 2 cyls. Pretty obvious I have a drive problem, 2 injectors works but 8 doesn't.

I've changed the injector driver to the Fairchild HUFA76429P3 mosfets, but didn't make any other changes. Do I need to change something else to get the drive for 8 P&H injectors?
 

bshaw

New member
Bwahahahahaha...it's alive, IT'S ALIVE!!!

Running pig-rich, but running!

To answer my previous question, the sense resistors have to be changed as well as the driver mosfets to run the 8 p&h injectors. You need 1/2 the resistance so paralleling in a pair of .1 ohm, 2W resistors did the trick.

Bill
'84 928s 5 speed (FrankenPorsche)
 
Top