12/03/2007: "ECM Peak&Hold mod /Befecsi hack"

Okay. So. It has started with an intermittent code 42. After double checking everything I've found a post which suggested that this code could come because of the overheated ECM. The six low impedance injector on one driver generating too much heat. It's time to fix this. The 749 ECM has two drivers and I've replaced both of them when I switched to the low impedance stuff. However replacing just the drivers leaves the ECM in saturated mode, so you are running your peak and hold injectors saturated. It works, both I don't like it because of two things. First it's running the piss out of one drive, while the other one does nothing. Especially in an 8 cyl configuration. Second the injectors overhead for opening and mainly closing is much longer (2msec vs 1.2msec in P&H mode).

Here is what I've done. Switched to the Sunbird wiring. The turbo 4cyl Sunbirds were using the same 749 ECM's as the SyTy's but they were equipped with four P&H injectors from the factory, cabled such a way that two of the injectors are running from one driver. I got the schematics of the Sunbird wiring, disassembled the ECM 32pin blue connector and made the following steps:

*moved pin C12 to C10; the SyTy wiring use C11 and C12 to drive the injectors, but these pins are internally connected. The 2nd driver's can be utilized via pin C10
*replaced the resistors on the board at position R76 and R93 with .05ohm 1% 2watt resistors (note: that's for a v8 application the v6 would require .067 if I'm right)
*there is a lot of misinformation about adding patches between pins D5-D6 and C13-C15. I've tried it, worked, but later found that the SyTy configuration C13-D7 patched and grounded were used in a V2 Sunbird wiring, so it does the same (actually Sunbird's harnesses used D6 instead of D7 but they are internally connected). Now I'm using this later one

The wiring modification required no soldering you just moving ECM pins, totally reversible. After the first start it was obvious instantly that the truck went way lean (I guess because the less opening-closing overheads) especially at idle. Anyway it was easy to fix that.

For the records I'm using memcal 16150500 from a TPI v8.

Hopefully this will work well at WOT too... Random pictures below...

Na. Ez vmi amit nem fogok magyarul leírni, a fenti leírás első sorban angol nyelven értő SyTy-s kollegáknak készült, másoknak ebben a részeletességben érdektelen. A lényeg, hogy eddig szaturált módban használtam a peak&hold befecsiket. Ezt most orvosoltam és remélem, hogy nekem ettől jobb lesz.