Anyone have a TBI intake laying around?

[robertisaar]

got into an interesting discussion on the gearhead-efi forums, so now i must know more....

anyways, supposedly, the 2.8/3.1 TBI intakes apparently have some fuel flow/distribution issues that requires a fairly unique injection scheme to keep AFRs between cylinders fairly close to desired. normally, with MPFI and most TBI setups, injector pulses happen at roughly the same time as a reference pulse, referred to as synchronous fueling. there are little subsets of that(single-fire vs double fire for MPFI, alternating vs simultaneous for TBI), but i'm not going to dive into those since they're not relevant here.

apparently GM decided that keeping things consistent was boring, so on certain TBI engines, the injector firings aren't tied to reference pulses... instead, the calculated pulse width gets turned into a duty cycle and the injectors work off of that.

so, i'm trying to figure out why air distribution seems good enough, but fuel distribution is crap... anyones have any pics of the TBI intake? especially the top half where the TB bolts on? trying to see if GM routed the ports oddly or have any other obvious reasons why this is necessary.

[geoffinbc]

Well with a TBI system the number of pulses is not a big deal nothing is synchronized to intake events. Pretty much as long as enough fuel flows it works. It is very similar to a carb. In fact TBI intakes are pretty much a copy of the old carb intakes. The intake is a dual plane system with each half feeding one side of the motor. The left side is completely separate from the right side except for a small EGR passage shared between them.

[robertisaar]

that's what i've been expecting as well... but for some reason, running a 3.4 w/a TBI unit with injector pulses synchronized to the reference pulses causes two cylinders to work great, two go stupid rich and two go stupid lean... or at least, this is what another member has found, he hooked all of this up to a 94-95 5.0/5.7 TBI computer and adjusted it to work with the 3.4 connected to a 4L60E.

it really doesn't make sense to me why this would happen with synchronized injector pulses, but not when running on a PWM scheme.

[geoffinbc]

But the TBI ECM has absolutely no idea what cylinders are firing. So how could you correct for three different fuelling situations between the cylinders. The TBI manifold even has the throttle body clocked so that the throttle bores are equal distant between the ports. I assume GM did this in order to improve distribution. Even though the system is simple and less refined than other port injection systems I doubt GM over looked a problem like that during a decade long run.

I would suspect his problem may have been due to a combination of other factors. Cylinder wear, cam wear, compression variation, ignition problems or poor spray pattern in the injectors could all be a source of problems.

[robertisaar]

and you're where i ended up.... i have no idea why changing from a sync to async injector firing scheme fixed the issue? i get the feeling that more than that changed when selecting between async and sync firing modes.

[caffeine]

I know the local tuning shop absolutely hates TBI ECMs. They say they make absolutely no sense and you're almost better off with a carb.

[geoffinbc]

No sense? They are so damn simple how can they not make sense?

TBI has got to be the most reliable system out there. It just doesn't fail. My Dad has seen countless Taxi's with over a million kilometers with nothing more than oil changes, cap, rotor, plugs and wires. The fuel system was all original and the injectors still showed a good pattern. He says you just wont see that today on new cars. You will always have a crank sensor, pump or injector go down before the million kilometer mark.