It's a stock H/C/I L05, and I for the life of me can't get rid of the rich spike (11 AFR) immediately after a part throttle acceleration followed by a lean spike (Up to 16). I zeroed all my MAP based AE to try and make it easier for me, because I was not even sure which (MAP or TPS) was responsible. I'm talking normal driving here, the TPS is barely breaking 10%.

Since it's going rich, then lean, I want less fuel over a longer duration, right? So I should be lowering the AE TPS PW, and also lowering the TPS filter to drag it out a bit? Or perhaps just leaving the PW where it is, and only lowering the filter coef. to drag it out?

 

You need to add in some MAP vs. AE. The TPS vs. AE comes in immediately as the accelerator is pushed then the MAP vs. AE follows extending the pump shot.

 

What is your Enable Min TPS% constant set at?

 

Lowering the filter coef's will drag out the AE. It will also increase the volume of AE, just reduce the PW tables to bring the volume back down. The last five columns of the EBL dump file are AE information. The ones that start with 'd' are the deltas to show where in the PW tables the AE PW is from. The AE columns are the PW's from the tables, the last column being the final AE PW that is added to the sync injection PW.

Note that this final value is the two PW's added together then compensated for by the CTS & RPM tables. By looking at the two PW columns (mapAE & tpsAE) you can get an idea of which is contributing when along with how much.

RBob.

 

About how many tenths of a second does it take for a PW adjustment to make it through the engine and register on the WB?

 

During sudden throttle transients or high load, there will only be a couple hundred milliseconds or so for the AFR to register on the WB if its within a foot or two of the exhaust ports.

Also, from my experience, its easier to keep the TPS and MAP AE seperate. The ammount of fuel required from the fuel settling out in the manifold is mainly a function of MAP as the rise in pressure allows the gas to settle out if the manifold temp is below the boiling point of most of the fuel components. A hot intake manifold will typically require no AE at all while a large, cold manifold will take everything you can give it. Basically the MAP AE will typically need to be active for up to a few seconds with the fuel volume determined by the temperature of the intake manifold itself.

The TPS AE is mainly for compensating in delays in fueling. Even the later computers that calculate the pulsewidths for the injectors in real time for each cylinder have some lag due to the sensors and such. The earlier ECMs had the fuel calcs fixed at 80 Hz, so during certain conditions you can experience a decen ammount of lag in the fueling when you slam the throttle down. This will typically appear as a very short lean spike. The engine will very momentarily hesitate and then take off like normal. Giving a few hundered milliseconds of TPS AE is really all thats needed to compensate for this.

 

It can take quite a while. Starting from the beginning, with 80Hz loops in the ECM that is a maximum of 12.5 msec for the ECM delay. This is the time between when an AE event is seen and queued up for injection. Note that this is for synchronous AE injection.

With a TBI system an injector fires every time a cylinder fires. This gets tricky as this means that at firing time another cylinder is halfway (90 deg) through the intake stroke. With the next in line at the beginning of the intake stroke (I am ignoring cam timing such as intake opening BTDC and closing ABDC).

The transport delay through the engine is dependent upon RPM. At 1,000 RPM it takes 120 msec from an intake to intake event on a single cylinder. At 2,000 RPM that becomes 60 msec. Usually once over 2,000 RPM the need for AE tapers off (intake velocity reduces wall wetting).

With the engine starting to exhaust prior to the intake stroke, that time can be made at 3/4's of a full cycle. Or 90 msec at 1,000 RPM and 45 msec at 2,000 RPM. With it extending to the full cycle time to complete the exhaust stroke (120 & 60 msec).

A WB sensor can easily lag another 200 to 250 msec. There will also be some lag in the WB controller. This would be dependent upon the level of filtering and the firmware update rate.

There is then the time it takes to travel out the port and down the exhaust pipe.

Worse case at 1,000 RPM for just the values I listed is: 12.5 + 120 + 250 = 382.5 milli-seconds. Or over 1/3 of a second.

RBob.