ion_four

I'm still working on my own little circuit to run larger injectors with boost, ala AFC-hack, but need a little info...
Which sensors determine dutycycle? MAP. TPS?(or is that just ignition)...
What is the injector dutycycle for stock injectors/pressure for a D16z6?

I am of the understanding that the AFC-hack really only utilizes altering the MAP signal, is that all?

Any help would be GREATLY appreciated.

falcongsr

i can tell you the MAP sensor has the greatest influence on injector duty cycle. other sensors can modify the D.C slightly (cold engine coolant temp sensor etc) but the map rules the roost.

sudden changes in throttle position cause more fuel to be dumped for a short period of time. the quick rushing in of air when the throttle opens causes additional fuel to be needed for a smooth power delivery.

also, i think the ECU adds a squirt of fuel when the throttle plate slams shut ..also for drivability. maybe it acutally removes fuel? hmm that doesnt really make sense though. *shrug*

ion_four

thanks. The circuit I was trying to use reduced the voltage that the MAP sensor sent to the ECM, but also reduced the voltage that went into the MAP! I think that's what was giving me the CEL, correct me if I'm wrong. I think that if I just reduced what was going into the ECM I might be cool...if I scaled everything down then I might be able to run boost and larger injectors...sigh...I wish I'd paid more attention in electronics lab

falcongsr

well the ECU goes into open loop mode around 2.8v or so. you will be reducing the voltage and so the ECU will be in closed loop mode while in boost and it will be taking out fuel based on feedback from the o2 sensor. not good.

Sonny

You can get around this by wiring in a relay on the TPS circuit. It doesn't matter where the MAP voltage is it if the TPS signal is 4.5 volts. The ECU will go into open loop at 25% throttle, regardless of the MAP voltage. That is the basis of Rafe Jaffey's TPS trick for JRSC tip-in detonation.

I concur with falconGSR's findings about MAP voltage. With my J&S Ultra v.2 and my JRSC, I was able to find that the ECU switched into open loop at about 2.8 volts from the MAP sensor.

Sonny

falcongsr

yeah but he wont have the pressure switch like we did on the JRSC.....not sure about this one. too tired to think...

Sonny

Yeah, the pressure switch certainly simplifies things. I think in this instance, you'd need to rig up some kind of relay that would pull the TPS voltage up to 4.5 volts at a predetermined MAP voltage (adjustable by user).

Sonny

ion_four

How is this bad?
I understand basically how the afc "hack" works and this is somewhat related;...the "hack" reduces the output of the MAP sensor by a specific amount (approx. 30-40%) so that the ECM reduces the D.C. to a point where the 450's idle correctly and when boosting up to 10-11psi... is this accomplished in a manner such that you will not excede the 2.8v that will put the car into open loop and thus, fuel will always be adjusted according to the O2 voltage...? Or is it that the computer will be referencing its preprogrammed tables for part-throttle vacuum operation and be applying the correct amount of fuel because of the larger injectors (which it is not aware of)? The second seems more likely...

On a related note; can one have ONLY the MAP sensor hooked up through the AFC and run the hack, or must other sensors such as TPS be modified as well for it to work?

I'm not too up on my electronics, but learning (again) and trying to figure a way to circumvent the AFC, since it seems kinda cheesey to me, since a stand-alone is in my (distant) future...

ion_four

Sonny, do you mean that the ECM goes into open loop at 25% throttle normally, without the Jaffey hack?! I would've thought it was much higher, atleast 50%...anyway...it seems like you're implying that open-loop is good (since the fuel is determined by the actual a/f mix in the exhaust) as opposed to the table which no longer applies (since there's boost). Seems like falcongsr is saying that you don't want your fuel determined by the o2 sensor...I guess it depends on how quickly it can compensate (or predict) by adding fuel when MUCH more air is present.

Sonny

Rafe Jaffey's TPS trick forces the ECU into open loop whenever boost is present. The main reason for this hack is because with the JRSC, the MAP sensor is located before the compressor. So, it cannot react fast enough to changes in manifold pressure (ie, boost). With the JRSC, it is possible to hit boost without the ECU going into open loop (ie, throttle is less than 25%). When the ECU is not in open loop, it is trying to maintain a 14.7:1 a/f ratio for emissions and fuel economy. This is too lean for boost and you get the dreaded tip-in detonation that plagues many JRSC users.

As far as the ECU switching into open loop at 25% throttle, it is definitely 25%. Look at this screen capture of a tuning session with my Hondata.



When the ECU switches into open loop, it stops trying to adjust the fuel (short term fuel trim) to get a 14.7:1 a/f. That is shown on the graph by the short term fuel trim going to 0.

I think what falconGSR is trying to say (and I completely agree with him) is that any time you have boost, you do not want the ECU in closed loop. The engine will most surely detonate since the mixture will be too lean.

Sonny

ion_four

It's not that I didn't believe you about the 25% throttle switching everything to open loop, just seems a little odd.

So, the afc "hack" stops the computer from going into closed loop (reading the oxygen sensor to determine fuel mix), correct? so when you have 25% throttle or more (TPS outputs 4.5v) it switches to open loop and uses precalculated fuel tables that are chosenmainly by the MAP sensor output, ignoring the o2 sensor? alright, I think I've got that straight.

SO, if I were to reverse-engineer the afc hack; I would want everything to be normal until 25% throttle, then have the output of the MAP sensor be scaled down by an appropriate amount for the 450cc injectors (and in the process be hiding the boost from the ecu)?

Does the MAP sensor play much of a role, if any, when the computer is in closed loop mode?

seems like I'm still missing something and it's late...sigh...
thanks, so far, sonny! you, too falcon; I'm sure I'll figure something out soon here...

ion_four

So, I think one last thing...thoughts on DIY afc hack...

Will a car using the afc "hack" be in closed loop at idle/part throttle? How does it idle correctly and get good mileage without the O2 sensor controlling fuel?