JPrevost

I've come to the conclusion that the IAC is junk. Not only is it a PITA, it also doesn't like to idle down to the programmed desired idle speed. After playing around with a MPFI zz4 90' pickup running off of the 7747, playing with my car, and tuning other less radical setups, I'm starting to think that the IAC is only good at 1 thing....starting the engine. I had my IAC unplugged for 2 weeks now and it's only stalled a couple times. My motor is suposed to idle between 650 and 750, I've got my min idle set at 500. The car has NEVER stalled coming to a quick stop, hard turn, no throttle etc. The only time it stalled was when I'm going back and forth from drive to reverse and drive.
The whole throttle follower thing is just pissing me off. I can't seem to find the settings in the 8746 hack file and when I do, they don't seem to change any of the parameters. I've already got the IAC step counts vs temp all filled with 10 and the desired idle speed is 550...it's NEVER gotten down to 550 and I'm starting to think that there is either something wrong with my IAC (doubt it) or I'm blind while looking for the IAC lag vs tps settings.
If anybody has any experience they wish to share, please do so.
If you're thinking that I unplugged the IAC when it was in it's park position, you're right, except that I have it set at 0 from 160 . I like the way the car drives without the IAC because it never seems to decel with the IAC. I actually gain speed and this might, just might, be my stock TC's fault. Without the IAC the car drives like a dream and idles down where it should after warming up!

Brent

iTrader: (1)

From RR's 747 IAC doc @ diyefi

0x5D7: Kickdown delay timer value (seconds).
0x5D8: Kickdown steps.

These two together form a post-startup rpm reduction. The kickdown
delay timer starts right after startup. Once the timer has expired,
the IAC is extended by the amount of kickdown steps.

A nuance: The kickdown steps must be less than the steps used in
the IAC position vs. coolant table value (0x5F5). Otherwise no
kickdown will occur.

Seems like you lowered the IAC steps VS coolant below the kickdown value.


I tend to agree that the IAC is highly annoying. Its really bad on newer cars. Espcially ones with manual trans. They hold the IAC open so long that the RPMs take forever to recover when shifting. I prefer the engine to slow down when I lift the throttle, not hang at that rpm.

Grumpy

You've left me without words,
and I'll just leave it at that.

RedIrocZ-28

I believe my IAC to be the Devil's Spawn! My car loves to idle like it has a vacuum leak. The Vacuum leak is that confounded IAC. I am with you, boycott the IAC

No really though. I have a 110* LSA cam with a dual pattern setup. 218* 224* .495/.502 lift. Could this be why my car has trouble idleing? Any ideas?

RBob

iTrader: (1)

Dang Jon <g>, won't get an argument from me. I've lost
my fair share of hair over them 'thar IAC's.

RBob.

Gary Anderson

iTrader: (2)

Gentlemen,

I'll cast my vote to ban the IAC as well. My manual trans is virtually impossible to drive with any kind of smoothness due to the IAC's slowness to respond. When I lift off of the gas pedal I expect the engine to slow down, not in 10-20 seconds. Obviously I'm missing some IAC parameters on my '7730 system. If anyone knows what I need to adjust to make this thing work I'm all ears. Thanks.

Regards,
Gary Anderson

kvu

sounds like the tq converter is staying locked on deaccel

leirch

Gary, sounds like your using an auto tranny bin as your base, I tried that and had alot of problems on my manual car, or your throttle blades are stuck open.


About the IAC discussion, I've been doing alot of evaluation of the IAC and I can aggree with every person on here saying that its just about worthless, one minute it will let X amout of air past and the next it will let 2x as much as before at the same load/rpm/throttle position!.. whats that about?

Ok enough ranting.

Down with the IAC!


Brendan

kvu

I will say it can mess with the brakes.It apears to hurt idle vacumn if not set right.Ok if it sucks the whats the official way to disable it and adjust the throttle blades?

Dominic Sorresso

Guys,

For the past month or so, I have been playing with the IAC parameters based on Grumpy's work and Rob Rauscher's article. I am using a 7747 on a Xfire Vette. Following their suggestions, the IAC is behaving very nicely.
When I start the motor warm, it idles right down to 750rpm. On decel, it idles down very nicely. And SOTP, the motor feels a bit stronger. INT is in the 124-128 range at all the critical cells. I will tell you it takes a bit of work, and I have burned about 60 bin files so far. But I am very close to endgame at this point. Right now I am in the process of getting open loop idle squared away. It still oscillates at coolant temps of 20-60C. Most of the time, idle IAC counts are at or near 0 with max IAC counts set at 70. Lower than that didn't seem to work. Before I did this, my motor would "flare" on start and especially on warm restart. Heaven forbid it would shut down when I got to a stop. I would need to let it settle down to idle and then move on. Otherwise, idle would not come downeven after driving it for a while. Don't give up hope. It really does work.

IROCZ88

I agree, I can't get the IAC to work on my car either. Really pisses me off. Is there a correct way to disconnect it and idle off the blades? I basically idleing off the blades now anyway cause it stalls with them shut all the way.

Pablo

iTrader: (1)

Well dominic clue us in on the steps you took to do it?

I too could never ever get my car to idle down where it wanted to idle because of the iac even with commanded idle settings were adjusted, iac vs coolant temp, the error term for retract/extend.. the only thing that made a rats *** worth of difference was this one table if i recall correctly that only supposedly added steps at reset for 70 kpa and up and even then it still didnt do much. Car was downright awfull to drive at low speeds too with the 5 speed. I never did disconnect it but i should have tried. Same spark table and everything but with a carb on top of the manifold (running the dist on the ecm still) the thing idles nicely and is smooth as glass to drive at low speeds

It always seemed to fight idling down for the first few seconds and then just finally gave up and idled high

7747, tbi

DAVECS1

iTrader: (2)

I am also battling this foe with limited success. After looking at the IAC motor and some of the source code hacks, it appears to me that the IAC is a PID controlled system. On the 730 binary files I have only been able to locate the P part of PID (porportional). Using WinBin there is a 2D array with some coolant idle parameters. I have some questions if anyone cars to respond.

1. What is the step function the IAC follows if it is a PID control?

2. What parameters adjust the integrative control and dirivative control?

Some of these flare ups described could definelity be from just changing the porportional control. On PID controls that is a rough adjustment and it sounds as if the IAC way overshoots its desired step function before returning to the desired control.

Come on guys chime in and we can beat this thing down.

Dominic Sorresso

Pablo,

To start with I have probably worn out the paper that I used to print out RR's article. I read that over and over and over again. Second, I used TC's Compbin quite a bit looking at other bin files both aftermarket, custom, Grumpy's final answer, and some factory stuff. The areas I was looking to address were startup flaring of the rpm's especially warm. Also the hanging of rpm's on decel and during shifting. The kickdown timer and steps wprked well to eliminate this. Also did some work on the DFCO. I am using a 7747 on a Xfire Vette and I was looking to optimize air speed through the venturi. By increasing airspeed here and minimizing bypass air through the IAC, I should get better atomization of fuel. So limiting the total number of steps for IAC is important. Some use 30, I found 60-70 to work for my combination. Its a lot less than 145 so now a lot more air goes through the butterflies. I played with Hot Bias but then set it back to where it was. Basically, I used some of the examples that are in the back of RR's article as starting points and went from there. I also played with the IAC small and large error coefficients for smoothing out the idle. I still need to get open loop idle correct since I still have some idle oscillation at cold start up when coolant temps are under 60C. However, motor starts cold with only a few cranks and it starts immediately at ignition when warm.

V8Astro Captain

iTrader: (1)

Dominic, I had the same cold start oscillation and I found that at 0x5E2 there is "RPM MAX HOT RPM FOR PK/NEUT" (conversion factor of 12.5). On my calibration it was set to 1100. My cold start-up wanted to go higher than this and it acted weird. Yeah I know it says it's for Hot start but it helped my funny start-up oscillation that I had from day-one.

Now I have it set to 1400 and it seemed to alleviate the oscillations.

HTH

Dominic Sorresso

HTH,

I'll be damned. I went to my bin file looking using the hex editor and lo and behold 0x5E2 had 58(Hex) in it. So you changed it to
70(Hex) I take it and this eliminated the oscillation. I'd like to understand why it does. I will try it tonight. Thanks for the lead.

V8Astro Captain

iTrader: (1)

Correct.

My only guess is that it's some wierd max RPM calibration for who-knows-what and the ECM is having a hissy fit because not everything is "lining up".

After playing around with the O2 sensor, cold start coolant temps, hot start coolant temps, open/closed loop qualifiers it still acted stupid at start-up. Then I changed the RPM at 0x5E2 and I haven't messed with it since.

leirch

I would imagine it has to do with your idle speed being changed from stock? Or are both of your cars/vehicles idling at the same stock speed?

Its looking for a base speed(650/stock?) and if yours was upped a little to get a better idle maybe that threw the entire calculation off?

RBob

iTrader: (1)

I had some IAC experiences this past weekend that may help with
someone trying to understand what can go wrong. It started when
I made a change in a one portion of the calibration. With this
change I then needed to rough-in the VE tables again.

The setup is a TBI on a warmed over SBC run with a '7747 ECM.

The plan was to take a short drive and gather data to get the
VE tables back in shape for final tuning. The idle portion of
the VE table ended up being rather lean. This caused the IAC
to open to ~ 90 steps in order to idle. Normally the steps
would be between 5 & 10.

The effect this had was to cause the engine to rev up whenever I
shifted (manual). As the IAC will not close to a lower count then
the idle setting, along with a reasonable VE value at higher RPMs,
the engine speed increased whenever I depressed the clutch.

This made for a jerky & hard to drive vehicle.


Once back at home base and making the required changes to the VE
table the engine RPM would drop as I shifted. A change in the VE
table corrected the IAC difficulties.

As I continued to drive and gather data there were times that the
engine did not want to idle down. The idle would hang up around
1000 to 1050 RPM (idle is set to 837 RPM). On a hunch I blipped
the throttle allowing the blades quickly to slam shut.

Ah, now the engine idled down. Checking the data logs showed that
the TPS was a hair higher then the required minimum for an idle
condition. What happened was that the throttle hung open just enough
for the ECM to stay out of idle. So, no idle control.

The cal has the TPS% idle threshold set to 1.7%. Anything over that
and the idle control logic is inoperative. I did change this value
to 2.3%. I also need to work on the throttle body a bit. It should
return to a fully closed position readily.

By the above examples it can be seen that the operation of the IAC
is closely tied to the rest of the ECM calibration. Small problems
in one area can cause larger problems in another.

As they say the 'devils in the details' (there's that devil again
when mentioning the IAC!).

RBob.

RBob

iTrader: (1)

There is another aspect of the IAC (on '7747 & '8746 ECMs) that
RR put in a list email and is not in his paper. This is very important
to obtaining proper idle control.

The kickdown steps must be greater then or equal to the IAC
position (from the steps vs coolant table) at the engine temperature
that you wish closed loop idle control. On the '8746 it is this same
point that the idle control via SA also occurs. Example from ANLU:


Note the kickdown of 65 steps. With the above table closed loop
idle control will not occur until the engine temperature reaches 86C.

(Halfway between the 66 steps at 80C and the 64 steps at 91C).

So for anyone running a 180 or colder thermostat, guess what?

This is true of both the '747 and the '8746 ECMs. There are two
other qualifiers for full idle control: 0 MPH and the TPS% less then
the value at $D514:

The 0 MPH is hardcoded in ROM. Once these three qualifiers are
met the ECM will go into closed loop idle control logic. It is also possible
to lower the temperature of closed loop idle control by changing the
above IAC position vs coolant table. Can make the bin value 65
from 71C on up. Then closed loop idle control will occur at a lower
coolant/engine temperature.

RBob.

V8Astro Captain

iTrader: (1)

This next quote is copied directly from the RR 747 IAC paper
RBob...I assume you are right because you seem to know how to read the source code. Am I correct in saying the kickdown steps must be greater than the steps in IAC vs Temp?

TIA

RBob

iTrader: (1)

(Note: double quoting broken, please see previous
post for relevant information)

Good question. Both of the above statements are correct.
There are two separate areas of logic that handle the
kickdown and the closed loop idle.

A good example (using the ANLU table above), at
startup with an engine temperature of 91 C and a
kickdown of 65 steps, no kickdown would occur. However,
the ECM IAC logic would immediately enter a closed loop
idle mode. So in a sense no kickdown is required.

The kickdown steps only need to be greater or equal
at the IAC Position vs engine coolant temperature
where closed loop idle control is desired.

Note that if the engine coolant temperature at startup
is 30.5 deg C then a kickdown will occur (note above
table). There will be NO closed loop idle though.
Not until the engine temperature reaches 86 deg C.

As a closing note, if the kickdown steps were changed
to 35 steps, guess what? The ECM would NEVER enter
a closed loop idle control mode. This is because no
place in the IAC position vs coolant table do the
steps go below 36.

The IAC is a juggling act, no doubt about it.

Hope the above makes sense.

RBob.

JPrevost

Rob, thanks for that extra explination of the IAC tables and values.
BTW, I'm still very active in the eprom burning and source code. I'm waiting for my friends wiring harness so I can do source code changes. Lots of ideas but still no harness for the test bench
I also just finished installing a 2800 stall TC and doing some agressive timing tuning. I actually have found that so far, my "mechanical w/vac advance" dissy table that I made from 4 parameters in my little crate motor booklet has been the best starting poing. No more backfire through the intake when I snap the blades wide open
Anywho, to make life more interesting, my friend's Mom's car is having IAC troubles, go figure! Except it's the other way around, it just stalls every chance it gets.
Ah, and while I'm still so far off topic I might also tell everybody that Big Daddy Don Garlits was at E-town for the old time drags event and being 5' from the Hemi while it's being prepped and then ripped apart is AWESOME. No words can describe the sound of one crew member getting the GREEN (yes, brown when it went in, green after a run) oil in his eyes and the sound of a clutch plate being dropped because it was so hot!!!
Now my last thing about the IAC, it's just a question. How do you actually figure out the TPS %? I can only find out the voltage with winaldl and it's "TPS %" is most likely not the same interpretation as the ecm's.

RBob

iTrader: (1)

Jon you have been quiet. Good to hear that you are still at it.

As for the TPS%, the easy way to get that value is to modify the
ALDL stream. Output RAM location $43 ('8746), or $49 ('7747) and
multiply that value by .39 for the percentage (0 - 100%).

If Joby would like to change WinAldl there is another method. Can
do what the ECM does internally to create the TPS% value.

Starting with the TPS ADC value (already on the ALDL), keep track
of the lowest value ever seen. This will be the TPS closed throttle
setting (a slow filter will also help keep noise at bay).

To calculate the TPS% value:

result = ((((current tps - lowest tps) * scalar) * 4) / 256

If the result is greater then 255, make it 255

Then:

tps% = result * .39

The scalar is a calibration value. It may be found at $D219 ('8746) or
$D287 ('7747). I have seen it range from 96 to 117. The scalar
would need to be an option value for WinAldl. It could then be
input by the user to match their bin value.

RBob.

Dominic Sorresso

RBob,

Any way this could be done using EASE diagnostics tool?

RBob

iTrader: (1)

My understanding of EASE is that it is a scan tool. If this is correct
then the first method of changing the ALDL stream to output the
TPS% variable will work.

RBob.