Restoring AE86 w/ 4AGE 16v + HKS ITB + Haltech E6X

[aceforever]

Ah, I'll have to do that at last resort. I'd like to work with the current igniter if it's workable.

Good news! I found a diagram of the igniter from here: http://forums.club4ag.com/zerothread?id=98235

Reuploaded for future adventurers:


Seems like IGf and Ne are outputs to the ECU. The shielded connector that's coming off the Igniter is actually for the distributor 4 tooth input. From the diagram it seems like the Ne from the distributor is only used for backup, so it shouldn't be necessary. Do you think that's right?

EDIT:
Igniter pinout

[jondee86]

Seems like on the ae86 efi diagram the ignitor connects to the tachometer. Does
this mean that my current stock tachometer won't work without the ignitor?

The AE86 uses a high voltage spike (generated by the coil back emf) to drive
the tacho. This spike is typically derived directly from the coil negative terminal.
Sometimes it is output from the igniter, but taking the signal directly works.

Alternaively, it is usually possible to program an AUX OUT on the the ECU to
provide a tacho signal. This can easily be conditioned to drive the OEM tacho.

Cheers... jondee86

[jondee86]

I have the entire data sheet that diagram is taken from. Later I will post the
accompanying text so you can get a bit more insight into what is going on.

The signal conditioning function for the Ne signal would be helpful, as some
people have actually used the igniter solely for this purpose when their ECU
was having problems working with the raw Ne signal from the distributor. The
Haltech will not have any use for the IGf signal, and will simply provide an IGt
providing it is receiving an Ne signal.

With the VAST system, the igniter controls the spark during starting, and
until the engine reaches some undisclosed rpm. Then the igniter switches to
using the IGt signal from the ECU. That could work

What is not entirely clear is the control of dwell. If this is controlled from
within the igniter (and I think it is), then the ECU will have to be set to let
the igniter do its thing, and not add dwell on top of dwell.

Cheers... jondee86

[aceforever]

I have the entire data sheet that diagram is taken from. Later I will post the
accompanying text so you can get a bit more insight into what is going on.

Wonderful! I'd like to see it when you get the chance.

The signal conditioning function for the Ne signal would be helpful, as some
people have actually used the igniter solely for this purpose when their ECU
was having problems working with the raw Ne signal from the distributor. The
Haltech will not have any use for the IGf signal, and will simply provide an IGt
providing it is receiving an Ne signal.

Ah! This is very interesting! I found a megasquirt info page about this too now that we know about VAST. http://www.megamanual.com/ms2/vast.htm
It looks like the Igniter converts the VR signal into a square wave signal. The page talks about using hall effect setting and needing some resistors.
Would I need to do that on the haltech? I assume not since the manual says it accepts hall effect directly. What's the difference between hall effect sensors and VR? Is it just that hall effect is square wave instead of other waves?

With the VAST system, the igniter controls the spark during starting, and
until the engine reaches some undisclosed rpm. Then the igniter switches to
using the IGt signal from the ECU. That could work

What is not entirely clear is the control of dwell. If this is controlled from
within the igniter (and I think it is), then the ECU will have to be set to let
the igniter do its thing, and not add dwell on top of dwell.

Cheers... jondee86

On another thread I found was coincidentally trying to get the same AE86 + HKS + E6X setup working. Perhaps from the previous owner or a shop tuning for the previous owner?
http://forums.haltech.com/viewtopic.php?f=11&t=1569
I jumped on the thread on the 2nd page, but on the first page a lot of people saying the igniter is "dumb" and to use falling edge at 3ms-5ms.

[aceforever]

Seems like on the ae86 efi diagram the ignitor connects to the tachometer. Does
this mean that my current stock tachometer won't work without the ignitor?

The AE86 uses a high voltage spike (generated by the coil back emf) to drive
the tacho. This spike is typically derived directly from the coil negative terminal.
Sometimes it is output from the igniter, but taking the signal directly works.

Alternaively, it is usually possible to program an AUX OUT on the the ECU to
provide a tacho signal. This can easily be conditioned to drive the OEM tacho.

Cheers... jondee86

Awesome! I'll check out the manual!

[jondee86]

Whole data sheet is here...
http://wenku.baidu.com/view/131a0a8e680203d8cf2f2401.html

I don't believe the igniter is dumb, as the documentation makes it reasonably
clear that the igniter is in charge of dwell. However, when you read the data
sheet you will see that the text does not make it explicitly clear that the VAST
igniter is the same as the ESA igniter when it comes to handling dwell. So...

EDIT: It appears impossible to find a clear indication regarding dwell in the
VAST system. Looking at various posts and articles, and reading the tech data
sheet again, I now think that it is more likely that the ECU sets dwell in VAST
and the igniter sets dwell in the ESA system.

Hall effect sensors provide a square wave (digital) output and VR sensors
provide an analog (rising or falling edge) output. Since the signal conditioner
in the igniter changes the VR output to a square wave, you would need to
use Hall sensor settings in your ECU.

Cheers... jondee86

[aceforever]

If the igniter is not smart and I'm using constant duty, do you know if it would cause any issues? In the stay of the haltech manual it warns you about ruining the igniter if the settings are off.

Also it's pretty interesting that VAST is somewhat similar to the EDIS system where it conditions analog rpm signal without the minus one tooth. And auto fail safe mode.

[aceforever]

EDIT: It appears impossible to find a clear indication regarding dwell in the
VAST system. Looking at various posts and articles, and reading the tech data
sheet again, I now think that it is more likely that the ECU sets dwell in VAST
and the igniter sets dwell in the ESA system.

That seems to confirm what the megasquirt page says about vast: http://www.megamanual.com/ms2/vast.htm

"IGt: The spark timing and dwell signal,"

I personally prefer the ESA system since it would have been easier to modify for the wasted spark configuration.

[jondee86]

Yes, the stock toyota igniter is compatible, set to constant charge, falling edge, 3.5ms

This quote seems to be representative of the opinions expressed in a number
of different forums. Of course you can also find opposite opinions

So it comes down to a bit of experimentation. There is no suggestion that the
igniter will self destruct on the first turn of the key if you get the settings wrong.
From what I see, getting it wrong means that the igniter may overheat and burn
out if the engine runs for a long enough time with the wrong igniter setup.

A digital temperature probe (finger) should be able to monitor for any unusual
temperature rise. And the MegaSquirt forums outline various techniques for
tweaking setting to obtain correct engine operation under driving conditions.

Cheers... jondee86

[aceforever]

Sounds like I need to test and confirm

And the MegaSquirt forums outline various techniques for
tweaking setting to obtain correct engine operation under driving conditions.

Do you have a link to this handy?

[jondee86]

This from your link a couple of posts up...

For dwell settings, set the maximum dwell duration to 3.0, then adjust it with
the car running and idling. Lower the dwell period in 0.1 millisecond steps, being
sure to blip the throttle between each change. If the idles smoothly, and accelerates
cleanly, reduce the dwell another 0.1 and test again. Repeat until the dwell causes
a stumble, then increase it by 0.2 milliseconds. Test drive the vehicle to make
sure the engine does not stumble do to ignition issues under any conditions.

Note that in some cases you may have stumbles even at 3.0 milliseconds. In
those cases, increase the dwell by 0.1 at a time until the stumbles are gone,
then increase it another 0.2 milliseconds.

This all seems quite reasonable And a hot igniter is an unhappy igniter, and as
far as I know, excessive dwell causes the igniter to heat up. So if your igniter
gets hot, you have probably got something wrong in your dwell settings.

Cheers... jondee86

[aceforever]

This from your link a couple of posts up...

This all seems quite reasonable And a hot igniter is an unhappy igniter, and as
far as I know, excessive dwell causes the igniter to heat up. So if your igniter
gets hot, you have probably got something wrong in your dwell settings.

Cheers... jondee86

Ah yes, I did read this. I thought you were referring to something else.

Just like any piece of electronics, heat always place a piece on durability huh?

I wonder what else I need to do for the electronics other than figuring out where the switched +12v is. I'll quadruple check the wiring diagrams and ecu manuals. I think I might be ready to connect all the wires up!

[jondee86]

In one of your pictures, there is a big, black, square(ish), multi-pin connector
high on the firewall behind the engine on the intake side. This connector will
have originally supplied 12V key-switched power for the ignition and injectors.
You can use this as the control circuit for your main power relay, as you will
now be taking power directly from the battery and by-passing most of the old
factory wiring.

More details here..
http://img842.imageshack.us/img842/5020/cplug.jpg

Cheers... jondee86

[aceforever]

Thanks! I found it. It was in plain sight!

I'm also curious what torque spec the fuel tank drain should be set to. I found an 4AG manual and an AE86 manual and I can't seem to find it. I'm worried about over tightening it and cracking the gasket.

[jondee86]

I don't think you will find a torque spec for that plug. Just do it
up firmly with your 3/8" rachet handle, but don't crank on it hard
like it was a wheel nut. Put gas in the tank and see if it leaks. If
it does, tighten it a little bit more.

This is the technique** I use for fuel lines, drain plugs, brake lines
and bleeders.... anything that holds fluid back. If you can find a
new gasket or washer, use that, as they have a better chance of
sealing first time round compared to already used items.

Cheers... jondee86

** Variation of the "torque it up until it breaks, and then back it
off a quarter turn" technique

[aceforever]

Ahh. Good tip. I'll try that.

I failed to get a wb o2 sensor controller and gauge. I'm hearing good things about PLX. What do you think?

http://www.plxdevices.com/product_info.php?id=MULTDM6

They're using the Bosch LSU 4.9 sensor. Supposedly better than the 4.2. I figured that it's probably easier to get a kit instead of having to deal with the current senso work with a new gauge/controller

http://www.plxdevices.com/product_info. ... EMOSMAFRV4

[jondee86]

In the pic where you are holding the plug of a WBO2 sensor, it appears
that there is both a male and a female plug pushed together. The male
side should have had a semi-rigid black cable leading to an O2 sensor
inserted into a bung in the exhaust downstream of where the header
pipes join to become a single tube.

The female side of the plug had several yellow wires in a black sheath,
and this should have lead to some kind of control or display unit. Generally
speaking, WBO2 sensors need a control unit to allow them to interface with
an ECU or gauge display. If any of these components are missing, then
you may well be better off starting over with a new WBO2 kit.

It appears that the Haltech can use a narrow band O2 sensor input to run
closed loop fuel control. Therefore, if you anticipate using this function,
when buying a WBO2 kit you need one that can provide a simulated narrow
band output for the ECU. Other that that the WBO2 will only be driving a
gauge for your in-cabin entertainment

I use an Innovate LC-1 with digital gauge but there are many others out
there... you just need to choose one that has the features you want at
an acceptable price.

Cheers... jondee86

[aceforever]

Just wanted to give an update on how this project is progressing.

Fuel System
The fuel level sender was misbehaving, so I took it apart and cleaned up the spiral wires like this youtube video: https://www.youtube.com/watch?v=hi0bHluXFX4 Now it works! Good as new.

Before:


After:

The empty fuel light came on, probably because of the thermistor in the sensor was all heated up.

---

I managed to jerry-rig some wiring to the pump connector and stuffed the fuel line into a catch. And to my amazement, the pump works!


There were some black crud in it.


After a pumping a gallon of fuel, there were little to no particles. How clean should the fuel be? (ideally nothing but fuel right?) There were some weird sounds coming out of the cylinder on the firewall or around it that I was hoping you guys would know. I'm currently uploading a video. I'll link it in shortly.

I think the most annoying aspect of working on the car so far is rounded or snapped bolts and dealing with fuel. That stuff is nasty and the fumes is horrible.

Rust Treatment

I've had really good results from submerging rusted items into Evapo-Rust yellow-greenish liquid. It's amazing how it easily it removes rust from screws and small parts. You can try to remove rust from bigger objects, but it's not that easy. They tell you to soak a paper towel and wrap the rusted area. It sort of works, but it's hard to make the paper towel stay stuck onto the surface.

Before:


After:


PlastiKote Rust Converter. This stuff really works! Just sand down the loose rust and apply the milky liquid on with a brush or paper towel.

Before:


After:


---

Bolts

I was dumb enough to snap a bolt when installing the fuel rail. Found out that maximum torque based on the markings on the bolt DOES NOT MEAN THE SAME as recommended torque. Luckily it snapped higher up so I could use a vice grip on them. A quick trip to the local auto parts store fixed the broken bolt issue. Got some nice class 10.9 ones.

[aceforever]

In the pic where you are holding the plug of a WBO2 sensor, it appears
that there is both a male and a female plug pushed together. The male
side should have had a semi-rigid black cable leading to an O2 sensor
inserted into a bung in the exhaust downstream of where the header
pipes join to become a single tube.

The female side of the plug had several yellow wires in a black sheath,
and this should have lead to some kind of control or display unit. Generally
speaking, WBO2 sensors need a control unit to allow them to interface with
an ECU or gauge display. If any of these components are missing, then
you may well be better off starting over with a new WBO2 kit.

It appears that the Haltech can use a narrow band O2 sensor input to run
closed loop fuel control. Therefore, if you anticipate using this function,
when buying a WBO2 kit you need one that can provide a simulated narrow
band output for the ECU. Other that that the WBO2 will only be driving a
gauge for your in-cabin entertainment

I use an Innovate LC-1 with digital gauge but there are many others out
there... you just need to choose one that has the features you want at
an acceptable price.

Cheers... jondee86

That makes total sense. Since I only had the sensor, and no controller or gauge. I went with the PLX sensor/gauge combo. It's a pretty good deal for the price, and pretty good reviews so far.

The sensor should be a direct swap. Just need to find out where to place the gauge in the cabin.

[aceforever]

Weird noise coming from cylinder on firewall. What is that thing?

https://www.youtube.com/watch?v=QkI8Y37alIE

Is it because something should be hooked up to that tiny plug?

I also don't have the FPR hooked up to fuel, could it be that?

EDIT:
Looking at this picture it looks like the FPR goes to a 23265 "Filter". That looks a lot like the 3 hole plug on the Tank Vacuum Box. Is that right?


Still not sure what goes on the cylinder on the firewall.

[aceforever]

Quick update. ECU hooked up to battery and ignition switch +12v. TPS also wired up. The ECU STARTS UP GUYS! I was kind of skeptical that it was gonna start up since it came in a pizza box.

[yoshimitsuspeed]

Nice work. There are not many people who would have gotten this far.

[aceforever]

Thanks yoshimitsuspeed for the encouragement! I really couldn't have done it without you guys! Especially jondee86. I also reeeeeallly want to drive this car. Hopefully everything works out

On a separate note, I've been searching for a place to tap +12v for the O2 sensor controller. Also wondering if ignition on or accessory would be better.

[aceforever]

According to this, I think I can use fuse #11 to go along with the intention of the fuse. (Guage 7.5A) I'm also thinking this must be powered only when IGN is ON. I'll double check.

[aceforever]

Also my clutch pedal support is heavily rusted. Obviously this could be cleaned up, but is this normal? Should I be worried? It also seems like there's a bit of liquid at the area surrounding the cylinder.

[davew7]

My clutch pedal bracket was also pretty rusty. It appears that Toyota installed the bracket, without any paint, because removing the other component parts to paint the bracket, there was not any signs of paint on any of the other surfaces. Dave W

[aceforever]

According to this, I think I can use fuse #11 to go along with the intention of the fuse. (Guage 7.5A) I'm also thinking this must be powered only when IGN is ON. I'll double check.

My suspicion was correct. I tapped the wiring coming out of the 1J connector that goes to the combination meter (Toyota speak for instrument cluster) and ran it to the passenger side for me to hook up to pos+ wire of o2 sensor controller. I grounded the negative onto the grounding point behind the kick panel. Turned ignition switch to ON and did initial testing. Everything powers up correctly (controller, sensor and gauge). Success!

[aceforever]

My clutch pedal bracket was also pretty rusty. It appears that Toyota installed the bracket, without any paint, because removing the other component parts to paint the bracket, there was not any signs of paint on any of the other surfaces. Dave W

Ahh. Seems kind of narrowsighted of toyota not to paint any metal in a car. I'll probably need to disassemble it and soak it in Evapo-Rust.

[jondee86]

Weird noise coming from cylinder on firewall. What is that thing?

That cylinder appears to be the "Charcoal Cannister" that is used on the USDM
AE86 as part of the emissions control system. It´s task is to absorb gasoline fumes
venting from the gas tank while the engine is not running. The cannister incorporates
an air inlet (underneath I think?) with a one-way valve. There is a small outlet on
top that connects to a vacuum source on the inlet side of the engine... I´m not
sure exactly where.

When the engine starts, air is drawn in thru the one-way valve, thru the charcoal
picking up some of the absorbed fumes, and they get drawn into the engine intake
manifold and burned.

I also don't have the FPR hooked up to fuel, could it be that?

The FPR maintains (controls) pressure between the outlet of the pump and the
inlet of the FPR. On the outlet of the FPR the pressure drops to close to atmospheric
in the return line to the tank. The FPR is pressure compensated by a small hose that
senses vacuum in the intake manifold. The idea is to maintain a constant pressure
differential across the fuel injectors by reducing the rail pressure as the manifold
pressure drops.

Looking at this picture it looks like the FPR goes to a 23265 "Filter". That looks a
lot like the 3 hole plug on the Tank Vacuum Box. Is that right?

Yes. As you no longer have an inlet plenum to provide a vacuum signal for the FPR
the signal is taken from the vacuum box. In due course you will find that the ITB´s
do not generate a very strong vacuum signal, so the calibration of the FPR will be
incorrect. This is not a problem, as you can compensate for the difference in your
fuel map on the ECU.

Cheers... jondee86

[jondee86]

It also seems like there's a bit of liquid at the area surrounding the cylinder.

When the clutch master cylinder in an AE86 starts to leak, it is very common
for the fluid to leak to the inside of the car. Sometimes it will run down the
firewall under the carpet and make a wet patch on the floor. Clutch/brake fluid
is harmful to paint, so it would be worth confirming if the cylinder is leaking or
not. If it leaking, they are relatively inexpensive to replace. However, it does
help if you are a double jointed midget with small hands, and can stand on your
head for some time while finding the nut that hides way up under the dash

Cheers... jondee86