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

[aceforever]

Hey guys, I stalled out and luckily a nice mechanic saw me and helped get me some tools to unbolt the fuse box.
I was able to diagnose the issue further and it seems like if I unplug the 2A plug in the engine bay, power returns.

In the electrical diagram it shows that the 2a power source is pin 7 that connects to a short pin and then a 30A fl fuse. Anyone know where that is?

Found it. The 30A Fuse for the rotors are at the top left of the fuse box with a short pin right below it.

[jondee86]

Generally speaking, if you lose power because there is a short in the wiring,
it is quite easy to pinpoint... just look for the smoke and/or flames

So losing all power to important stuff indicates a break or non-functioning
relay. Looking at the #2A plug does not show that it powers anything that is
critical to engine function. But the #2E plug is the one that powers the EFI
once the key is switched ON. Test the EFI relay and make sure it does not
have any corrosion on the pins.

Cheers... jondee86

[aceforever]

Haha I didn't see any smoke. When I was able to reproduce the electrical issue, I tried removing all the fuses and relays. I still have 4-5V from fuse link to ground.
When I unplugged #2E, it didn't work the first time I unplugged it. But later on, when I unplugged it, it gave me power again.
I also noticed that with the #2E plugged in, when I unplugged #2A, power resumed.

I reseated the two plugs in, and I now it seems to be working correctly?

My friend came over today and helped me drive the car to an alignment shop without stalling. As he was driving I was tuning on the laptop. We got -57kPa to ~12kPa and 0-4000RPM tuned so it's not lean. Around -3kPa and above it goes too lean so I raised the inj timings there some more just now. Hopefully later I can test it out.

[jondee86]

The #2A7 pin and the #2E5 are both powered up from the same red/blk wire
that originates at the fusible link block on the battery positive terminal. It
would be worth checking that you don't have a bad/loose connection at the
terminal.

AFAIK the only way you can get 4 or 5V in a 12V circuit, is to measure the
voltage between two resistances in series, that are connected between 12V
and ground. So working both ways from at the point you measure 4-5V, trace
the circuit and see what those resistances are ? It is very possible that one
will be a bad connection to power or ground.

Cheers... jondee86

[aceforever]

I definitely shorted my igniter from the bolt that holds the red/black wire and ground before when I was working on ignitor stuff. (the fuse link bolt cover was semi-cracked) I think that has to be it. I'll check the resistance on the those circuits.

I was able to hack a way to log O2 sensor 0-5V wideband sensor into the ECU using the Trim signal in to record the voltage, and it's logging all the data I need to make changes! I also noticed that I may be able to use the wideband 5V directly as 02 sensor since there's an option to choose between 1V and 5V in the O2 sensor setup.

While my friend was here, we worked on bleeding the clutch and boy it grabs so much sooner now from when I let go of the clutch! It's pretty awesome! Before, it grabbed around 20% clutch pedal depress, now it catches around 70% clutch pedal press. There was a ton of random crap in the old brake fluid that was coming out.

[aceforever]

I'm currently in the process of wiring up the wiper motor and I'm facing some issues. It looks like I'm always getting 11.9 voltage in the wiper motor wires always. I traced it down to the wiper and washer switch. When I opened up the back a little, it looks like there are some brown stains. I think it is corroded and who know's what else is corroded. Is it worth taking apart and fixing up? or would I keep on getting issues?

I found the part for around 55$ on some site. I was wondering if there is an alternative switch I could use from a different car? Thanks!

[jondee86]

For the moment I think you would be better off just squirting some WD40
inside the switch and working it thru its range of operating movements about
50 times. With a bit of luck that will clean the contacts enough to allow you
to continue to use the existing switch. 11.9 volts is close enough to 12volts
and that is the nominal operating voltage, so the washer and wipers should
work OK.

Then put the wiper switch assembly on your list of things to look out for at
swap meets or junkyard visits. Sooner or later you will pick one up for just
a few bucks. Keep it as a spare and use it when the existing one dies.

Cheers... jondee86

[aceforever]

What I mean is, in the off position, something is short circuiting so that the motor circuit gets 11.9V. I'm currently taking the switch apart to diagnose the issue.

EDIT: I took apart the switch assembly and created a circuit diagram for the wiper and washer switch w/ int to help with the diagnosis.

[jondee86]

I found this...
http://forums.toymods.org.au/index.php?t=msg&th=13120&&pl_view=&start=0#msg_106182

And this, which is interesting... and I want it
http://www.dorikaze.net/showthread.php?9154-Got-my-JDM-Variable-Int-Wiper-Switch

And this which is for wipers with the intermittent function...



In various places I have seen reference to a "wiper relay" and it is shown on this
diagram. But the way it is drawn leads me to believe that the "relay" components
are actually contained within the base of the switch. The diagram for the switch
without intermittent function is a little different.

Cheers... jondee86

[aceforever]

Thanks for looking! The first link looks like for a headlight switch and the circuitry there is quite different as the Wiper Switch (at least the one with int) is more complicated.

I saw the 2nd post as I was searching before, that variable INT switch is discontinued everywhere I looked. If you find one though, you know who needs one Hint: ME

Here is my progress so far on repairing the one that came with the car.


Like my homemade 12V power source? haha. I ordered a breadboard and some jumper wires to help test this out. It's coming on Thursday.

EDIT: another question. I currently have the trigger angle to be 21 BTDC, however when I was talking with a Haltech rep sometime ago, they recommended me to go with 80-90. Should I reseat the distributor gear and make it 80-90? Since currently we're only using an ignition advance under 20 BTDC, it should be fine for now. I'm thinking in the future, when we need over 21 BTDC, would the ECU have trouble sending the timing signal to the ignitor?

[jondee86]

Hahahahahah I was searching for wiper switch and that popped up. It
looked like a wiper switch and I didn't pick up the fact that it was a light
switch Never mind... it was a bit interesting any way.

It also reminded me that there is only one washer pump, and that works
for both the front and rear wipers. When you operate the dash switch for
the rear washer, it energises the washer pump plus a flip-flop valve that
re-directs the flow from the front washer nozzles to the rear one.

I think way back I suggested 50-60deg as you will not have any reason to
go beyond 40deg BTDC total advance. I guess 80-90deg would also work.

Cheers... jondee86

[YoShImUrA]

Hi Jondee, actually, some cars (like my 1986 USDM hatch) came with not one but two motors for the washers.

Speaking of this, I once acquired an adjustable intermitent speed (slider? potentiometer?) washer button from a RHD car.

I've always wondered if there was a way to make the LHD one work with that potentiometer. I never found anyone that had done it but I'm curious... Just asking if you know anything about it

Sorry for the slight jack thread...


Sent from my iPhone using Tapatalk

[jondee86]

Ahhh.... I did see some stuff about two washer motors, but kind of
by-passed it, as my "Euro Spec" cars only ever had one plus the diverter
valve. If by slider you mean the rotary bezel that surrounds the wiper
switch, it should be possible to make it work.

That is pretty much the conversion that is described in one of the links
I posted above. The intermittent function basically just switches the low
speed motor on and off with a small timer circuit. AFAIK the wiper motor
is the same, just different switches.

I must have a look in my parts box and see if I have one lying around

Cheers... jondee86

[aceforever]

I think way back I suggested 50-60deg as you will not have any reason to
go beyond 40deg BTDC total advance. I guess 80-90deg would also work.

Back then I didn't know I could reseat the distributor gear so that it doesn't match the OEM suggested alignment. I'll seat it to 50-60 Thanks! (btw I feel very comfortable about syncing up distributor and ECUs now! It makes me feel pretty good)

[aceforever]

So my breadboard came. I managed to burn out an LED while testing without the switch. Apparently you need at least a 1k ohms resistor in an LED circuit After I tested lighting up an LED to see if I do have power. I connected up the switch with two other LEDs to indicate slow and fast motors.

https://www.youtube.com/watch?v=auip6wCN0Rw

The connector for the wiper motor is totally different than the harness connector. So I cut up the wiper motor and put the flat single connectors on. Heatshrinked around it, and then finally electrical tape all around the connection after I tested everything. Seems like it works

Here's a video of it in action:
https://www.youtube.com/watch?v=aR8InxmsSsY

[jondee86]

btw I feel very comfortable about syncing up distributor and ECUs now!
It makes me feel pretty good

Yes... I'm sure it seemed rather daunting when you first pulled the ECU out
of the pizza box But a few months of DIY and learning with a bit of trial
and error, and now you can feel the satisfaction of having done something
that took you outside your comfort zone for a while !!

Cheers... jondee86

PS: What was wrong with the switch ?

[aceforever]

It was shorted at the circuit board due to some rust. Scraped that off with a screw driver and it works again.

Yes... I'm sure it seemed rather daunting when you first pulled the ECU out
of the pizza box But a few months of DIY and learning with a bit of trial
and error, and now you can feel the satisfaction of having done something
that took you outside your comfort zone for a while !!

Haha yes! Now I'm not only fixing cars, I'm fixing stuff around the house too! Just took apart the dishwasher to figure out what was causing lots of particles on the dishes A lot of unscrewing later,I found the filter with tons of gunk to be cleaned off.

[aceforever]

Guess what came in the mail today!



I got too excited and didn't order a baseplate and oil. If I was doing this for a profession, I would absolutely go bankrupt because I forget parts.

[aceforever]

So today, I was trying to sort out the speedometer issue.

At the instrument cluster side of the speedometer cable is a square bit.


At the gearbox side of the speedometer cable is a metal bit.


This metal bit goes into this housing that is hooked up to this plastic gear


As you can see my gear is very worn and that's probably why it isn't working. I'll go ahead and order a replacement since everything else was connected properly. I even checked to see if the metal bit transfers the spinning motion correctly to the plastic bit. Also tested it with the instrument cluster hooked up. Spinning the metal bit does indicate speed.

The problem is either the plastic gear being worn out, or some other issue with the gearbox. Either way, simplest way to test is to buy a replacement.

P.S. For future adventurers, when you pull out the plastic gear housing transmission fluid will leak out! Make sure you have a paper towel or rag handy to stop the fluid!

[jondee86]

The plastic gears have a few options on the number of teeth...



There should be a number stamped on the end... this one says 20.
Make sure you get the correct one or your speedo will not be reading
correctly. The other gear inside the gearbox is steel, so it will chew
up the plastic gear if the cable does not turn easily. If you can spin
it with your fingers you are good.

Cheers... jondee86

[aceforever]

Thanks for reminding me. I've read up on these speedometer driven gears and in the US, there are two variations. 20 teeth for SR5 rear-end and 19 teeth for the GTS rear-end. The reason why there are two is because SR5 has different final gear than GTS rear-ends. I've made sure that my rear-end is the GT-S version by checking on the number of bolts on the drive side of the diff. GTS has 10 bolts, whereas the SR5 has 8bolts.

I've decided not to take out the gear out again and just go with 19, since that is supposed to be the right version for my car. Transmission oil is nasty stuff!

[aceforever]

I also got a chance to play with the horn / centering steering wheel.





I took apart the aftermarket steering hub and took a look. It seems like the positive track isn't connecting with the +12V coming from the horn. The hub adapter's positive track is too wide. Also found out that the center spline/bolt wasn't grounded. Not sure how it looks like on a normal hub. Will need to engineer something to ground the center. And probably look at getting a different hub adapter.

Good news is, I managed to center the steering wheel as much as I can.

[aceforever]

Looking at ignition maps and was trying to see what ignition timing does on an engine and what the relative change in ignition timing is when you go higher rpm and higher load.

Not sure but looking at some sample ign maps, it seems to me like higher load means more retard in ign timing, and lower load means more ign advance.
Higher RPM usually means more ignition advance but it varies less so than load.

From: http://www.aeu86.org/forum/showthread.php?tid=11876


From: http://forums.club4ag.com/zerothread?id=74402





And from what I understand based on information from videos and forums. Ignition timing greatly controls the power output of the engine. It also can cause rod knock (and I presume other bad things?) if tuned incorrectly.
The injector timing, to a certain extent, doesn't affect power output as much if kept within a range near 14.7 AFR.

In my case, I think tuning the engine so it's not running too lean or too rich for the highway drive to the tuner would suffice. Please correct me if I'm wrong.

Another question about going lean. Since there's always going to be a little latency in the ECU trying to compensate for extra air when changing throttle input quickly. Is it normal for AFR to go lean for a brief moment before it catches up and I get more fuel? What's the desired behavior here? I found that throttle pump settings only helps somewhat and cause the AFR to go rich after a brief moment after the initial leaning.

[jondee86]

Here is a nice ignition map that fits in with my personal opinion of what
an ignition map should look like



Don't worry about the numbers on the axes as I don't know what engine this
map is for, and it doesn't matter. What is important is the shape of the map.

1. Because the combustion process needs to reach peak cylinder pressure at
a specific number of degrees after TDC for best torque, it follows that as the
engine rpm increases, the spark must be initiated earlier to allow time for
the combustion process to take place.
So we can say that ignition advance increases as rpm increases.

2. As the fuel load increases and the mixture becomes richer, the flame front
travels faster and the combustion process completes more quickly. This means
less advance is required when the engine is under heavy load.
So we can say that ignition advance decreases as load increases.

So you are correct in your understanding. And these two conflicting requirements
result in a ignition map shaped as shown above. Of course there are many ways
of making these simple rules seem complex and out of reach of ordinary mortals
However, countless millions of cars happily ran for billions of miles with a simple
mechanical distributor that used centrifugal weights to advance the ignition with
rpm and retard it with vacuum... no electronics required !!!

By using a sophisticated dyno it is possible to establish exactly the right amount
of advance for best torque at any particular combination of rpm and load. This
process will produce a "peaky" 3D map as shown in your post. But I have proven
that a car will run perfectly well with a basic "smooth" map, very much like the one
shown above.

Cheers... jondee86

[jondee86]

In my case, I think tuning the engine so it's not running too lean or too rich for
the highway drive to the tuner would suffice. Please correct me if I'm wrong.

The ignition timing you already have programmed should be fine. And 14.7 AFR is
also fine for general round town running. You only need to start dialling in a bit
more fuel at big throttle openings and high rpm's.

Another question about going lean. Since there's always going to be a little
latency in the ECU trying to compensate for extra air when changing throttle input
quickly. Is it normal for AFR to go lean for a brief moment before it catches up and
I get more fuel? What's the desired behavior here? I found that throttle pump
settings only helps somewhat and cause the AFR to go rich after a brief moment
after the initial leaning.

It is likely that the E6X cannot inject extra fuel quickly enough to stop the lean
spike that follows a rapid throttle opening. More modern ECU's will inject the fuel
asynchronously,... that is to say, immediately, without waiting to add it to the next
scheduled injection event. If the extra fuel arrives late, then you will get a lean
spike followed by a rich spike.

I think the two controls you have available set the amount of fuel to inject, and
how quickly to inject it ? The "gain" control most likely responds to the TPS rate
of change, so if you still have some lag in your TPS response, that could be adding
to the "lateness" of the throttle pump action.

Cheers... jondee86

[aceforever]

Thanks for walking me through some of the theory of ignition timing. I'll stick with the ones we programmed on for now and see what the tuner comes up with. I'm pretty interested in the theory behind the values. I'm also wondering what the process is of tuners when they do tune the ignition. I've heard from some places that you tune up until you hear a knock and then detune it down a bit. My ECU doesn't have a knock sensor so I presume the tuner needs to do it by ear? Or perhaps use a 3rd party knock sensor and hook it up to one of the connections on the ECU for data logging?

I didn't realize that the throttle pump correlates to the TPS values! That means it's very important for me to get the TPS signal to not have that delay issue. Thanks for the tip. I'll play with that. If even after that I can't get the ECU to inject extra fuel quickly, is it okay for that lean spike?

My base plate for the ITG filter came today. I'm trying to find a place to drill the holes out. I've made marks for the center of the holes and just need the machinery. Wish me luck!

I've also installed new wipers on the car. 17" and 19" RainX Latitudes work really well. I got them because came with a bunch of adapters and I wasn't sure which connector the car had. Apparently it's the Small J-Hook for future folks.

[jondee86]

Any decent tuner will have a "knock box" which will allow him to listen
for knock with headphones while tuning. Typically the system would consist
of a donut style knock sensor attached to the block using any suitable 8mm
threaded hole, an amplifier with selective noise filtering, and headphones.

If the dyno allows the engine to be held at a constant load and speed, the
tuner will adjust the timing to achieve best torque. He will do this at a number
of points and use these as a basis for your ignition map. Best torque will
usually be several degrees before the point where knock can be heard. If the
dyno does not allow for constant load holding, the tuner will make his best
guess at what advance you require. The difference in results between these
two approaches can be substantial.

The factory ECU will use the TPS to monitor the position of the throttle plate
to determine when the engine is at WOT and at idle. It uses the rate of change
in the position of the throttle plate for the throttle pump function.

A momentary lean spike will not harm the engine. Neither will a rich spike.
All that happens is that for that moment you have less than ideal combustion,
and the engine can tolerate that. I had exactly the same problem for some
time until I found the correct settings for the throttle pump. ITB's need a BIG
shot of extra gas when the throttle snaps open at low rpm, but little if any
extra gas after 3500rpm.

Yeah... good luck with making those holes. I did mine with a cheap holesaw
designed for cutting drywall. It was ugly

Cheers... jondee86

[aceforever]

As always, thanks for your knowledge and wisdom!

I've been reading the E6X manual and it shows this setting:

Throttle Pump Dead-band
This field defines the percentage change in throttle position that must occur before the
throttle pump is activated. This feature allows for “jitter” in the throttle that would
otherwise over-fuel the engine. The valid range of values is 1-20%.

I think it's to compensate for TPS that bounce around when you're holding at a certain position. I don't know what the defaults are at this time. But I feel like this number should be pretty low for me.

Also something neat for when I face the atomspheric MAP. Seems like it's only good if I face atmospheric after 70% as that's the minimum

Full Throttle Map
This feature allows the user to adjust a special fuel map that is used only when the
throttle is wide open on normally aspirated engines. With some manifold and or
throttle designs, pressures in the manifold can reach close to atmospheric pressure
before full throttle is applied. This effect can make tuning difficult around full
throttle. This map allows the full load settings to be easily set without interfering with
lighter load settings. This field enables or disables the use of this map.

Full Throttle Threshold
This field defines the throttle position at which the ECU considers to be full throttle.
This field can be set between 70 and 100.

[jondee86]

Throttle Pump Dead Band
This appears to be a way of "de-sensitizing" the throttle pedal so that
accidental small movements don't activate the throttle pump function.
In many driving situations, providing the throttle is used gently to increase
speed, there is no need for extra fuel. The change in manifold pressure
will automatically move the fuelling to a richer point on the fuel map.

If we use the original name of accelerator pump it makes it clear
that this function was designed to help avoid lean stumble when the engine
was being asked to accelerate the car. Sudden acceleration would call up
a big squirt of fuel, while slow acceleration would get little or no extra
fuel from the pump.

Full Throttle Map
This function addresses the major shortcoming of ITB's when used with
speed-density tuning. Because the MAP reading reaches atmospheric before
the throttles are fully open, it is difficult to tune the top end of the fuel map.

My first attempt at tuning was to use SD only tuning, relying on the MAP
signal to determine how much fuel to provide for a given RPM. My reasoning
was that in the low to mid RPM area of operation, the 15% to 20% MAP range
was greater than the 5% to 10% throttle range that I would be using. Since I
wanted the best fuel control possible below about 3500 RPM this seemed
like a good choice.

I configured my VE table based on recommendations on the MS forums so
that over half the table was dedicated to the MAP region above 80%. In fact,
I split the MAP range from 92% to 100% into 5 rows of 2% increments.This
method worked ok at lower RPMs and small throttle openings. The mixture
was maintained fairly well under these conditions.

Over about half throttle or high RPMs this method was not ideal. In the upper
RPM portion of the table, I had more than 10% changes in fuel required from
just a 2% change in MAP. This made the tune incredibly sensitive to noise on
the MAP signal. As a result, the VE table would bounce around several bins
creating inconsistent fueling and mixture.

Here is the last VE table I used for SD only tuning. This was run on the 45mm
ITBs with my 2.0L engine.

VE table for SD only tuning. Notice that 2/3 of the table is dedicated to the
MAP region of greater than 80% with almost all VE increase occurring only in
the upper 15% of the MAP range.

A 3D view of the same VE table. This view really shows the nature of the ITBs
relative to MAP. You are trying to tune an engine using only 15% of the available
MAP signal. Notice the steep slope of the curve where large fuel changes are
made over just a 2% change in MAP signal.

I have copied and pasted the above as it illustrates very well the problems that
can be encountered when tuning ITB's using the MAP sensor for load input. It is
not impossible to tune if the ITB's are correctly sized for the engine, and there
is a stable MAP signal. But if the ITB's are over-sized and the engine has bigger
cams than stock, a TPS tune is a much better option.

The default value for the Full Throttle Threshold is 95% so effectively what the
Full Throttle Map is designed to do is ensure that you have safe fuelling when
the engine is at WOT. This is not unusual as the stock ECU runs in open loop at
WOT for the same reason.

Cheers... jondee86

[aceforever]

Thanks for walking me through tuning for WOT. I'll have to play around with it after I get the air filter on.

Update:

I was able to drill out the holes for the baseplate. However I did drill one screw hole a bit off and now it might leak air in. I'm thinking of using gasket maker to seal that up. Would the vacuum suck that in? I feel like if I shape the gasket in the shape of a grommet, there should be enough material that it wouldn't be able to pull the gasket in.




Currently taking TPS apart to see if I can clean out the contacts.