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

[aceforever]

Ahh good to know thanks for that tidbit YoShImUrA!

[aceforever]

For best stock belt performance/safety unbolt the belt assembly at the base of the b pillar, pass it through the base belt hole on the left side of the seat, and bolt it back up to the b pillar. pass the buckle through the base belt hole on the right side (it might be easier to remove plastic sheath for this) - it's how oem's do the belt on cars equipped with fixed back seats that have high side bolsters at the base.

Followed what you recommended. Went out for a test drive and it seems to fit very nicely. Thanks!


The cold weather makes starting the car a bit hard. I knew from before it takes two starting attempts to start the car. So I adjusted the priming map to double the amount it had, and then now it starts up with one attempt to start. Surprisingly simple.

The car was also running quite a bit rich during the warm up (10-11 AFR) I adjusted it to more like 13 AFR. I think this was what partially causing fuel in the engine oil prior.


Here's a video sample of it revving quite happily. Before it would shudder for a bit.
https://www.instagram.com/p/BA7-NZil5-b/

[Deuce Cam]

I never tried looping the belt through the upper hole like that. I haven't seen it done. Not sure if it's a good idea.

I meant loop it through the hole on the bottom left side of the seat on the leg bolster (where the lap belt would go on a harness). Also pull the buckle through the hole on the right side leg bolster next to ebrake lever (where the other side of the lap belt would go on a harness).

[aceforever]

Ah, okay. I did route the belt through the bottom left hole. And slip the belt through the other side. I'll redo the belt so it's not in the upper hole.

[jondee86]

Looking at the belt path, I would say that the belt would be more effective
passing thru the top hole as it shows in your pic. The belt will be laying directly
onto your shoulder, whereas if the belt goes around the outside of the seat, you
would need to move forward some distance before the belt gripped your upper
body. Routing the belt thru the lower LH hole and the buckle thru the RH hole
will make sure that you are held tight and low.

Just throwing that in for discussion purposes

Cheers... jondee86

[YoShImUrA]

I'm with Deuce on this one. No manufacturer has done it like that, and for a reason. The belt would literally choke you on certain decelerations, especially if you're short as you claim, the way it's put on the pic above. I'm not even talking about a crash situation. The belt has to hold you from the top of your shoulder, not leaning under any circumstance from your neck.


Sent from my iPhone using Tapatalk

[aceforever]

It does feel a bit tight even without any movement. I'll try both ways and make a decision.

[jondee86]

It does feel a bit tight even without any movement. I'll try both ways and make a decision.

Seems reasonable To be clear, I'm in no way suggesting that you should use
a seatbelt in a manner that may be dangerous or less effective. I'm just looking
at the simple mechanics... where the seatbelt should be located to best restrain
forward movement in the event of a sudden stop.

Firstly, the belt needs to be in close contact with your upper body so that there
is no "free play" to be taken up before the belt starts to restrain movement. Ideally,
the diagonal belt should pass over the center of mass of your upper body to reduce
any tendency to roll over or under the belt when load is applied. It is no good if your
left shoulder is restrained, but your head and right shoulder can roll forward over
the belt. Certainly, if the belt was pressing into your neck while driving, it would
be uncomfortable and possibly dangerous.

Comparing your pic and this one...



... it can be seen that the design of the seat and the location of the anchor points
will also have a bearing on what is comfortable and/or safe. I've not looked at a car
factory fitted with fixed back buckets and 3-point belts, but I think they probably
don't use the upper seat holes because they have no control over who will be sitting
in the seats. And of course there may be other technical or safety reasons that
I don't know about. And because... everyone likes pics


Truckers seat with integrated belt...


Load testing seat belt...



Cheers... jondee86

[YoShImUrA]

Lol!

You can check how Porsche does it with his bucket seat interiors, like on the GT3, GT2 cars.


Sent from my iPhone using Tapatalk

[SgtRauksauff]

Before routing the seat belt, sit in the seat and find out how the belt will fit on an actual person, preferably the one who will drive it.

In my '85 hatch, I've got a Corbeau Forza II seat, bolted to some flat stocck that's right on the floor. The seat does not slide. I have a stock seatbelt on the outside of the seat, and do not use it through any of the holes in the seat. However, I do run a 5-pt harness when autoXing or tracking the car. If I were to route the stock seatbelt through the seat at all, I would route the lap belt through the seat hole, but leave the shoulder belt portion outside of the seat.

In a normal harness, the shoulder straps go straight back, or slightly down to the anchor points. In a crash, they do not really have any effect on the seat itself, but rather hold the person down and into the seat, towards the floor, in a straight front-to-back orientation.

When using a stock harness, at least on these cars, the anchor point is routed through The b-pillar loop for the seatbelt, at or significantly above shoulder level. In a crash, the person in the seat would have the shoulder to the outside of the car restrained more than the inner one. Their torso would rotate a bit, but as the pressure builds on teh torso strap, the waist strap is tightened, holding them down and into the seat.

If you were to route the shoulder belt through the upper hole in the seat, in a crash, the tension on the seat belt will tighten against the waist as normal, but it will also pull up and forward on the seat. in a hard enough crash, it could twist/damage the seatback.

[aceforever]

Thanks for the advice everyone. It's wonderful how much information you all have

Just received a package with these CF popup headlight covers. I bought these from Retro-spec made in the US. They look pretty good. I did find a little scoff on the side of one, but I can probably buff it out. I ordered back in November, and now it finally arrived. Had to email them a bit for them to get me with shipping information. Seems like either they are very busy or a bit off game there.




The backing looks like some kind of fibrous material.

[aceforever]

Suspension update:

While lifting my car up in the air. I measured the wheel center cap angle and the angle of the strut with my phone's level tool.

One side gives me 5deg and 20deg

Other side gives me 0degrees and 15degrees.

While on the ground both wheel center caps give me 5degrees. Have not measured the struts yet.

[jondee86]

That ^^^^ information would seem to indicate that the angle between the stub
axle and the strut is the same on both sides = good And if the wheels have the
same same angle when the car is sitting on the ground (did you roll it back and
forwards a bit before measuring ?) that also = good. But 5 deg is rather a lot of
camber for road use = not so good

Does one side hang down lower than the other when the car is in the air ?

Cheers... jondee86

[YoShImUrA]

If the car's been corner balanced, I'd consider a difference in height from one side to the other normal, as they have been adjusted to a different height.

Did you end up getting the RCAs?

[Deuce Cam]

Interesting. It's good the angle difference seems to be the same on both sides, I wonder how accurate the app and method for measurement is though? Wasn't the front camber -4.5*/-4.7*?

Even without a corner balance there can be a difference in height due to the weight distribution. An example is the stereotypical left rear lean on a fixed height rear suspension setup.

I'm also curious if one side hangs down more than the other when hanging in the air, but it might not tell the whole story since, iirc, the camber plate locations were on opposite ends of the spectrum on each side. I'd also be interested to know the total expanded length of both the strut/shock/spring assemblies. Can you verify the springs are the same on both sides? Either by markings, or by measuring and comparing the overall length/thickness of coil/# of coils on both.

[aceforever]

Does one side hang down lower than the other when the car is in the air ?

I'll have find some time to do that. What are we looking for here?

Did you end up getting the RCAs?

Nope, not yet, I'm trying to figure out why the settings are different first, before spending money that I might not have to.

I wonder how accurate the app and method for measurement is though? Wasn't the front camber -4.5*/-4.7*?

The app doesn't know negative and positive. But I can confirm they are negative. So -5 degrees, so not far off.

I'm also curious if one side hangs down more than the other when hanging in the air, but it might not tell the whole story since, iirc, the camber plate locations were on opposite ends of the spectrum on each side. I'd also be interested to know the total expanded length of both the strut/shock/spring assemblies. Can you verify the springs are the same on both sides? Either by markings, or by measuring and comparing the overall length/thickness of coil/# of coils on both.

Sure I can check out the spring thickness and length when the car is up in the air. The drive side is definitely riding higher up when the car is on the ground.

[aceforever]

Tuning Update:

So I did notice when I'm coasting down hill, or even going up hill on partial throttle, I'm getting a rich condition. I think I need to data log better to figure this stuff out. Any tips here?

I think there's a way to make the Haltech Closed loop system work, should I just get that up and running and not worry about the fuel map being off?
What about decel fuel shut off?

[jondee86]

Does one side hang down lower than the other when the car is in the air ?

I'll have find some time to do that. What are we looking for here?

When the car is on the ground the height is set by the springs. But when the
car is in the air, the droop is set by the shocks.

Now, to get 5deg difference between one side and the other, there would need
to be a BIG difference in droop (and I don't think that this is the explanation).
But if there was an assembly difference... different length inserts, inserts with
different stroke, one insert with a bottom spacer and one with a top spacer etc,.
then you could have some difference in droop. And that would be part of the
explanation.

Cheers... jondee86

[jondee86]

What about decel fuel shut off?

Yes, you should implement this feature. The manual seems to indicate that the
fuel cut is active on closed throttle above your nominated rpm threshold. I assume
that this is the "fuel restore" threshold ? By comparison, the factory ECU cuts fuel
on closed throttle (and high vacuum) above 1700rpm and restores fuel when the
engine speed drops to 1400rpm.

As you have a high idle, you will need to try setting the fuel restore threshold
300-400rpm above your warm idle speed. This should mostly take care of the rich
condition on deceleration.

A rich condition under load on low throttle just indicates that you have too much
fuel dialled in on that part of the fuel map. Generally speaking, it is best to get the
fuel map as close to optimum as you can, before switching to closed loop. You do
not want the ECU to be making big percentage changes as this can create its own
set of problems.

Better to log TPS, RPM and AFR on an uphill grade where you hold the rpm reasonably
steady to reduce transient changes in AFR. Then you can get a better idea of where
to make changes in the fuel map. It is a tedious process, and if you don't have some
decent uphill grades close by, you might find that it is more cost effective to do the
tuning on a dyno.

Cheers... jondee86

[aceforever]

Now, to get 5deg difference between one side and the other, there would need
to be a BIG difference in droop (and I don't think that this is the explanation).
But if there was an assembly difference... different length inserts, inserts with
different stroke, one insert with a bottom spacer and one with a top spacer etc,.
then you could have some difference in droop. And that would be part of the
explanation.

Would the camber plates have anything to do with it too though? Since one is set to near max positive, another set to near max negative?

[aceforever]

As you have a high idle, you will need to try setting the fuel restore threshold
300-400rpm above your warm idle speed.

Currently warm idle speed is 1500-1600, so about 2000rpm should be good for now then.

You do not want the ECU to be making big percentage changes as this can create its own
set of problems.

I'm definitely going to take it into a professional at some point after I get the throttle bodies checked out. But prior to that, I'd like to head to meets and drive around without having to worry about dialing in that portion of the map. Is it safe to assume that getting closed loop running for the time being is better? What's the disadvantage in having big percentage changes?

Thanks!

[aceforever]

CF headlight cover fitment update

They're actually very well done. I did notice that the driver side popup headlight assembly was definitely damaged before. Luckily I have a replacement assembly on standby.

[shagymc]

Headlights look great. As far as the closed loop is concerned, you should find out what percentage the ECU actually is allowed to change up to. I have the sprint 500 and it allows up to 15%. Hooking it up is easy but you'll need a wideband O2 sensor. I am currenty running the innovate motorsports version and it's great. There's an output wire tht goes directly to the wideband input of the ECU and that's it. You'll still need power and ground still obviously but as far as wiring is concerned nothing to difficult! Keep up the good fight!

[aceforever]

Headlights look great. As far as the closed loop is concerned, you should find out what percentage the ECU actually is allowed to change up to. I have the sprint 500 and it allows up to 15%. Hooking it up is easy but you'll need a wideband O2 sensor. I am currenty running the innovate motorsports version and it's great. There's an output wire tht goes directly to the wideband input of the ECU and that's it. You'll still need power and ground still obviously but as far as wiring is concerned nothing to difficult! Keep up the good fight!

Thanks for the information! I'm currently running a PLX controller with BOSCH LSU 4.9 sensor to (the best of my ability) tune the fuel map. When I read the manual it said only narrowband O2 sensor could be hooked up to the ECU, however I do see an option to use wideband in the software. So while hooking up the O2 controller to ECU is relatively easy, I'm not 100% confident on if it's a good idea or not.

[jondee86]

What's the disadvantage in having big percentage changes?

From the E6X manual...

Note: It is preferable to keep the increase and decrease limits small (say around
5-10%). Excessive swinging of the air-fuel ratio can result in surging and poor
operation of the catalytic converter. The closed loop algorithm should never be
used as means to correct bad mapping.

From my brief look in the manual it appears that the E6X is only capable of running
closed loop with a narrow band O2 sensor. This means that it is limited to bringing
the AFR to (or close to) stoich. And as the manual states, stoich is fine for cruise
but not when the engine is under load... for example, accelerating hard in higher
gears or pulling up a grade.

The Haltech advice is to set upper and lower rpm limits for closed loop operation,
so that the ECU drops back into open loop under high load (rpm) conditions. These
limitations are inherent in older ECU's such as the E6X. Modern ECU's have much more
flexibility in closed loop, and usually will have an auto-tune function as well. The ECU
will compare actual AFR to preset AFR targets, and adjust the fuel map accordingly.
And of course they will accept a WBO2 input to allow correction to any preset AFR
target, not just to stoich.

However, if your WBO2 has the ability to output a simulated narrowband output
(and I'm sure it will) you can use this as the input to the E6X.

Cheers... jondee86

PS: I have only looked in the manual, so if software says you can use the E6X with
a wideband input... disregard the above

[aceforever]

Seems like modern ECUs can get you pretty close to a well tuned map without manually tuning it. It's a pretty good idea. I wonder why earlier ECUs didn't feature it. I suppose most people are going to maximize the tune anyways with an aftermarket ECU back then.

Looking through the manual more:

O2 Sensor Type
The type of sensor used should be selected here. Choose the sensor that best describes
the sensor you are using. It is also possible to attach a 5 Volt sensor, such as a UEGO
probe, to the E6X. Since these sensors are expensive, and have limited life, it is
unlikely that they would be used in general running of the car for closed loop feedback
control, but rather as an aid during tuning

So it is possible, but not well documented. Like it doesn't mention what min temperature the setting for activation is. I'm guessing coolant temperature? The default is 63C, which seems pretty high for air inlet.

Here's some settings I mocked up for use later. Most of it is defaults, only change is O2 Sensor voltage and sensor type settings to adjust for wideband. And lower maximum increase/decrease at +/- 10%.

PLX controller sends a voltage based on this formula AFR = Voltage * 2 + 10. I reversed the formula, Voltage = (AFR - 10)/2 to figure out what voltage is necessary. (14.7-10)/2 = 2.35V = 2350mV

There's no in-depth AFR target map like you mention. Man I kind of want a new ECU now just for auto tuning fuel maps. But boy would that be a bunch of money and work. Once my older setup is all set and done, I'm guessing it shouldn't require maintenance and tweaking, so the benefits of getting a new ECU just for that doesn't seem justified.

And fuel shutoff at decel is pretty easy to implement, not many settings there.


So with this setting on, I should see AIR/LEAN on the AFR gauge when decelerating.

[aceforever]

Also found this:

A.9 Fuel Consumption
Poor fuel consumption is a result of a too rich mixture. If the fuel consumption in traffic is
poor but the highway consumption is good, then it could be the areas of the map used for
accelerating that need to be leaned out. The throttle pump may also waste fuel if its values are
too high. Also try using the fuel cut off on deceleration feature.
If the city consumption is fair but highway cruising uses too much fuel then lean out the light
load bars on the 2000 rpm to 3000 rpm maps (This is where most cruising takes place).

The manual explains the exact problem I was mentioning. Fuel richness at cruising and gave tips on how to optimize it.

[jondee86]

Aftermarket engine management keeps evolving and getting more complex as
it tries to keep up with the increasing complexity of factory engine management.
One of the advantages of having an older car is that the electronics are quite
basic, and the ECU only has to deal with what is necessary to run the engine.
So the E6X can do the job, only the interface is a bit clunky by today's standards.

Autotune is great for getting a basic fuel map quickly. But to fine tune the map
and tune the high rpm/WOT part of the map, you still need manual intervention
and some dyno time. I used to spend a lot of time making logs and using them
to make small adjustments to the fuel map. My aim was to keep the corrections
made by the ECU to less than +/- 2% of the map values. That way if the WBO2
sensor died the engine could run just as well in open loop as it did in closed loop.

Looking at the information you posted above, I can see that the WBO2 sensor
can be used, but the ECU still only targets stoich (or whatever voltage you put
in the box). This is effectively exactly the same as using the simulated narrow
band output provived by your PLX gauge. The minimum temperature setting
is to make sure that the engine is up to a temperature where your various fuel
enrichments corrections have tapered down to zero.

Cheers... jondee86

[aceforever]

So the E6X can do the job, only the interface is a bit clunky by today's standards.

My day job is writing software programs, and it saddens me deeply when I use the haltech software. Other than the antiquated UI design, which itself is very retro. It had bugs in the program too. There was a problem with the 3d graph fuel map window where it wasn't displaying the current TPS/RPM when using Alpha-N, so it was incredibly hard to tune. That's another reason why I went with speed density instead.

My aim was to keep the corrections
made by the ECU to less than +/- 2% of the map values.

I like that thought process. I'll try to achieve the same.

I used to spend a lot of time making logs and using them
to make small adjustments to the fuel map.

That's a good approach. At least the E6X is new enough to have data logging.

This is effectively exactly the same as using the simulated narrow
band output provived by your PLX gauge.

Yes I agree. However I guess since I can target a specific voltage within 0-5V, it's just a tiny bit better. But yeah, it's effectively the same thing.

Hood update:
Took some time to apply the heat reflective tape. Personally I think it looks pretty nice. Probably a bit overkill but I did this as a precautionary measure to keep the top coat of the CF from dulling out or yellowing from engine heat. I heard mixed stories of whether it was okay not to protect it, so I'm playing it safe.

[YoShImUrA]

Looking like a McLaren F1! Haha!

So interesting the ecu stuff. I only wish I knew what the hell you guys are talking about. Lol


Sent from my iPhone using Tapatalk