86 RX EFI issues .

[discopotato03]

Still behaving so alls good - so far .

My main concern was finding out if the problem was mechanical (pump/reg/filter/tank pickup/pump power supply) or engine/computer electronics gremlins .
Now that I reckon I know it was electronics the engine management system can go in with no reservations about fuel supply .

3 ECU plug distributors , yep my original diaphragm was leaking and the spare was totally shot so I had it rebuilt by that mob in QLD .

By the looks of things MPFI EA82's have fuel pressure regs that are unique and probably not available for sane money . My RX Turbo is a bit over 23 yrs old and things like reg diaphragms don't last forever . I reckon if I'm going to replace it an adjustable one it opens options and a guage lets you know whats going on .

Gotta go , cheers A .

[Gannon]

Yeah, people speculate A LOT about how o2 sensors work in a EFI system, who knows really, except maybe a JECS programmer.

Im only going off the narrowband o2 meter i had in my L series (4 plug 1989 ECU) and whenever the vac/boost gauge was showing negative pressure, it would bounce between rich and lean (closed loop mode) about 3-4 times per second (that was when it was new, it got slower as it aged) but as soon as manifold pressure went positive, it would stick at the rich end of the scale (open loop mode). I was quite impressed with this as the EA82T does not have a pressure sensor so it must have calculated the switch-over point from predefined maps

[twilightprotege]

i ran my o2 off on the way home, and just had a bit of a drive around home with the o2 on - i couldnt tell much of a difference to be honest.

[discopotato03]

Gannon from memory most probes of my cars era were single wire and quite crude . Don't quote me but I think they were a 0-1 volt device with a switch point at about 600 milliamps which represented stoi or 14.7:1 AFR .
They were only supposed to work over a narrow range to trim idle and light load mixtures at that 14.7:1 . I'm told they are very slow and most driver inputs send them out of close loop mode .

I think we can safely assume that at any notable load or on boost the system works on mapping only .
When back at home I'll have to dig out the 86 EFI system chapter because I think I remember reading that TPS and that pressure sensing gadget on the 3 ECU plug cars sends the computer into open loop mode .

The 4 ECU plug cars appear to use a much more sophisticated system inc ignition timing control . No vacuum cans or mechanical advance system in the distributor , theirs is purely a spark distribution system (cap/button) and the light sensors to tell the ECU where the crank and pistons are .
The Mass Air Sensor also tells the ECU the volume and temperature of the air going in so it can add fuel by mass to get accurate AFR's .

As for why cars like mine act up I don't know . I must look through my collection of L looms to see if the shielded oxy sensor wires have problem areas . Maybe the shielding can be broken and the probes signalls scrambled .

A .

[Gannon]

Could these issues with the o2 sensor be caused by ground loops between the exhaust and the body. Normally the input circuitry of the o2 sensor part of the ECU is designed for a 0-1v input. If there is more than 1 volt on that wire (caused by a bad ground somewhere else) it could be messing with the internal voltage regulation of the ECU. This theory could explain a lot of your issues (like the CE light coming on)

Before i sold my car, i had a funny issue with my o2 sensor and narrow band gauge. The higher the electrical load (headlights/ AC blower/ radiator fans) in the car, the lower the voltage on my o2 sensor wire and thus the lower the reading on the gauge. It must have been from a bad ground and current that normally went one place, was going another and affecting the signal

[steptoe]

there is some sort of pressure sensor on the 3 plug - the green boost light

[Gannon]

The 3 plug ECU has two pressure switches, one to activate fuel cut that connects to the ECU. I think it activates at around 10psi. The other is as you said, for the boost light, set at around 2 psi

[steptoe]

and that first one to activate, maybe it is the GO RICH dude message to ECU ?

[nncoolg]

Yeah, even though It 'only' works the light, Im 99% sure it is connected to the ECU aswel, and of course the overboost one is.

[discopotato03]

I seem to remember reading something in the 86 EFI WSM about one of those pressure switches telling the computer juice time . Have to read it when I get home .

The thing I'd like to know is what Fooj used to switch in/out close loop mode . I suspect it was either or a combination of the 3 plug cars throttle pedal microswitch and the throttle position switches .

Anyone ever find out for sure what the pedal one does ?

Cheers A .

[steptoe]

In my auto case I thought it was kickdown switch, if manual also has same it may just be an early warning for ECU to disengage the injector cut which is thought to be activated above 1500rpm and in throttle off position - part of the lesser known fuel saving advantage efi has over carby, and vapour injection LPG has over smogger devices of old

When I have run my EA82T LPG with both the round electrical connectors unplugged between body and engine,RHS, with just oil pressure and temp wires hooked up - the boost light does not work. I may have just deprived the switch of its earth supply, but dunno

[Gannon]

3 plug MPFI non turbo throttle switches make contact at full throttle, 3 plug MPFI turbo throttle switch makes contact at about half throttle.

So it must be the throttle switch that tells the ECU to go rich load/open loop.

[steptoe]

Can just see someone making up a box of 50 switches to interface betwen their ECU and rest of car to switch off or on each little ECU function to see what happens..... " right, lets tell the ECU it is all of a sudden a manual instead of fitted auto" or switch off the turbo non turbo discrimination circuit...

[nncoolg]

...just thought you guys would find this interesting...

Recently sold my Liberty and bought a MY99 WRX auto for my wife, a lovely car, but suffered from a bit of surging. Since we just got it and even though it has been serviced well in it's life, did the mandatory 'just bought car' servicing-oil,filter,tranny oil,filter,fit K&N,change coolant,2 bottles injector cleaner,diffs,fuel filter, still surgey, so....

Unplugged the o2 sensor and buzinga-no surge

The o2 sensor is in the stupid Lseries spot-in dump pipe, obviously too cold for it, even with original manifolds & heat shields.

So, yeah, you can't just unplug it as it will give CEL because of the heater circuit, so you just remove the sensor pin from the plug and bind it up, al good!

here are the results of the fuel usage:

With o2 Usage - 29.4 MPG
Without o2 Usage - 29.5 MPG

(these are both driving to/from work and around town) - typical weekly use for the wife.

I am wondering how much money subaru owners spend each year trying to get their cars diagnosed, and goodness knows what they 'need' to fix the problem!

Anyone want to go in on making 2" long 'Subaru surge-fix looms'? - a real 'my family sticker' killer!

[steptoe]

Bosch produced a leaflet designed to be issued to retail customers - it looks , pointing out that O2 sensors have a sevice life of about 80,000km. Maybe if they were replaced at these intervals there'd be fewer problems out there in car diagnostic world

[Gannon]

O2 sensors have a sevice life of about 80,000km. Maybe if they were replaced at these intervals there'd be fewer problems out there in car diagnostic world

A very good point Jono

[SCOOBIDOO]

The 3 plug ecu i ahve quite a bit of knowledge on.

To begin..code 35 will not effect the performance of the vehicle,mine has run this code for long and when the egr controller solenoid is conected to the wiring loom..the code is gone...i also know that the 2 controller over boost / and boost p/hgram activated switches have no effect on performance, one turns on the light on the dash....the other cuts injecter trigger when over boost...and do not need to be fitted to the vehicle for full performance,one actually reduces performance,they dont controll the "hey give me more fuel map" the full throttle switch on the tps does that along with a tricky word called INTERPOLATION....which really means computer precise guessing...
also the throttle switch turns off injector trigger ,no foot on throttle = no injecter trigger from 2000rpm-down to 1200 then the switch becomes redundant below 1200 and above 2000rpm...but between 2000 down to 1200 rpm..it controls the injecter pulse trigger....i have no turbo trigger / no auto trigger and no cal trigger and pull 11:1 on my wideband AEM on WOT.... even @ 15 psi...go figure???? mr fuji is much smarter than we know....

[nncoolg]

just updated the post above, results are in.

o2 sensor will remain disconnected, wife will remain happy

[steptoe]

what about a new O2 sensor to waste money on ....29 MPG out of an auto wrx ? Dat's good