EA82 SPFI injector question

[aiiadict]

EA81 (rebuilt), in a VW bus.

 

VW carb was ugly, FI was cheap. So I stuck it on the engine.

 

Wired it up. Modified the Dist. to fit. Flipped the IAC over so it

doesn't bump the dist.

 

verify timing..

 

Ok.. I verified timing is good. crank a bunch, no start.

 

good spark. pull plug wires, put plug 1 wire a bit away from case, turn engine by hand... good spark right at TDC.

 

ok.. try to run again. No action.

 

 

next, check fuel.

 

 

pop off the intake boot, turn it over. No spray into the throttle

body.

 

 

I'm looking at the schematics, and I can't see how fuel gets into idle

circuit. with throttle valve closed, the only air entering the manifold

is through the bypass valve. the inlet to the manifold is below the

throttle valve. The only place I see that fuel enters is from ABOVE

the throttle valve...

 

 

how does gasoline enter manifold at idle?

 

 

ok, so I'm messing with the throttle body.. put power to START on

ECU, fuel pump turns on..

 

 

no spray

 

 

hold start on (about 5 to 7 seconds)

 

 

all of a sudden I have BIG spray, and it fills the throttle body. I'm

not cranking the engine, so I pull the plugs to let it air out.

 

 

a few minutes later I try again. The same. about 5 seconds of no spray, and suddenly I have a throttle body full of fuel.

 

 

I wonder if the injector was stuck before, finally became unstuck, and

now sticks open?

 

 

I pull the injector. Put 12v to it. it clicks happily.

 

no longer applying power....

 

I put air pressure to the intake hole of the injector. I get flow to

the output hole. I don't think this is supposed to happen (IE, fuel

should flow only if power to injector) Anyone know if injector should allow pressure to go from input to output with no connection to +,-12v

 

 

There is a screw on top. Not supposed to be able to remove it (tiny

tack welds)...

 

 

I remove the screw. there is a spring under it. there is an o-ring

on the screw.

 

 

the o-ring on the screw seems to be damaged/old.

 

 

any tips on cleaning or fixing the damaged injector (I'll go get a new-

used one at the yard tomorrow)

 

If I get a new o-ring for the spring holder screw inside the injector, should this fix the leak?

 

please email as well as posting (aiiadict at gmail dawt com)

 

 

Rich

[aiiadict]

EA81 (rebuilt), in a VW bus.

 

SPFI was cheap. So I stuck it on the engine.

 

SPFI was from EA82, 1991 loyale.

 

fuelpump, airflowmeter, distributor, ECU, harness.

 

(coil came from different SPFI sub)

 

 

Rich

[GeneralDisorder]

Check out my write up and make sure your wireing is setup properly:

 

http://home.comcast.net/~trilinear/EA81_SPFI.html

 

There's also a link to the 89 FSM with all the troubleshooting in it.

 

GD

[daeron]

yes, the partial FSM is the entire engine section, and in Part2 you will find the fuel injection chapter. This system is equipped with some wonderful stuff, for such a low-budget simple little computer. However, it sounds like a new injector assembly may be in order since you've already disassembled this one and discovered a bad o-ring.

 

if you can't find a link to the FSM, I will put one here, too.. http://www.ch601.org/engines.htm

 

Good luck, it shouldn't be that difficult... if you've gotten this far before running into any problems, you obviously know what end of a wrench to use.

[aiiadict]

 

I'm looking at the schematics, and I can't see how fuel gets into idle

circuit. with throttle valve closed, the only air entering the manifold

is through the bypass valve. the inlet to the manifold is below the

throttle valve. The only place I see that fuel enters is from ABOVE

the throttle valve...

 

how does gasoline enter manifold at idle?

 

I see... tiny hole that the idle adjustment screw regulates flow.. It

connects the throttle body above the throttle valve to the air bypass

chambers...

 

I have this wired correctly. I don't know about all of the smog stuff

(I cut ALL connectors off the harness, except for the ECU, and plugs

that go to sensors)

 

I'm going to need to add a speed sensor to the VW, so I can have

the ECU shut of gas while coasting (this is an awesome feature, especially

since the VW is automatic)

 

I did this project to (finally) acquaint myself with FI setups. This system

is really cool! Easy to understand. The wiring seems scary at first, but

if you spend some time on it, it becomes less confusing. All the wires do

is connect the ECU and the sensors/injector. Some go to the smog equip.

 

+to ECU, - to manifold (the - took me a while to trace :-), + IG switched to

red/lavender. Fuel pump grounded, fuel pump relay to +IG switched, coil to +ig switched. START on ECU to start signal (crank engine), and GROUND THE COIL.

 

In a few docs' Ive read, people claim that the coil gets supplied with - to make it fire. This is NOT true.

 

The coil is grounded to engine (connect a wire between bracket and engine case). There are two 2 conductor connectors on the Dist. 1 connector has Black to capacitor, cap to ground (radio spike absorber), the yel of this connector goes to tachometer, which isn't necessary for ECU operation as far as I can see.

 

The amplifier transistor does "amplify" the low current ECU output, but the transistor is actually used as a low to high current switch, IE all on or all off.

 

the other connector is to the "amplifier transistor". Must be similar to PNP transistor. One wire to + switched ignition. One wire to the ECU (ignition control). The wire to the ECU is coil control. the ECU sends out a low current signal on this line. The amplifier transistor "amplifies" this, and allows +12 (from +IG) to flow to through the coil, HIGH current.

 

So that would make the +IG connector the collector... Emitter to the ECU, and the base to the + power of the coil.

 

(been a while since I took electronics in college)

 

Rich

[daeron]

honestly, fuel injection systems are almost all as simple as this one... FI, to alot of us youngsters, is much easier to understand than carburetors.. less vagueness.

 

Its all a matter of numbers.. the computer takes numbers from varying sensors, and calculates the numbers it needs for spark and fuel supply. If your sensors are off, then the numbers will be off. The 80's Soob SPFI is just a wonderfully simplistic device that doesn't allow itself to get too complex, plus it has the on board diagnostics with ECU LED readout rather than needing an OBD scanner.

 

You picked a great project to familiarize yourself with FI systems. I am a Datsun guy, and I have a pdf on my computer that is the "fuel injection bible" for the Bosch L-jetronic fuel injection installed on my 75 280Z.. despite the fact that it was tailor written to the particular subsystems installed on the Z, its a FANTASTIC read on L-jetronic system operation, and the L-jetronic FI system is basically what is used on virtually all modern cars.. maybe the last 3-6 years have seen something of a departure from that, but as far as Fuel Injection Theory goes, its worth reading. PM me with your email address and I can email it to you, if at all interested.

[aiiadict]

any tips on cleaning or fixing the damaged injector (I'll go get a new-

used one at the yard tomorrow)

 

got it

 

New injector sprays good when power applied, stops spray when power removed.

 

previously, with the original sticking injector, the injector got stuck on, but I don't know if it was under power when it happened. the fuel pump could have pumped the line up to high enough pressure that the failed injector snapped open and let the pressure out...

 

with new injector:

 

try:

apply +-12v,

result:

spray

 

try:

apply +-supply to computer, ign on

apply +12 to START on ECU

result:

pump relay clicks on/off and pump turns on/off while START = 12v

injector = 0v (with analog meter.. the digital one takes too long to take a reading)

 

Ok, so now the injector isn't getting power..

 

injector resistance = 1.7ohms

injector + harness resistance (at ECU 43,48 RedBlack and RedWhite) = 1.7 ohms

 

so, good continuity, no short. harness isn't adding resistance to injector circuit.

 

No resistance between injector terminals and case (it's insulated properly)

 

on page 50, section 2-7 of the FSM, it says I should replace the ECU..

 

How did it go bad? did the sticking injector cause the ECU to fail?

 

there's probably an "amplifier" transistor leading to the +V side of Injector+ on the ECU (pin 43).. I wonder if that burnt out somehow. I'm going to test for +12v on the +injector supply inside the ECU and see what I get. maybe the connector on the end of the harness, or on the ECU has failed... I'll find out.

 

Rich

[aiiadict]

there's probably an "amplifier" transistor leading to the +V side of Injector+ on the ECU (pin 43).. I wonder if that burnt out somehow. I'm going to test for +12v on the +injector supply inside the ECU and see what I get. maybe the connector on the end of the harness, or on the ECU has failed... I'll find out.

 

Rich

 

I looked inside the ECU, there is a transistor running to the Injector power supply output pin +. Makes sense, I get - GND at the injector even with IGN off.

 

I try:

+12, gnd to end of wiring harness (to connector that goes to ECU), pins 43, 48. Injector sprays.

 

+12, gnd to inside ECU, so testing the connection between ECU and the smallest connector. spray from the injector.

 

ok analog meter on pins 43, 48 inside ECU. Ignition on. computer flashing code 5 (from what I read, this is specification code?). Apply +12 to START on ECU / starter solenoid.

 

0 to +2 volts coming from ECU, going to injector.

 

Computer is flashing code 14, "injector: abnormal output".

 

Can anyone tell me if this is normal injector voltage? It could be the slow response time of my meter giving me false readings.

 

note: injector will not get any power if crank position sensor doesn't sense movement... IE you have to either pull it out and spin by hand, or leave it in and crank engine!

 

Rich

[daeron]

have you checked the ecu output for the injector signal while ignition is in start? You did specifically mention checking the injector voltage then, but i couldn't find if you said you checked the ECU pin for voltage while cranking, too... To the best of my knowledge, there is no "dropping resistor" or anything else in line between the injector and the ECU; but the injector may receive a grounding signal from the ECU and get constant positive voltage elsewhere.. It may be that you've got your injector wired backwards.

 

I searched the FSM and GD's SPFI conversion writeup briefly, and couldn't discover an answer to the polarity of the signal from the ECU to the injector, but I suggest also checking for continuity between ground and the injector "firing" pin from the ECU.. it may be that the ECU is switching ground to the injector when you are looking for it to switch 12V+.

[aiiadict]

have you checked the ecu output for the injector signal while ignition is in start?

 

Yes. ECU output on injector power supply pins = 0 to 2 volts while cranking.

 

 

You did specifically mention checking the injector voltage then, but i couldn't find if you said you checked the ECU pin for voltage while cranking, too...

 

Tested at both injector, and at ECU. While cranking, 0 to 2 volts at both.

 

To the best of my knowledge, there is no "dropping resistor" or anything else in line between the injector and the ECU; but the injector may receive a grounding signal from the ECU and get constant positive voltage elsewhere.. It may be that you've got your injector wired backwards.

 

injector power supply pins on ECU:

 

48 = negative supply

43 = positive supply

 

48 is always connected. 43 (positive) is switched on/off by the ECU

 

I searched the FSM and GD's SPFI conversion writeup briefly, and couldn't discover an answer to the polarity of the signal from the ECU to the injector, but I suggest also checking for continuity between ground and the injector "firing" pin from the ECU.. it may be that the ECU is switching ground to the injector when you are looking for it to switch 12V+.

 

 

pins 48 and 43 show polarity of Injector power supply.

 

at the injector, RW = positive, RB = negative. The wiring harness is unmodified between the ECU and the injector (it is as it was in the loyale when I pulled the system)

 

Injector always has GND. the "firing pin" from the ECU (pin 43) is +12v.

 

I've tested everything I can think of so far. I'll test throttle idle switch tomorrow. perhaps it is sending "NOT IDLE" and "extremely low throttle" (IE, below "acceptable" range.. which may cause the ECU to tell the injector to open TINY ammount (2v worth), which is under-rating for injector to open.)..all because throttle sensor is sending too low signal for throttle position.... Maybe. I'll cross my fingers :-)

 

 

Rich

[GeneralDisorder]

Injectors don't run at 12v - that's why you have to use a noid light to test injector fireing on TBI and MPFI. I'm not sure what 12v would do to the injector - quite possibly damage them, but I AM sure you are making this much more difficult than it needs to be. Do the ECU power and ground checks in the 89 partial FSM pdf. You have wireing issues. The injectors do not fail and neither do the ECU's. In over 4 years on this board I've never heard of such a thing - the chances of you coincidentally running across such an extremely rare type of failure after butchering the harness is nill. Your problem likely stems from your complete reworking of the wireing - you should have done that *after* you got it started.

 

GD

[aiiadict]

>Injectors don't run at 12v - that's why you have to use a noid light to test

>injector fireing on TBI and MPFI.

 

Ok... that's good, because they definitely are not getting 12v

 

>I'm not sure what 12v would do to the injector - quite possibly damage

>them, but I AM sure you are making this much more difficult than it needs to

> be.

 

12v for very short periods of time doesn't appear to effect them.

 

 

>Do the ECU power and ground checks in the 89 partial FSM pdf.

 

I've gone over every pin in the system. The only work I did on the harness is pulling out headlights, etc. Between the ECU connector and the intermediate connectors/sensor connectors, the harness is %100 original.

 

 

>You have wireing issues. The injectors do not fail and neither do the ECU's.

 

That's why I've been testing the wires :-) The injectors do in fact fail. The first one I got from the junkyard was definitely bad. I hope the ECU is ok.

 

 

 

>Your problem likely stems from your complete reworking of the wireing - you >should have done that *after* you got it started.

 

 

I didn't completely rework the wireing though...

 

I've verified that the +,- pins to injector are properly wired, not shorted, injector is within resistance range, injector is insulated...

 

 

 

thanks for replies!

[aiiadict]

Injectors don't run at 12v - that's why you have to use a noid light to test injector fireing on TBI and MPFI. but I AM sure you are making this much more difficult than it needs to be.

GD

 

Hi,

 

What is a NOID LIGHT? Is it something I can make?

 

Do you know what voltage level I should get at the injector during START?

 

I hope I'm making it more difficult than it needs to be! Because I've tested this thing over and over again. But I am learning how FI works, which was the main goal of the project.

 

I'm going to go through the entire section 2-7 today, and do *ALL* of the tests to see if I can find what's wrong.

 

Rich

[daeron]

Hi,

 

What is a NOID LIGHT? Is it something I can make?

 

Do you know what voltage level I should get at the injector during START?

 

I hope I'm making it more difficult than it needs to be! Because I've tested this thing over and over again. But I am learning how FI works, which was the main goal of the project.

 

I'm going to go through the entire section 2-7 today, and do *ALL* of the tests to see if I can find what's wrong.

 

Rich

Rich,

 

What GD meant was that Injector Failure and ECU failure are both EXTREMELY uncommon failures.... 95% of the problems with this engine when fuel injected, are either: TPS, CTS connections, dirty MAF, blown intake manifold gasket, blown timing belt, blown headgasket, or missing a distributor rotor screw. OK, maybe only 85%, but still......

 

The word on the street is, pretty well the only way to fry an ECU is to connect battery cables backwards (this fries them quickly and easily) and POTENTIALLY disconnecting the positive cable while the engine is running.

 

That being the case, the fact that you've already run into one sticking injector GREATLY reduces the odds that you've got another one. Not only that, but the fact that you had one of the uncommon problems lowers your odds of having another uncommon failure. That's why he is using the language he is.

 

The FSM should tell you how to test the injector firing signal.

 

GD, this system drops the firing signal down to the appropriate voltage inside the ECU? I know my Datsun has a separate "dropping resistor" to do that, but the Datsun ECU switches ground to the injectors and feeds them constant positive voltage thru the resistor...

[aiiadict]

The FSM should tell you how to test the injector firing signal.

 

GD, this system drops the firing signal down to the appropriate voltage inside the ECU?

 

I cant find the injector firing signal test in the manual....

 

unless there is a resistor (very small) wrapped up in the harness, the ECU does all the work...

 

Rich

[aiiadict]

Success!

 

After playing with the 91 loyale ECU for a LONG time, I accidently dropped my hot-start wire into the open computer. No more 91 ECU :-P

 

I had tested both ECU's on the bench, with the 91 wiring harness and the 91 distributor. The 91 ECU would make the injector click when spinning the 91 dist. The 87 would NOT click with the 91 distributor.

 

So I plug in the 87, clip the 87 dist. to the harness. I get spray when I spin the distributor! The spray is VERY visible, so if you have any troubles starting, crank with the intake boot off and YOU WILL SEE SPRAY.

 

Ok, cut the connector off the 91 distributor. Same with the 87. Solder the 87 distributor to the 91 connector. put it all together... no run!

 

Check spray... none.

 

check spark... none.

 

ECU is reporting crank angle pos error.

 

So, it turns out the connector off the 91 distributor is bad.

 

I remove it from the 87 dist...

 

I use a 4 prong ?headlight? connector that was junk out of the harness.

 

solder it up. tape it up. put it in the engine.

 

crank..... pop pop. crank.... pop pop. Not backfiring, but firing and popping out the exhaust! It sounds like it's trying to start.

 

ok, connect the throttle cable, go to the drivers seat. push gas pedal half way and crank.... RUN!

 

It runs and sounds great!

 

it won't idle. This could be that the 91 AFM is not compatible with the 87 ECU (I'm going back to the yard and getting the 87 AFM) airflowmeter... both are the "hot wire" type, not flapper valves.

 

I also need to make sure the idle switch is operating correctly.

 

it took hours to track... I wonder if the 91 ECU would have worked if I had found the problem in the Crank pos sensor before I fried the ECU ?

 

thanks for everyones help!

 

Rich

[GeneralDisorder]

MAF's are interchangable, that's not your problem. I have used many MAF's and many ECU's in my swaps and all work together. Make sure hot-wire bypass is clear of oil and dirt, and check operation of the TPS idle switch.

 

Most likely it was just the wireing (as I said it would be) with the distributor. Either ECU should have run fine.

 

GD

[daeron]

you hit the nail on the head: the idle switch. Its part of the Throttle position sensor, which has a VERY NARROW window of proper calibration.. you need a set of feeler gauges and an accurate ohm meter to calibrate it, but its a fairly easy process, outlined in the FSM. The FSM refers to it as a "test procedure" but it then goes on to say that should a new unit be needed, it must be adjusted to the specifications used in your "test procedure".

 

If the TPS is off, it will have difficulty idling. If the IAC is in poor shape, it will have trouble idling. If you have any vacuum leaks, it will have trouble idling.

 

(again, trying to brainstorm here, not condescend :-p)

 

anyhow, congrats and good luck.

[aiiadict]

If the TPS is off, it will have difficulty idling. If the IAC is in poor shape, it will have trouble idling. If you have any vacuum leaks, it will have trouble idling.

 

anyhow, congrats and good luck.

 

I checked the idle switch. it was definitely not operating properly. I hand adjusted it so that it CLOSEs when throttle is released fully.

 

I'm cleaning up the wiring a bit before I try to start again.

 

Vacuum leaks: there should be none. I got a full set of good vacuum lines and they're all connected as they should be.

 

IAC was removed, cleaned, and reinstalled facing the drivers side (so it isn't in way of dist.).

 

Thanks again.

 

Rich

[Snowman]

Wow, I'm sorry I missed this thread being off the board for a while. Sounds like a lot going on there!

 

Good to hear that it runs though. I'd be interested to see what happens once the TPS idle switch is set up properly.

[aiiadict]

Wow, I'm sorry I missed this thread being off the board for a while. Sounds like a lot going on there!

 

Good to hear that it runs though. I'd be interested to see what happens once the TPS idle switch is set up properly.

 

 

I'll post pics of everything later this evening. I haven't tested the idle yet. I adjusted the idle switch in the TPS with a meter. It's ON with no throttle, turns off with throttle. It definitely was NOT set correctly before (it was never turning ON)

 

Rich

[daeron]

the TPS switch should me calibrated to a slightly finer degree than simply ensuring that the idle switch functions properly.. The procedure involves two different feeler gauges, set in place between the throttle drum and the throttle stop screw. One gauge simulates a 1* throttle opening, and the other a 2.5* throttle opening.. and the TPS should set specific resistances (withing a range) at those two points, as well as a certain spec resistance at WOT.

 

If you've already gone through all that, then cool.. I just wanted to make it abundantly clear in case you still have any running issues... Sounds like you are on the right track, though.

[aiiadict]

the TPS switch should me calibrated to a slightly finer degree than simply ensuring that the idle switch functions properly.. .

 

Ok... It idles now (after making sure TPS idle turns ON when no throttle...sounds good) but sits at about 800rpm for about 30 seconds, then runs up to about 900rpm for about 30 seconds, then goes back down to about 600 and runs a bit rough, then back up to 800...

 

I'll be adjusting the TPS switch today. I still need to figure out which test connector goes to which ECU pin so I can time it correctly.

 

Rich