Fuel pump & main fan run continuously in key off, but not in start or on

[propane]

 

 

 

 This is the same 1996 legacy AWD wagon 2.2L w
5spd that’s in the diff howl thread/ topic. This newest problem is entirely
different (electrical)  & felt
justified in started new thread. Prob is that fuel pump runs in key off, &
not in on. Same for main fan. ( & I assume in start as well cause it wont
start now. Previously, the fan would cycle on, then start ran fine.  So far its killed the battery( have a good
replacement battery in now, which I leave disconnected till I can find out more)
probably from running continuously & I didn’t hear it The only electrical “work”
I’ve done was to R &R the front ABS sensors; & unplugged  the O2 sensor (when I dropped the exhaust to
change steering rack).  UPDATE: temp guage  swings to max, & Tach jumps.with key off , goes nil with key on. Wonder if I fried a relay by using the battery charger to jump start? also  changed out the O2 sensor.
. What
is supplying power to fuel pump? I first went under the dash and sure enough
the green connectors were plugged together. So I unplugged those. Still the
same. Fan runs unless I pull the 20Amp fuse in the eng Main fuse block.  no clue as why.. thx

forgot to add- I'm sure I had a CEL light previously, now there's nothing, as in the initial key on postion looks like every other indicator light is there.


 

Edited June 23, 2014 by propane

[Fairtax4me]

Make sure the battery connections are clean and tight. Also check engine and chassis grounds, and the grounds for the ECU on the intake manifold. That ground is down low on the back corner of the manifold on the drivers side below where the fuel rail loops around the back of the manifold.

 

Other than that. I would have to think the ignition switch is bad or the ECU fried. Might want to pull back the carpet on the passenger floor and check for water. The ECU is under the metal plate on the floor.

[propane]

Battery terminals were cleaned when new battery subbed in, this car used to start instantly, & w replacement battery cranks fine. I pretty sure its not getting fuel since the pump only  runs when off. Doubt its getting spark, either. grounds are good, this is dry climate & was garaged for entire time. (have carpet back & will visually check ECU) I subbed in a known good used ignition switch. Same symptom. Further, since I disconnect the battery each interval when trying something new, I noticed that it doesnt act up  till the key on triggers something. . THis flips something on to power both pump & fan, when  I then roll it back to off, but not in on. I wonder If I fried the ECU when using the battery charger 60 AMP start mode. thx

Edited June 23, 2014 by propane

[Fairtax4me]

ECU grounds the fuel pump relay and fan relay to turn them on. Possible the ECU got cooked. The charger could have caused a high voltage spike.

 

There was an issue on the 1st gen legacy where the voltage regulator in the alternator would go bad and send voltage back to the fuse panel that would cause weird problems. Could try unhooking the alternator and see if you have the same symptoms.

[1 Lucky Texan]

in this car, are there also some white or other connectors that should stay unplugged ?

[propane]

No joy by disconnecting Alt, I notice the rev counter jumps to 500RPM  , opposite of the temp gauge pegging max high to max low outside of is range. fuel pump still runs in off, nil in on. no white connectors I see , I rem a thread where a pair of black connecters in addition to the green are mentioned  Still trying to narrow it down.

[Fairtax4me]

Did you disconnect the plug and the charge cable?

 

Do all the guages go crazy like that with the key in Off position? Crazy guages could mean a back feed from somewhere.

Anything else like lights or radio affected by this?

[propane]

 

 Yeah, I disconnected Alt. plug and main
charging wire. Only other gauge is fuel and it is not involved , no radio
noise, EDIT: yes there is a brief static sound when beginning the test.

    not sure what symptoms there would for radio or lights – cant see any
diff.  The thing that is unusual is the
instant jump of the rev to 500RPM (eng. Not running) and Max ohm swing by the
temp gauge,(as well as fuel pump) ONCE 
the battery has been re connected at the neg.edit: & HAVE cycled the key once, doesnt quit until the bat is discoed again.  

I have a 1991 FSM to see
where the fuel pump relay is unless somebody knows off hand, It seems like
something gets tripped & then stays energized feeding power to its self,
plus the fuel pump & temp gauge, rev counter  Edit:  previously, when doing a compression test, I unplugged the coil - not sure if this a good or bad practice? also, the drivers low beam plug for headlight was melted, I seperated out the wire, to fix later, noticed the whole plug had been chopped in, using wire nuts.  EDIT : found the relay for fuel pump, will take it out & test it per haynes manual instructions

 

 

Did you disconnect the plug and the charge cable?

Do all the guages go crazy like that with the key in Off position? Crazy guages could mean a back feed from somewhere.
Anything else like lights or radio affected by this?

Edited June 24, 2014 by propane

[Gloyale]

WAAAY up in the upper left hand corner of the dash is the Main relay (brown 6 wires) and the Fuel relay(green connector 4 wire) possible that the problem is there.

 

Either stuck relay or crimped ground.  My guess is a pinched ground keeping relay closed and pegging instruments.

[propane]

 Summary of Problem:  started after I tried to start car (was in a
hurry to test drive to see if suspension fixes solved diff howling) broke my
rule and used the charger to boost waay dead battery & attempt to crank over car, subbed good battery since then.( something drained this don’t
know.) green wires had been connected along, was “going to get to those” after
driving it.  That’s when the fuel pump
& main fan started running in off,along with temp gauge pegging hi, rev jumping to 500RPM, .. but go dead in "on " position, (subbed in diff ignition switch, unhooked green connecters) Car didn’t start BTW.

same pattern repeats once I disconnect the neg battery and start over

While I’m chasing
down fuel pump relay, I wonder what its saying with the main fan (& gauge
peg) involvement? Did I fry the main relay? ECM?

Edited June 24, 2014 by propane

[propane]

WAAAY up in the upper left hand corner of the dash is the Main relay (brown 6 wires) and the Fuel relay(green connector 4 wire) possible that the problem is there.

 

Either stuck relay or crimped ground.  My guess is a pinched ground keeping relay closed and pegging instruments.

 took the whole bracket out holding both relays , after unplugging. I wonder if Subaru's engineer;s thinking was that relays buried up high were more on less fail safe? (I only tested for continuity + for power circuit, no ohms for switched circuits (2 and 1)  By unplugging both, the temp gauge was deprived of power. Tachometer continued to jump to 500RPM w/o relays plugged in What I forgot in the above summary was the absence of the check engine light which comes out of ECM as does tach... my 1991 FSM shows them connected I think.  I'm about 90% sure the brain box got spiked by battery charger jump start. A local pull n save  has a 1/2 off sale this Friday. Will post results .

[Fairtax4me]

Go here: http://www.main.experiencetherave.com/subaru_manual_scans/

and get the FSM for 96. There are quite a few wiring differences in that 5 year span.

[propane]

swapped ECU wasn't the problem same symptom  the key has to be that fuel pump runs w/ key off ---   Does it have its own controller/ timer? thx 

  Down loaded Factory service manual, nice but I couldn't find a wiring diagram for engine control,

 Update: didn't scroll down in FSM PDF, its all there !! thx

like in my 1991 FSM I realize the diff as OBDII was introduced  in 1996 . Someone suggested to check for fused wires underneath of main wiring  block in eng compartment. Was thinking of recheck main & fuel relays under power to see if their is cross over from switched voltage. Dont know where to start looking for "pinched grounds" as Gloyale suggested. I suppose the next trouble shooting step would be to trace power supply circuits to afflicted components, looking for a common denominator

Update: pump doesn't run if I have had the battery discoed- only after switching on - off ... that sounds like a relay flipping on & being burnt internally  I need to check voltage (back probing the connectors for main Fuel pump) relays are cheap to sub in - Can't hurt at this point.

 key off, fuel pump runs, temp swings max up;

 key on Rev pegs to 500 RPM;

 no longer starts

 found battery discharged probably from fuel pump running all night;

gave it a battery charger boost  no start  gauges go wild

 changed brain boxes , no difference

 Update: No check eng light, bulb good, wondering if I botched an O2 (front) connector, & that is tied to the light.

Edited July 10, 2014 by propane

[Cougar]

I suspect you may have damaged some wire insulation when you did the jump to the battery and some wires may be shorted together now. The wires in the ignition switch harness may be where the damage is at. Disconnect the connector for the switch and see if that stops the trouble. There may be other damaged wires also.

[propane]

Can close out this topic- It was the main engine harness unplugged, left barely hanging by last person who "worked" on the car. After finding the rear diff drain plug, & oil drain plug finger loose, dripping I shoulda caught the style of maintenance. I must have nudged it when securing the O2 sensor cable OEM style so it doesn't tangle with the CV axle. 

 Not having the temp sensor connected must have been the trigger for the gauge to go max on. I had to use another speedo cluster, original was in bad shape & temp needle was half twisted off.

 fan ran because of green connecters at first, then ?? because of temp no reading???

fuel pump because of main relay grounding it ???

 

 Only thing I can think of to be gained from this "wild goose chase" was that no fuses ever blew, no melted harnesses, needle on temp kept trying to bury itself past max. Follow me over to the "differential howl thread"  tomorrow it was front bearing- wheel bearings on  Now that was something learned.

 THx for the FSM link, & all the comments,

Edited July 15, 2014 by propane

[Fairtax4me]

Ground problems cause all kinda problems. The ECUs main grounds run through the harness connector to the intake manifold.

 

Glad to hear you found the problem!

[Gloyale]

Yeah......I should have known......

 

Now that you say it was the main engine connector.......I remember seeing that before.

 

Main plug disconnected = fans run and ECU/Dash does weird things.

[propane]

I took it to Auto Zone to have codes read; sorta expected something like the 0325 knock sensor, other one baffles me, 0180 I think, which was fuel temp range exceeded or something. all of the reading of the factory service  manual, I do not recall any temp reading hardware??

 with the fuel pump running so much, did this trigger the code??

Am thinking of clearing the codes w/ battery disconnect. It has a "low oil rattle on start up, maybe setting the knock sensor code, or I buy a new one from Amazon for a $15, looks like I'm almost done with this part of the reconditioning, to put this back on the road. I'm still getting a dragging front brake, it releases some of time

[Fairtax4me]

Rattle at startup? What kind of oil filter is on it? If its Fram you need to change it.

 

Regardless of the oil filter, a rattle will not set a knock sensor code. The sensor detects only a specific frequency range and does not set codes when it picks up knocks in that range. You will only get a knock sensor code if the resistance of the sensor element is out of range (internal damage or corrosion), or if the sensor is unplugged.

[hippyloc]

I'm having same fans/fuel pump running with key off issue. Freaking been going nuts for a couple days trying to figure it out after a 5sp swap. I see it has turned out to be your main engine harness. I wish that I would have looked here to start. Would have saved a lot of frustration and disappointment. Almost went postal on this poor beast. Was beginning to think it was some sort of cosmic joke, the way things have went lately. Thank goodness for the USMB community and the willingness of it's members to continue being great human beings.

[el_freddo]

@hippyloc- you got your issue sorted and it was the same loose main plug? 

Cheers 

Bennie

[hippyloc]

It was actually a combo of problems. 

First: connector B12 was not even plugged in, like I convinced my self many times "Of course that's in, how could I forget that?" Well I did forget, even had tucked one end up behind the rest at the firewall.

Second: the brain, ECU, was coming out of a 2yr. coma and spitting out all kinds of CELs.

Third: Wore down batt. narrowing causes for no start.

The odd fans and fuel pump timing was result of batt. Charge being lower than multimeter was saying and poor grounding. 

My mm was getting like 30-40 ohms when checking grounds. Plus parking lights no work either switch. All other lights wrk, turn, brake, reverse and headlights. All of the chassis grounds were garbage. Welded on some stainless bolts, in appropriate locations. Added extra ground cable from battery to strut tower. All new gnd locals tested <5ohm. Still hunting down park lights no wrk issue, but cars starts easily now.

 

[hippyloc]

Always double check. I keep forgetting to do that.