Need help with fuel pump or ECU problem

[dancoco808]

hi guys,

have a eg civic with b18c and hondata s300. car is unable to start. previously no problems. today when doing some interior work on the car and removing loosely run and unused audio cables;

i noticed the positive wire from the ecu relay to fuel pump had been poorly installed (under the seats and along side raid) and the insulation had been worn right through. the wire had made contact with the body/ground. im suspecting this caused a short and possibly caused damage to the fuel pump or ecu or relay?

i have checked all ECU fuses in the engine bay and kick panel (although box says 7.5A, but has 15A fuses in it), and none of the fuses were busted.

all accessory power works, engine cranks fine, but does not start/fire. im unsure if i am meant to hear the fuel pump prime when turning the key to acc/on, but i cannot hear it.

anyone have any experience on the matter?

cheers,
- dan.

[chowdaa]

Dont want to jump to conclusions, but have seen few eg ecu capacitors leak, short the board and blow the fuse, i found a 10a where my 7.5 was meant to be after smoke had poured out of my ecu. From what you are telling us, sounds like previous owner had issues with old motor or current motor blowing fuses, whether it be from short or faulty ecu and just kept upgrading the fuse. Possible ECU damage.

How is your main relay, just trying plugging in and out with ignition on and see if you can hear the pump.

Does is smell burnt? can the top be opened for visual inspection?

[dancoco808]

hey mate, thanks for the reply. tried taking out relays in and out, no difference. i have a slight burnt smell thats very very faint, vague and non-specfic from the engine bay fuse box.

here's a shot of the ecu

http://img.photobucket.com/albums/v334/coco69/civic%20blog/IMG_66190800.jpg

the big red wire from the relay goes to the fuel pump under the back seat.

http://img.photobucket.com/albums/v334/coco69/civic%20blog/IMG_66170800.jpg

the red wire from the relay to fuel pump that was ran dodgy, worn through and made contact with ground.

http://img.photobucket.com/albums/v334/coco69/civic%20blog/IMG_66200800.jpg

[GenesisEG]

If that's the wire the powers the fuel pump, then it won't blow the ECU (or at least it shouldn't). The ECU should only be controlling the coil, and is electrically separated from the contacts themselves. It's more likely that if the power cable shorted out, you'd have a problem with either the fuel pump fuse, the relay, or the fuel pump itself. The ECU should only have a problem if the coil or the ECU cables short out, which it doesn't appear to have done.

I assume you understand how a relay works, you seem to know what you're doing.

First, try measuring between the coil contacts and ground, you should get some voltage there if the ECU is firing up the relay.

Then, if you are getting voltage, check the continuity between the contacts on the relay when it fires up. When the coil is energised, the contacts should close and you'll get power through to the pump. If the power is hitting the coil but the contacts are not closing, the relay is kaput.

If you're still not getting anything, but the relay is working fine, check to see if the line side of the relay is getting power, it should be getting power all the time.

If you are getting power, and the relay is working, then the pump is probably gone.

If you're not getting power, the fuse may be gone.

Please keep in mind, I'm saying all this from a purely theoretical standpoint, as I don't really know for sure how cars are wired up (I'm an electrician and security technician). It just makes sense to me that when the ECU wants the pump to work, it fires the coil, which closes the contact, which turns on the pump. If it's not working, then check it in that order. Make sure it's firing the relay, then check the relay contacts, then the power source, then the pump.

Get rid of that fuse too, put the right capacity fuse in there.

Oh, I just realised...

It looks like that bracket on the bottom might be the ground for something, probably the coil. I only see 3 wires on that photo going into the relay, and that doesn't make sense. There should be 4 wires, otherwise you don't have a circuit. If it is the coil ground, then it won't work unless it's grounded. I might be mistaken, but just something else to think about.

[dancoco808]

thanks for the reply mate, ive already tried replacing the relay with a new one, and no change.

will test each of the coils/contacts for a more detailed idea of whats going on as you suggested.

turns out the bracket on the bottom of the relay isnt used for ground as it doesn't reach into the relay itself... i thought this also, but was able to unclip the bracket quite easily and noticed it didnt have a connection to the relay. there are actually 4 wires connecting to the relay, the 4th is behind in the picture. the central pin 5 is unused. but the thing is, prior to this event, the car has started fine.

is there a way to test the fuel pump? i.e. supply it with another 12v source and see if it primes? is this safe to do?

[GenesisEG]

If it's a 12VDC pump, which I guess it would be in a car, then yeah, should be fine. Just take a lead straight from the battery to the pump.

Before you try that though, have you tested the relay to see if it's being fired up?

Measure your DC voltage on the line side of the relay to make sure that your fuel pump positive cable is powered, and then measure from the load side of the relay. The load side SHOULD get 12V when you turn on the car.

If you're getting 12V constant on the line side, and 12V on the load side only when you turn the car on, then you know the ECU and relay are fine, and you can put it down to the pump.

At least that way, you can eliminate an ECU problem before you risk frying your pump! =P

[dancoco808]

went down and poked the relay with a multimeter... still nothing :( but heres some results that may help diagnose.

(btw following is all with car on 2 clicks 'on', not ignition)

old relay has power
http://img.photobucket.com/albums/v334/coco69/IMG_66310800.jpg

old relay, but no power to fuel pump line
http://img.photobucket.com/albums/v334/coco69/IMG_66060800.jpg

new relay has power
http://img.photobucket.com/albums/v334/coco69/IMG_66320800.jpg

new relay but also has no power to fuel pump line
http://img.photobucket.com/albums/v334/coco69/IMG_66330800.jpg

so regardless of relay, there appears to be no power to the fuel pump line

other pics:

kick panel fuse box, noticed alot of wrong fuses, slowly replacing them
http://img.photobucket.com/albums/v334/coco69/IMG_66360800.jpg

engine bay fuse box, replaced ECU and backup fuses to correct, does it matter that the battery fuse is 100 instead of 80 as it says on the sticker?
http://img.photobucket.com/albums/v334/coco69/IMG_66230800.jpg.
http://img.photobucket.com/albums/v334/coco69/IMG_66240800.jpg

[90LAN]

looks like you dont have a main fuel relay

look for a silver/grey box with brown end on it

see if you have one

[GenesisEG]

The fuses should be rated no higher than the specified rating.

If it's too high, then it's not protecting the cables and equipment as it should. It'll allow too much current to pass through in the event of a fault. The point of the fuses is to provide a controlled weak-link in the circuit so that in the event of a fault, the fuse will melt first, protecting the components on that circuit. If it says 80A, put an 80A fuse in. It's not worth risking it for the sake of something so cheap and simple.

If the new relay is not allowing the 12V through to the load side of the contacts, then the coil is not firing, which is probably controlled by the ECU. Unless you're really unlucky and the new relay is a dud too.

Can you identify which wire provides power from the ECU to the relay? If you can, try measuring the voltage on that wire. The ECU should fire a small amount of current down the wire to energise the coil in the relay and close the contacts. It looks like the ECU is not firing up the relay, so double check the ECU positive wire just to make sure. If you get no voltage on the coil circuit when you turn on the ignition, then the ECU might have been damaged.

Sucks, but at least you've isolated it down that much.

[dancoco808]

looks like you dont have a main fuel relay
look for a silver/grey box with brown end on it
see if you have one

yeh i cant seem to locate one either, is that because of being an non-efi stock eg4 breeze, that now has an engine swap by any chance? will look again. are you referring to this lan? (pic taken from http://www.tegger.com/hondafaq/mainrelay.html)

http://www.tegger.com/hondafaq/mainrelaylocation/hondacivic94ex_mainrelay.jpg

you reckon at this point in time its safe/a good idea to run a 12v to the fuel pump to see if it works? or given the fact that it receives no power is pointless?



Can you identify which wire provides power from the ECU to the relay? If you can, try measuring the voltage on that wire. The ECU should fire a small amount of current down the wire to energise the coil in the relay and close the contacts. It looks like the ECU is not firing up the relay, so double check the ECU positive wire just to make sure. If you get no voltage on the coil circuit when you turn on the ignition, then the ECU might have been damaged.

Im not too sure of the wiring colours etc, as im a toyota guy and ive just owned this honda for about 2.5 weeks, also its got an aftermarket computer to confuse me even further. i will try tho...

[GenesisEG]

If it's not receiving power, then it really doesn't matter if it works or not.

Try it, I mean the chances of breaking it are slim. Just run a lead from the battery positive to the lead where it plugs into the relay. If it fires up, then it's probably an ECU problem.

[dancoco808]

If it's not receiving power, then it really doesn't matter if it works or not.

Try it, I mean the chances of breaking it are slim. Just run a lead from the battery positive to the lead where it plugs into the relay. If it fires up, then it's probably an ECU problem.

so by that theory am i able to run a 12v positive lead from ignition switch wire to the fuel pump as a band-aid fix to move the car around?

its currently stuck underground 2 levels in my apartment carpark.

are there any other relays/fuses that i've not located?

[90LAN]

yeah if you do that it will be constantly on
but good enough to move it for now

[GenesisEG]

About the fuses or relays, I know not. I'm really not familiar with the wiring of cars. Everything I've said so far is from my electrical knowledge in general.

If the fuel pump is powered by 12V through a relay, then something must control the relay. That would probably be the ECU.

The thing that worries me about running the 12V from the ignition is that I don't know if the pump is supposed to run on a continual 12V supply while the car is on. I'd suppose so, but not knowing how the fuel system works, I can't say for sure.

Electrically, I see no reason why it wouldn't work. If you run the ignition power from the cable that powers your cigarette lighter or something and take it down to the relay, join it on to the load cable (the one that was shorted out)... It should work. Just make sure you turn your ignition to accessory before turning the car on to pre-pressurise the fuel pump before starting the engine.

It might get you out of the poop, but I can't guarantee that it won't cause you any more grief. Sorry I can't be of more help.

If you get it started, drive it straight to an auto electrician that can do a diagnostic on the ECU.

[dancoco808]

i cant seem to find any sign of a main relay / main fuel relay at all...

with my 2jz, our fuel pumps have 2 speed operations and its a common mod for bpu supras to cause the fuel pump to constantly run at full duty. im hoping that constant 12v to the fuel pump on the civic would be okay too... (http://mkiv.com/techarticles/12v_mod/12v_mod.htm)

but anyway, i went down to try and hook the fuel pump up to the ignition 12v, and its now starting to struggle to crank and dying. i can hear the fuel pump prime and engage when the key is set to the 'on' position, which i guess is a good sign the fuel pump may not be dead.

is it possible that the battery is flattening due to all this testing? i assume that the battery is not likely to be the sole problem and that a change would not be the solution though.

1. is it okay to jump the car with the aftermarket ecu?

2. if i jump/change battery, how far would it be safe to drive with the fuel pump running off the ignition switch? (enough to say take it from southport to slacks creek/HPR?)

btw, thanks guys for all the help and suggestions, much appreciated

[90LAN]

test the battery to see if its flat with a multi meter

[grifty]

went down and poked the relay with a multimeter... still nothing :( but heres some results that may help diagnose.

(btw following is all with car on 2 clicks 'on', not ignition)

old relay has power
http://img.photobucket.com/albums/v334/coco69/IMG_66310800.jpg

old relay, but no power to fuel pump line
http://img.photobucket.com/albums/v334/coco69/IMG_66060800.jpg

new relay has power
http://img.photobucket.com/albums/v334/coco69/IMG_66320800.jpg

new relay but also has no power to fuel pump line
http://img.photobucket.com/albums/v334/coco69/IMG_66330800.jpg

so regardless of relay, there appears to be no power to the fuel pump line



wait if ur getting power to the relay but the relay is not closing the circuit then something is wrong with the relay or the wiring

also 11.8v seems abit low, maybe its too low for the relay to activate?

also which relay pin number are u getting power to?

[dancoco808]

test the battery to see if its flat with a multi meter

12.6v across the terminals

does the pivot volt stabilizer have anything to do with all this?


wait if ur getting power to the relay but the relay is not closing the circuit then something is wrong with the relay or the wiring/
also 11.8v seems abit low, maybe its too low for the relay to activate?
also which relay pin number are u getting power to?

thing is ive changed the relay over to a brand new one. same problem. from memory, 30 seems to be ground, 85 and 86 are other powered sources, and 87 is to fuel pump with 87a in the centre left unattached.

[grifty]

when the key is in no 2 ignition position you should be getting power at 86, if not then ur ecu is not sending the sind signal to turn the fuel pump on

[dancoco808]

when the key is in no 2 ignition position you should be getting power at 86, if not then ur ecu is not sending the sind signal to turn the fuel pump on

I'm pretty sure im getting power at 86, just nothing sent through to fuel pump (87). I'll go check the exact contacts in a sec.

Try another relay? Try another battery?

[dancoco808]

i think there may be 2 problems:

1. wiring is going to the wrong pins of the relay
2. battery is starting to flatten/weaken

on the car at the moment:
30 - solid blue w/ white stripe: 0.0 volts
85 - thin blue w/ red stripe: 9.4 volts
86 - thin green w/ blue stripe: 8.2 volts
87 - to fuel pump: 0.0 volts

according to above results, on the car atm:
30 - ground
85/86 - power/switch
87 - fuel pump.

according to the markings on the relay itself it should be:
30 - power
85 - ground
86 - switch
87 - fuel pump

so... are the pins going to the relay wrong? anyone know what the colours of constant power and switch are?

clearly my voltages are dropping also. which also might be the reason why it hasn't started even when i ran ignition power directly to the fuel pump.

i will go get a new battery and place the wires in the correct position on the relay, hopefully this works ><

[dancoco808]

okay, have new battery in and wired the fuel pump directly to ignition wire. car starts and works fine, maybe smelling a bit extra fuelly (although it doesn't have a cat). ultimately i'd like to get the relay harness up and running though. but its good to hear the car running again, and great that i can move it out of the basement carpark.

so here's what i've worked with regards to the relay:

when i had the following on the relay:
85 - ground
30 & 86 - power/switch (unsure which is which)
87 - fuel pump

car would start, idle for about 5 seconds, then turn off. after the first start, the car would no longer start again, unless i unplugged the battery then it would start once again for a few seconds before being unable to start again. i was unsure about which wire was power and switch, so reversed them with the same result.

i then tried running brand new wires (not from the ecu), directly from battery to 30, ground to 85, ignition switch directly to 86 and from 87 to the fuel pump. the same result occurred as above.

it's got me stumped, cannot figure out why the relay circuit will power the fuel pump or keep it going.

im considering making the 60km drive with constant 12v to fuel pump to take the car to the guys who did the engine swap (HPR). anyone reckon it will be okay to do this?

[GenesisEG]

So many things to reply to, so I won't reply to them all.

First, to attend to your immediate problem.

Make sure it is a 12VDC coil. If it's not, don't even think about trying to run the relay with 12V or you'll fry it. There's a chance it might be lower, if it's being controlled by the ECU. It should say the coil rating on it.

If it is 12VDC, take it to the battery, lets test it.

Connect your ground pin to ground using any wire.

Since I can't tell for sure which pin is for the coil and which is for the pump, I'll need you to do a quick test for me.

Put one probe of your multimeter on the ground, and set the multimeter to resistance.

Then put the other probe on pin 30 and then 86.

One should have a completely open circuit, the other should show the resistance of the coil. I'll assume that 30 will be open circuit and 86 will be the positive of the coil, but I need you to test to make sure before I tell you to blow up your relay...

Let me know how that goes.

I just want to eliminate one problem at a time. Once we can confirm that the relay is working fine, we can worry about everything else.

[grifty]

ok here is how u should have it wired up

30 is power going to the replay

86 is the wire from the ecu

85 is negative

87 is the wire going to the fuel pump


if you were getting power at 86 and the relay wasnt turning on that means you werent getting power to 30, id either run a new wire or try to find the fault in the existing wire.

edit: i just read the above (stupid me)

can you just check if contacts 30, 86 and 87 are getting power when you are turning the ignition on?

[dancoco808]

okay some decent progress:

genesis, i checked the relay, it runs a 12VDC coil. as you mentioned and in keeping with the way the relay works, the open circuit showed complete continuity, while the other showed resistance. thx for the diagnostic pointers.

grifty, i ran thru all the pins again to check against the layout. i disconnected the wires from the relay and tested them off and on ignition. checked em all again just like u mentioned, i must have missed the fact that some of my wires weren't powering up.

-------------------------------------------

just got up from downstairs and this is what i did:

86: i noticed that blue/red wire from the ecu was reading 12v when the key was turned. i used this as ignition and plugged it to 86.

85: i ran a brand new ground instead of using the ground wire in the harness, this would eliminate any problems with the wire.

30: i traced the thick blue/white wire that i was using as constant power, noticed it to go thru the firewall and into the engine bay battery side. i assumed this was the wire providing constant power. however, it was reading no voltage either on or off ignition. this may be where the problem lied. for the time being (and seeing as i dont have a head unit) i tapped into the constant 12v line off the headunit harness (with fuse) and into the 30 pin on the harness as main power.

87: i connected this line to the fuel pump positive.

result: car fired up and maintained idle for the first time off the harness!

im gonna go back down and neaten up my wiring. there are now 2 wires that are being unused from the harness (i assume these to be the dud power, and possibly dud ground).

with regards to the pin 30 12v power wire, can i run a new wire from the battery +ve terminal thru the firewall and directly to the relay pin 30? or does it need to be a wire with fuse for safety, or come off the ecu? otherwise im pretty happy with my 87/86 and 85.

[grifty]

i would use a fuse from the battery+ to terminal 30 to give me peace of mind that a short circuit will not burn up all the wiring, there cheap aswell anyways.

[dancoco808]

cheers, will do that tmrw. currently running from a 10A, reckon that would do the job?

is it normal for the relay to get very hot also?

also, on testing, ignition, idle and acceleration, signal to the fuel pump was always above 12v. i guess its always running full duty, this normal?

[GenesisEG]

Regarding the wiring, I'd re-run the existing wiring to bring it back to stock as much as possible. This problem began because of mish-mash wiring, and whether you keep the car or sell it, the world will be a better place if the wiring is tidy and predictable. If you can re-run the wire in the loom back to the fuse panel, all the better.

Regarding the relay getting hot, yeah, the coil is a giant resistor and generates heat as current passes through it. Since the car needs fuel all the time, it makes sense that it'd be working full time to supply that fuel to the car. Since the coil is only a switch and not dependent on how much fuel you're using, it's running the coil at full capacity even at idle. That's why it's hot.

Congrats on getting the car up and running, good job. Once it's all tidied up, you'll be able to enjoy it with peace of mind.

[jdm_b16a]

I'm going to join the discussion here because basically I have exactly the same problem (or at least the consequences of a similar problem). Even though the car has had an engine transplant and an aftermarket ECU I don't understand why you need another relay when you should have an EFI Main Relay (like in the photo). The only answer seems to be that the conversion shop went with the standard non-EFI relay and then just added a bunch of extra wires to complete the conversion to EFI. Always a recipe for disaster, especially as the wiring looks like its been done with crimp connections, notorious for working their way loose over time.

The first thing is always to test the fuel pump is actually working. As you have found out, and others have suggested, a simple 12V feed to the correctly grounded pump will tell you if it is working. So far so good.

Your main issue is with power and you seem to have solved that, at least temporarily, with some new wiring. But I wonder if you are getting VTEC engagement as that is a function, in some ways, of revs. My understanding is that the fuel pump needs to be switched via the EFI Main Relay, and communicating that to the ECU. I'd be interested to know your final solution.

Peter

[dancoco808]

Regarding the wiring, I'd re-run the existing wiring to bring it back to stock as much as possible. This problem began because of mish-mash wiring, and whether you keep the car or sell it, the world will be a better place if the wiring is tidy and predictable. If you can re-run the wire in the loom back to the fuse panel, all the better.

Regarding the relay getting hot, yeah, ..... That's why it's hot.

totally agree. yeah, i'd love to bring it back to stock as much as possible, the only problem is that i'm having to work with a car thats was in a state very far from stock, with poor quality and incorrect wiring.

was getting a little worried about the heat, thanks for clearing that up :)


...I don't understand why you need another relay when you should have an EFI Main Relay (like in the photo). The only answer seems to be that the conversion shop went with the standard non-EFI relay and then just added a bunch of extra wires to complete the conversion to EFI. Always a recipe for disaster, especially as the wiring looks like its been done with crimp connections, notorious for working their way loose over time.

But I wonder if you are getting VTEC engagement as that is a function, in some ways, of revs. My understanding is that the fuel pump needs to be switched via the EFI Main Relay...

thanks for joining in mate. the more the merrier, as every single person joining in so far has been incredibly helpful. :)

agreed. i seem to not have a real EFI main relay at all. they may have kept the standard non-EFI relay (eg4 breeze w/ carby originally) and just added this separate floating 5-pin relay to act in its place. the wires have also been done in a very average manner. the thing is im not sure what to do about it but run the relay i've been given, except with secure and quality wiring.

that's an interesting point. i will try to see if VTEC engages or not soon when i've got the car to safe operation.

---------------------------------------

its good that the car is up and running, but i'd like to get the wires running the way that they should in a safe and reliable fashion. i still want to try and figure out the original problem as i still have a few question marks. have learnt so much about the car already in the 2 weeks that i've owned it (my first honda), thanks to the help of many of you.

so i've plugged the following:
30 - fresh wire to battery w/ 10A fuse holder (not sure if this is right)
85 - fresh earth wire
86 - wire from the ecu that has voltage when key is set to 'on' (i.e. ignition switch)
87 - line to the fuel pump

there are two wires left over that are not being used from the original harness. i have closed their ends for now so there is no chance of contact. these would be 30 and 85, where i have run fresh wire from their sources. this would assume that they would have been ground and constant power. i can't seem to trace them back that far as they get lost amongst the harness. both of them test zero voltage both off and on ignition.

in the stock format, does anyone know the layout of the constant power? e.g. does it go from battery to fusebox to relay? or battery to fusebox to ecu to relay? the origin of the original power wire seems to be the biggest question mark for me.

[dancoco808]

visual summary:

http://img.photobucket.com/albums/v334/coco69/relay_original1.jpg

voltages tested at the relay pin/wire junction to be more accurate
http://img.photobucket.com/albums/v334/coco69/relay_problem1.jpg

also with new battery
http://img.photobucket.com/albums/v334/coco69/relay_fixed1.jpg
car is up and running with this layout

some thoughts:
?safety concerns
?which fuse to use on power wire (currently 10A)
?pin30 power source from battery instead of oem fusebox or ecu
?when the engine is running, there is constant 12v+ out at pin87, how is the circuit i've set up different from running the direct wire from ignition to fuel pump in regards to fuel pump duty/operation

[grifty]

i would double check that the wire to 86 is coming from the ecu. To find the correct fuse rating for the fuel pump i would check the interior or exterior fuse box lid. other than that it looks all good.

[GenesisEG]

As far as the current rating of the fuse, what is the rating of the pump?

Just divide the watt rating by 12 and you'll get the current rating of the pump. The fuse should be slightly higher than that.

I would hazard a guess that the original power to pin 30 would have come from the OEM fuse box. I'm not familiar with the wiring but some cars have a fuse box in the engine bay and one in the dash, check both.

It doesn't really matter, as long as the pump feed is protected, it'll be fine.

As for pin 86... I dunno. I would say it comes from the ECU, it needs to be powered BEFORE the car turns on to pre-pressurise the fuel.


?when the engine is running, there is constant 12v+ out at pin87, how is the circuit i've set up different from running the direct wire from ignition to fuel pump in regards to fuel pump duty/operation

There probably isn't much difference, but since fuel management is run by the ECU, if, per-chance, the ECU wishes to cut off fuel from the engine, and you've run the coil feed from the ignition, it won't be able to.

There should be 12V constant at pin 30, probably straight from the battery, although running it from accessory should be fine too, since accessory is going to be running when the ECU starts.

From pin 87, there should also be 12V constant, as long as the ECU is holding the contacts closed. It probably won't need to open the contacts unless you have a serious problem with air/fuel ratios or something.

Regarding the VTEC issue... I highly doubt the fuel pump relay is going to have any effect whatsoever on it. The relay turns the fuel pump on and off, it can't control how much fuel comes out of it. The only problem you could possibly have is if the pump can't keep the pressure up, and that would be because the pump is too small.

Anyway, let me know how you go. Sounds like it's all under control. Just double check the rating of the pump and make sure the fuse is rated accordingly. Remember, watt rating of the pump divided by 12, that's your current rating. Make the fuse slightly bigger than that.