theorie

iTrader: (31)

Here's a strange one. I spent the whole day yesterday trying to figure this out. Let's see if anyone else has had this issue or can shed some light on this...

Background:
I stripped one of my FDs to the chassis (everything except the windshield/rear glass came off the car), and started building it as a 20B "weekend warrior" car. 1/2 way though, I decided I wanted to use the car as more of a daily driver (putting the 20B in my other FD), so I finished reassembling this car with one of my 13B motors. Here's the before and after:

Beofre:


After:


I finished assembling the car and it's back on the road, with one strange issue...


Issue:
With the car running, turning-off and removing the key does not turn off the ignition and the motor will continue to run. The Power FC commander stays lit-up and functional, and the odometer stays lit-up and displaying mileage -- I call this "zombie" mode. In this state (with the key out / motor running) there is a slight dimming to the odometer backlight, and the motor seems to have weaker spark.

If the car is already off, turning the key forward two clicks (to "run" position) without starting the car, then turning back to "off" also results in the ignition staying on "commander on, odometer on, dimmed lights).

The easiest way to kill everything and "reset" the state to "off" is disconnecting the battery (I have a breaker installed next to the battery). Pulling the "ROOM" fuse also has this effect.


Assumptions:
I'm guessing there is some sort of short between a constant-power wire and an ignition-switched relay-powered wire, but I haven't identified the exact spot where this would be happening to cause this state. It seems like it might be related to the "ROOM" fuse circuit, but perhaps it's coming from elsewhere.


Observations:
#1 If left alone, the ignition will eventually turn off (as evident by odometer and commander screen going dark). This ranges from anywhere between 3 seconds to 3 minutes (see #2).

#2 The longer the car is turned on (keyed-forward to "run" position, with or without motor actually running), the longer the ignition stays on after keying-off. It's almost like there is a capacitor on the ignition...it charges when you key on, and it then discharges (keeping ignition powered) after keying-off until it's fully discharged. No, I don't have a turbo timer. I also haven't identified anything on the ignition circuit that could do that.

#3 In the zombie state, pulling the EGI relay will turn off the ECU and kill the motor (if it's running) but it does not shut off the ignition, and if you plug in the EGI relay the commander comes back on (with the key still out).

#4 In the zombie state, pulling the "ROOM" fuse will kill the ignition and reset it back to "off". I saw Dale's thread regarding a drain on the "room" fuse, but that seems to be a different type of issue.


Notes:
#1 I've tried swapping all the relays with ones off another car - problem persists. I also tried swapping the alternator - problem persists.

#2 I've fully-disconnected all accessories (radio, gauges, boost controller, etc.) - problem persists.

#3 The fuel pump is rewired directly to the battery with a relay (relay is activated by power coming directly off ignition switch harness). Yes, zombie problem still persists even with the relay disconnected.

#4 As part of the build, I relocated the front fuse box (the skinny one by the hood latch) and also the fan relays that sit near the passenger-side fender liner. I've triple-checked all the wiring and everything is good with these relays. Side note: as I understand the diagrams, none of these relays should have an effect on ignition anyway -- the car will run without them, you just won't get very far when the fans don't kick on and the car overheats.

#5 Pulling the "ROOM" fuse will kill the zombie state, but if you key-forward with the fuse removed (accessory or run), both of the fans instantly turn-on to high-speed. Is this some kind of safety feature of the ECU?

Note that the "ROOM" fuse powers the interior lighting (glove box light, ignition key light, door/cargo lights) but it also seems to be upstream of the engine control, fuel system, ignition system, and dash cluster / warning lamps. Hence why I suspect this circut -- all those systems are remaining on with key-off.





As I mentioned before, pulling the "ROOM" fuse tells me there is something shorting upstream of the "ROOM" fuse (i.e. between the fuse and the battery)...but where?

Any thoughts or suggestions on how to trace down this issue?

fendamonky

iTrader: (6)

I've had this exact same problem with my car. I believe the culprit (for mine) is actually somewhere in the front body harness. My issues initially started when I was in the UK, the car would randomly get power when ambient moisture ramped up, or when I had been driving in the rain, or when the car got washed. I did NOT trust the electrical/wiring skills of people on that island, so I learned to deal with it by just disconnecting the battery. Eventually it became so bad that sometimes it would stay on regardless of if it was wet out or not.


When I began to pull the body harness I noticed that I had a LOT of corrosion going on in there, so I'm thinking that was the cause of my issues. Unfortunately my car has been down for quite a while after my engine died (life got financial priority..) so I'm not sure if swapping the harness will provide a fix.


I would check there though, IIRC the long rectangular box sitting near the hood latch may be a good place to start.

Hope you get it sorted!!

Ceylon

iTrader: (3)

Have you done much work on the harnesses while they were out? Simplifying etc.

Sounds like you'll likely have to start tracing wires, starting with any part of the harness that appears to have damage to the insulation.

Failing that, you'll have to isolate Room Fuse circuit and do some continuity tests to see which wire is connected back to live. If you have one, a tone wire tracer may help.

Nice work on the car btw, looks mint .

Not sure I'd judge an entire countries electrical skillset based on one car . FD's are cheaper over here, though, so more available to bodgers .

fendamonky

iTrader: (6)

Nah, just the local "experts" where I was at in E. Anglia... I had a weird short that cut my dash lights and rear running lights when my headlights were on. The local auto shop that was recommended to me by a mechanic I trusted "fixed" my issue by adding about 1 meter of wiring and circumventing my dimmer switch.

I was DDing my FD so I wasn't about to run it down to RE:worx (from St. Neots, Cambs) for them to chase wires and guess at solutions. I didn't have the time, nor the quid, to do that!!

Neutron

iTrader: (10)

Have you by chance checked to see if you are getting voltage through your grounds when the car is in this state? Something, like water, could be conducting voltage from a constant power source and back-feeding enough voltage to cause your car to stay on. If this is a relay you would probably see some voltage through a ground wire.

I have seen something similar when a rubber boot to one of the radiator fans had filled with water. Very similar symptoms to what you are experiencing. Took forever to find the culprit.

theorie

iTrader: (31)

Problem solved. CPU2 (not the flasher unit, but in the same black box) was backfeeding power. Swapped in a spare and car shuts off just fine now.

caredden

iTrader: (148)

Nice to see some resolution in these threads. Very helpful. I'm glad it worked out for you.

theorie

iTrader: (31)

I seriously checked all my relays, all my fuses...everything, and couldn't figure it out. Then I gave up and moved on to something else...I finished installing an aftermarket keyless entry (not an alarm, just keyless).

I had just finished wiring up a mini-relay to send the security system the "ground-pulse" (which disables the factory alarm). Shortly after the car alarm started going-off randomly (honking, pop-ups actuating, etc.), even without opening the door.

I traced the alarm problem to to CPU2, and when I popped it open it smelled like some of the capacitors had gone bad (similar to what happens with the capacitors in the gauge cluster). I swapped in ECU2 board off my other FD, and found that the alarm problem was fixed, and also the ignition issue was fixed as well.

So yeah, if anyone runs into this problem...check CPU2.

Ceylon

iTrader: (3)

Glad you found it, problems like this can be a right pain to locate the source .

fendamonky

iTrader: (6)

Good ****! I'll def add that to my own list of things to look at before firing the new engine up.

Cheers