Mitsubishi Eclipse 3G Club banner

1 - 20 of 24 Posts

·
Registered
Joined
·
18 Posts
Discussion Starter #1 (Edited)
'01 GS Sportronic 125kmi with an unknown, working auto-start system. (It seems to be tapped into the inspection port in the front fuse/relay block with all other wiring being under dash.)

Headlights are dim; so, I started digging in and was going to go straight for a socket or harness upgrade, possibly a "replacement" harness with its own relays that use the original H4 socket voltages for relay triggering inputs only. However, I wanted to do my homework on it, and did some searching here.

It looks like someone may have changed out the H4 sockets in the past (since there appear to be splices into the harness, and the H4 sockets have longer than expected leads, plus there are no "release tabs" on the sockets).

I removed the H4 socket from the driver's side and probed it with a multi-meter whose common probe was connected to the - battery pole. H4 ground appears good with 0.00V, low beam looks good with 13.7V when lights are on (low beam), but high beam lead shows stable 7.1V with lights on "low" :scratch: and 13.7V on high. This is all with engine off, and those 13.7V readings are within a few millivolts of readings across the battery terminals.

I pulled the bulb and connected it to the socket: readings are the same with bulb connected or disconnected. As one might expect, both filaments light (dimly) with switch in low beam. This ~7V "leak" is preventing me from upgrading with a relay harness b/c the high beam relay will always be closed, and I want to figure out where this voltage is coming from.

High and low beam relays click as expected. (Do those relays just pull out? I was reluctant to pull harder on them to inspect.)

I haven't probed any fuses or relays in the front fuse/relay box.

Any thoughts on where this stray voltage might be coming from? Can a diode hack between high and low beam relays' switched power (designed to light both filaments on high beam) fail this way (passing no voltage "forward" but leaking some "back"). I am not aware of this hack having been installed, but I've seen it referenced here and thought it might be a possibility -- though it'd have to be under/inside the fuse/relay block since I don't see it on top. How about the high beam relay "leaking" even though it's supposed to be open with low beam's engaged?

I think it's most likely that high-beam relay, some corrosion in/on/around the front fuse/relay block, or voltage leaking between wires in the harness, but I'd like any input or personal experience with this issue anyone has.

Alternately, best diagnostics steps from this point would be awesome!

Thanks in advance!
 

·
Registered
Joined
·
18 Posts
Discussion Starter #2
Ok, let me try this way:
Is there any way the front ECU or headlight switch control could cause the front relays to pass this stray voltage into the high beam + feed?

I'm thinking the best next step is to pull the high beam fuses in the front fuse/relay box and see whether the voltage leak persists (i.e., ruling out anything in the relay/ECU/switc). Any electrical gurus that could second that as a good diagnostic next step?
 

·
Registered
Joined
·
18 Posts
Discussion Starter #3
How about the relay question? Do these relays just pull out, or are there tab releases I'm not seeing, or a tool to use?
 

·
Registered
Joined
·
18 Posts
Discussion Starter #4
Pulled out all the relays, and they test fine. Also pulled all the low and high beam fuses. Relay connection points in the box show expected voltages (and ground switches properly by the ECU) as do unprotected side of the fuse connections as long as all fuses are pulled.

However, I find continuity among fuse-protected sides of the fuse connections across low and high beams (on BOTH driver and passenger sides); so, it appears there's some cross (or electronics) in the harness bridging high and low beam feeds. Whatever it is, it only passes reduced voltage: 7V leak from low to high, 2.5V leak from high to low. Thoughts?

I'm right on the edge of throwing in the towel and wiring up a harness that taps into the unprotected side of the fuses (and never putting fuses back into the fuse block so the original light feeds in the harness are never powered up again), but I'd like to actually sort out the problem.
 

·
Registered
Joined
·
18 Posts
Discussion Starter #6
100% stock 00-02 relay box. Continuity among fuse-protected sides of fuses for high and low beams...ALL four fuses (IIRC, fuses 7, 8, 9, and 10). I expected not to be there on ANY of them since it uses separate fuses for each filament, but it definitely should NOT exists across high and low beam fuses' protected sides. I'll dig up some pics, but nothing is modified (in the relay box anyway). I'm tearing into next what appears to be replacement H4 sockets where they are connected back into the harness. The splices look a bit bulky, and I wonder whether there are any diodes in there across high and low beam feeds (that a prior owner might have inserted to keep a HID low beam lit when the high beams are engaged...just a theory -- I have no evidence of a HID kit having been installed).

...and thank you for responding. I was starting to wonder whether I'd get anything on this post.
 

·
Registered
Joined
·
18 Posts
Discussion Starter #9

Hacked together from Tearstone's awesome online manuals! Thanks, Tearstone folks!

This is the schematic for the area I was testing in the relay box. A-08X and A-07X are the low and high beam relays respectively. Their 3 and 4 pins are always hot, and the front ECU provides their 1 pins with ground to engage the relays in response to headlight switch signals. The ECU and ground-side of the circuits are omitted from the diagram. When closed, the relays pass current out their 5 pins to the unprotected side of either 6 and 7 or 11 and 12 fuses. It is expected that 6 and 7 will be continuous on the unprotected side (not red) but discontinuous on the (red) protected side (when the fuses are removed).

With all related fuses and relays (A-08X, A07X, 6, 7, 11, and 12) pulled, the connectors represented by the red dots were all continuous. This would have been a MUCH better test had I disconnected the H4 connectors from the bulbs.

I also checked relay operations and ensured that the 1 connector on each relay did ground in concert with the headlight switch. They both click, and lights operate exactly the same way if the unused fog light relay is substituted for either (i.e., round robin switching of all three yields no changes in symptoms).

With the 6, 7, 11, and 12 fuses pulled but the relays back in, expected voltage reaches the (not red) unprotected side of each of those fuses as and when expected per headlight switch position.

It is only when those fuses are replaced that anything gets weird. With 6 and/or 7 in (and 11 and 12 out), and the high beams on, the red dots on 11 and 12 show 7.3V (stable). With the reverse arrangement, low beams elicit ~2.5V at the red dots on 6 and 7. I suppose each of these symptoms could be caused by a horrifically bad ground in the headlight harness, but the ground at the H4 connectors tests pretty solid (even with the lights on).

I'll test for 6, 7, 11, 12 continuity with the H4s disconnected next, but I'm interested in any thoughts.
 

·
Registered
Joined
·
32 Posts
I would use a jumper on the ground side just to be sure. As you probably know when electrical problems get really strange it can be traced to a bad ground in allot of cases. I would also be interested in what happens if you replace one fuse at a time and take your voltage readings again.
 

·
Registered
Joined
·
791 Posts
Great diagnostic so far, seems definite that it's something on the headlight side of things. I'm wondering what'd happen if you ran new wires between the fuses and headlights. Are there any burns inside the harnesses to the bulbs?

How'd you pop the relays by the way? Was there a release tab?
 

·
Registered
Joined
·
18 Posts
Discussion Starter #12
I'm glad someone else was unable to find that information: No release tab on the relays -- just grip, and wiggle gently but firmly while pulling out.

No obvious harness damage, but, as I wrote above, I'm suspicious of the splices where what looks like replacement H4 sockets were installed.

Unless other evidence manifests, I agree that it pretty much HAS to be something in the harness between fuses and lights and/or a horrifically bad ground.

Lug: as I replaced fuses, things pretty much stayed the same: insert a how beam fuse, turn on the lows, and both high beams' protected fuse conductor will show 2-3V. Reversed for highs yields 7V on low. Weird. :scratch:

Thanks for the responses, folks! I appreciate it.

(Wow! There's a lot of smiley's available on this board!)
 

·
Registered
Joined
·
32 Posts
I should have been more clear. I meant remove fuses 6,7,11 and 12 and the reinsert only one fuse check your voltages again. Repeat until you have checked voltage with each fuse being the only one of the four installed at that time. This should help to narrow down where the problem is.
 

·
Registered
Joined
·
18 Posts
Discussion Starter #14
You were clear. I had done that. I think I recorded the original test results poorly :rolleyes: : there is no continuity across sides, only among high and low beams on the same side.

It turns out that the bulbs cause all the "weird" readings I was getting. I tested one on a motorcycle battery. With a common ground and the low beam powered, ~7V leaks into the high beam circuit (on a bare bulb with ground and power test clipped to battery)...both filaments light. With the high beam powered, only the high filament lights but between 2 and 3V leaks into the low beam side.

What's good about that is that it (the voltage leak) disappears from the harness when the bulbs are disconnected; so, a standard relay harness (that uses OEM harness feeds only to switch the replacement harness' relays) will work.

I got some additional (though still insufficient) light output by directly grounding the H4 connectors to nearest grounds. I think even more will be gained with a whole headlight harness upgrade (either from the fuses "out" or a whole relay harness). I'm going to compare output on a bulb powered by a fuse tap to output powered directly from battery; if not sufficient, I'll go the whole relay-harness route...otherwise, fuses out only.

Finally, even after having read all the reports about the undersized wires in the stock harness, I was downright astounded and amazed at how embarrassingly, maybe dangerously light duty the OEM headlight feeds were (when I opened the splices to the upgraded H4 connectors). There is no way these wires should be powered by 10amps: 20 gauge at best, probably lighter.
 

·
Registered
Joined
·
18 Posts
Discussion Starter #16
I finally found the issue, and I'm as sure as I can possibly be without having done the work to fix it. I cannot believe I overlooked it, and I even violated one of my own diagnostic rules in doing it: keep looking at the suspicious things, and check all prior work. I should have kept looking at the replacement H4 sockets. :noplease:

I got all the rest of the hardware I needed to make a relay harness with Hella relays, 12 gauge wire, 10 amp ATM fuse taps, and new H4 sockets. I wanted to avoid the expense of the relays and their pig tails if possible but wanted to be absolutely sure I could get adequate light from the four fuse supplies in the relay box (being unsure about the quality of ANY of this OEM wiring at this point); so, I connected one fuse tap to a test clip and compared low beam output from that vs straight from the battery. They were indistinguishable; so, I wouldn't need the relays. To be double sure the work would be worth the effort, I compared to what my existing harness with upgraded grounds was outputting. It was also indistinguishable from straight battery power! :shoot2:

It was a WTF moment. So, I tested the bulb on my Maxima's battery -- same output, same characteristics. Stuck it in the H4 connector on the Max; holy sh1t nuclear explosion of light! Ok, something's way off here, and I'm obviously confused. I need to think. I start packing up my tools but decide I'll probe the H4 sockets on my Maxima while operating the lights (to see how it's putting out so much more light at the sockets than when testing on the battery). Well, that car does a bunch of weird stuff when switching lights, but, bottom line, it was using the bulb completely differently than the Eclipse.

I'm totally baffled, and drive home, but, of course, it's just. eating. at me. So, my brain is on slow simmer and I decide it's time to start drawing this out. I try to figure out how I would wire up a series of relays to emulate what the Maxima was doing (and gave up on that...it's weird), but it got me to thinking about how the filaments work and the potential for running the current backward through them (which is part of what the Maxima does). Then light dawns on grassy knoll, and I just. know. The dorks who replaced the H4 sockets wired them up wrong, and I never checked that the socket pin outs matched what they are supposed to be. I had always just assumed one wouldn't bother doing a rewire wrong (esp on both sides the same way).

You remember the circuit:

Here is the H4 connector:

The green is the way it's supposed to be wired:
Pin 1 is supposed to be the low beam hot
Pin 2, the ground
Pin 3, the high beam

The red is the way it's actually wired
Pin 1 to ground
Pin 2 to high
Pin 3 to low (round robin style)

A decent pic showing the layout of H4 filaments:
low beam filament has the shield and both filaments share that common ground lead down the right side.


You get the picture:

The "low beam" power is energizing the H4 bulb's high beam pin. Current flows through the high beam filament, over the common ground in the bulb, and out the grounded low beam filament. This lights both, weakly, and the high beam filament lights lazily.

High beam energizes the H4's ground pin, and current escapes to ground "backward" through the low beam filament to the grounded low beam pin.

Derp, de derp, DERP!!! (Could hurt some prior owner right now.)

Thanks everyone for your help!
 

·
Registered
Joined
·
791 Posts
Wow. Can't believe they wired the new bulb sockets in wrong. That's probably something I wouldn't have checked either- I mean, who does that kind of thing??? Obviously the previous owner of your car did! :D

Are you still making your own harness?
 

·
Registered
Joined
·
18 Posts
Discussion Starter #19
HRT, almost no chance I'll make a harness unless rewired sockets produce poor light... too much hassle. I'd only make my own for satisfaction sake since parts cost almost what I could buy one pre assembled with same materials. Also, most fuse taps will not work to replace all four fuses in the relay box due to physical constraints. A harness would have to splice into OEM harness or plug (signal wires only) into existing H4 socket...my plan if I have to make one.

OK, Bitter :rolleyes: Where was that nugget of wisdom last month? LOL.
 

·
Moderator
Joined
·
54,393 Posts
Sorry, you didn't ask me for help and I can't baby sit each and every help thread. A fisherman does not catch every fish in the sea, a cop does not pull over every speeder, and I do not read every problem thread.
 
1 - 20 of 24 Posts
Top