Issues with LED High Beams (9005) Non DRL

[covertrussian]

I just got a LED high beam bulb set ~20watt/bulb from Sealight. After installing them, with low beams on, the high beam dash indicator is partially on and the LED's themselves are partially on too.

 

Switching from low beam to high beam works perfectly fine, the high beam LED's and dash indicator illuminate properly, but when I switch back to low beams, the headlight ON indicator disappears (the bottom left corner of the dash one).

 

If I remove one LED bulb and install the one stock 60w bulb, everything functions fine. It's just odd that high beams would try to be on when the switch is set to low beam only.

 

I know that LED's don't fare well for high beams on 08+ Legacy's, since they use high beam bulbs at half power for DRL's, but 2005 don't do that, so my copious amount of searching has found nothing thus far.

[nevets27]

Don't have any good or solid info but from what I've seen (on various cars) it's because there is no load or the leds are feeding back power to the dash.

 

I think if you put in a diode or a resistor (depending on the actual problem) it might fix it.

 

Just a thought or path for you to look into.

 

Sent from my LG-US998 using Tapatalk

[covertrussian]

One of the things I found so far is, since leds barely take any power, bleed over (or backfeed) from another bulb could be causing it. Could be ground related, or a small resistor is needed.

 

Similar issue from motorcycle guys with resistor fix:

https://www.r3owners.net/threads/high-beam-indicator-problem-using-led-headlights-solved.30241/

 

RX8 thread, from post #10 on:

https://www.rx8club.com/new-member-forum-197/strange-high-beam-indicator-light-262629/

 

At first I figured it's similar to canbus issues, but since 2005 don't seem to have lights on canbus, that wouldn't be it. Once it warms up I'll start taking fog and corner bulbs out to see if that solves the backfeeding.

[apexi]

One of the things I found so far is, since leds barely take any power, bleed over (or backfeed) from another bulb could be causing it. Could be ground related, or a small resistor is needed.

 

This probably won't help, but I'll mention it anyway..

 

I remember a few threads where people complain about one headlight bulb being full brightness, and the other side illuminating dimly. I guess that's what happens when 1 of the 2 headlight fuses in the engine bay blows. So that makes me think the headlight bulbs can backfeed each other? I don't know, over my head..

 

Edit: I'm an idiot, you're talking about high beams, and I was thinking low beams.

Edited November 16, 2018 by apexi

[xt2005bonbon]

One of the things I found so far is, since leds barely take any power, bleed over (or backfeed) from another bulb could be causing it. Could be ground related, or a small resistor is needed.

 

 

I just switched my front end with a face lifted jdm one and had to wire the OEM HID low beams and so forth. While I was doing that, I learned that the low and high beams have their grounds going through the DRL module (see page WI-141 of the FSM). So there are probably gremlins in that box . All I know so far is that the DRL module switches the ground for the low beam when you turn the combination switch two notches. I believe the DRL module adds a resistor in series for the low beam so a lower voltage is applied for DRL purposes. Now this does not answer your question. I wonder what would happen if you isolate the ground of the high beam LED from the DRL module and connect it directly to ground. Worth a try.

 

 

Edit: actually, looking again at the page from the FSM, my suggestion may not work. It'll probably leaves the bulb on all the time with the low beams. Looks like it is a red wire that goes to the DRL module that should be separated from the DRL module? Your car may go boom if you try though

Edited November 16, 2018 by xt2005bonbon

[covertrussian]

My DRL's are actually disabled, via unplugged harness under dash. I have an idea on how to test this, but I'm trying to find some decent (and cheap) low beam and fog light LED's first :-D.

[xt2005bonbon]

Why don't you check the voltage at the USDM headlight wiring harness (low and high beam) for each combination switch position? Wouldn't hurt.

[xt2005bonbon]

Update?

[covertrussian]

I'm hunting for some decent, reasonably priced, H7 fanless LED's first. I'll install them together and see if it still grounds through the bulbs, if it does then the fun will begin.

[rougeben83]

Had exactly the same problem. Voltage bleed is coming from the foglight circuit (it's connected to the highs to switch off when highs are on). I think it has something to do with subaru using the ground for switches, not the other way around like in most DC circuits. If you do the independent foglight modification, it should eliminate this problem - it did for mine.

 

Oh, and my DRL's have been disabled for the past 3 years, FWIW...

[covertrussian]

Had exactly the same problem. Voltage bleed is coming from the foglight circuit (it's connected to the highs to switch off when highs are on). I think it has something to do with subaru using the ground for switches, not the other way around like in most DC circuits. If you do the independent foglight modification, it should eliminate this problem - it did for mine.

 

Oh, and my DRL's have been disabled for the past 3 years, FWIW...

 

That's exactly what I figured too. I was gonna unplug fog bulbs and see if it fixes the problem, if it does then find a better way to ground stuff.

 

Good to know on how the circuit works though, makes it harder to actually properly reground them .

[Dishwasher]

Good H7 LED's don't exist. Stick with halogens or go with an HID conversion (including projector swap).

 

Go down the Test Lab rabbit hole if you dare:

 

https://www.youtube.com/channel/UCVOiJIYQyCbXwkrKYXw8Thg

Edited December 5, 2018 by Dishwasher

[covertrussian]

Good H7 LED's don't exist. Stick with halogens or go with an HID conversion (including projector swap).

 

I will disagree with that, at least as of this year, prior to this year LED's were really bad. Check out this page: http://bulbfacts.com/led-kits/chart/projectors.html

 

He uses H7 projector for testing, even $50 LED's are doing real good, but they all with fans. I don't like the noise from those, plus there is not much room in there.

 

I'm eyeballing LED's for amperage reduction reasons

[nevets27]

For the record 9011 HIR bulbs are an easy and good high beam upgrade. But LEDs are improving and at some point will likely be a good replacement.

[Dishwasher]

Those output photos all look terrible on that sight, splotchy with a weak hotspot. I wouldn't buy any of those, but to each his own.

 

I'd check out hidplanet.com and headlight revolution for more info on LED testing. I like the idea of reducing wattage (although you run into issues in the snow as LEDs don't generate enough heat to melt snow). The problem is that LEDs don't simulate filament placement well enough. They'll get there eventually, but they aren't there yet, even in 2018. Diode Dynamics SL1 is probably the closest to a good replacement, but they don't make it in h7.