Subaru Forester Owners Forum banner

1 - 20 of 52 Posts

·
Registered
2018 XT Touring CVT
Joined
·
1,294 Posts
Discussion Starter #1 (Edited)
I am wiring LEDs in series with low beams for DRLs.
Subaru switches the ground on the headlights. The DRL resister is in series with the ground connection.
I will wire LEDs in series with the lowbeams, eliminating the DRL resistor.

UPDATE: DONE, it works and looks great.

LEDs for DRLs:icon_idea:

The 12V LED strips replace the DRL resistor. The low current of the LEDs won't light up the H1 bulbs.
The low cold resistance of the H1 lowbeams has no effect on the LEDs.

I used 24" side emitting 12V LED strips, warm white color, from Oznium.

Cut them to fit if needed, and use silicone sealant. The white stripe lead was the positive.
Use red spade terminals, cut and filed to be skinny enough to plug into the stock DRL plug from the lowbeams.
Looking into the plug, the positive was on the right.
Make a harness, and silicone seal all splices, then tape them up for strength.
Keep the wiring in front of the radiator, not on top, it seems too hot. Secure in place.
Use duct tape on the DRL plug after connecting the spades, to hold them in place, and to weatherseal.
Seal the now unused DRL resister connector with duct tape. It holds and seals better than electrical tape.

I put them along the bottom of the headlights, it's a choice, there are lots of places to mount LED strips.
I left extra wire on the harness for additions. I can unplug this and replug the DRL resister for stock setup.
For show or offroad use only. Do not do this if it is not effective, and/or illegal in your state.
I am not recommending this mod for anyone.
 

Attachments

·
Administrator
2018 2.5i Premium CVT
Joined
·
18,970 Posts
Bare LEDs or a LED strip? Bare LEDs in series would have the same current through them as the headlights--for a moment or two. Till they said goodbye.

I'm away from home now and can't check much further, but I think you could do something like Headlight +, LEDs, appropriately computed limiting resistor, DRL control circuitry, ground.

Or headlight +, 12 volt LED strip, DRL control circuitry, ground. If you could find a bright enough LED strip. It would probably work OK if you left the low beam in the circuit as suggested, since it has pretty low resistance. Though the LEDs wouldn't be full bright. Just put the LED strip across where the DRL resistor used to be.

More when I get home tomorrow and have the chance to look at the diagram.

See my LED tutorial for more info on working with LEDs, both bare LEDs and LED assemblies. It's reachable through this sticky.

There was an error in my statement of the order in which components of the DRL circuitry are connected. See this post for correction.
 

·
Registered
2016 Forester XT
Joined
·
453 Posts
A lightbulb has a variable resistance depending on temperature. As bbottomly pointed out, your LEDs would have full current for a few milliseconds at every start up, shortening hte life of the LEDs.
I recommend wiring in parallel, resistors are dirt cheap and a LED circuits as simple as it gets. If done properly you wire each LED in series such that their total forward voltage is just under +12volts (or 13.5v to be safe) and then use single resistor for the rest.
 

·
Registered
2018 XT Touring CVT
Joined
·
1,294 Posts
Discussion Starter #5
I plan on using LED strips that are made for 12V. I do not need to light up the road at all, just something bright appearing, which fits LEDs well.
 

·
Administrator
2018 2.5i Premium CVT
Joined
·
18,970 Posts
A quick test with my 2003 shows that a rather light duty LED strip lights up fine when wired in place of the DRL resistor. Non-quantitative findings are that it gets almost the full voltage, its resistance being so much higher than that of the low beam filament that the current also flows through. Since current is so minimal, the low beams don't light up at all.

I still need to consult the wiring diagrams and make some better measurements before I sign up to anything! It's the engineer in me. That was also the source of my original concern about blowing out the LEDs. In my terminology, 'LED' by itself means simply a LED, with no surrounding circuitry. A set of LEDs, appropriately configured for operation at 12 volts, is 'a strip,' 'an assembly,' 'an array,' or words to that effect.

bashfulbernie: I had to dig into your profile to find out what year Forester you have. Not everybody would go that deep. Help us help you by filling out the vital details of your Forester in your signature. Year, model, transmission type ought to do it just fine. That way it will automatically show up on every post you make, and we can fine tune any advice on questions you might have in the future.

I'll be back later after more research. But for now, I'd say you're on the track of a winning solution.
 

·
Registered
2018 XT Touring CVT
Joined
·
1,294 Posts
Discussion Starter #7
I edited my profile and signature to read: 2006 XT Manual
 

·
Administrator
2018 2.5i Premium CVT
Joined
·
18,970 Posts
Good move! I was continuing my research on the 2002 that you previously had listed there. But it seems that they're all pretty much the same in the DRL department.

I've been moving kind of slowly today as a consequence of getting to bed at 3AM after flying home from Florida, but I'll try to do a few more tests and take some measurements tomorrow.
 

·
Registered
2018 XT Touring CVT
Joined
·
1,294 Posts
Discussion Starter #9 (Edited)
UPDATE: not going to do this

I imagine I could also have these LED strips blink with the turn signals.
The very small added current should not affect the blink rate much.:icon_idea:
I might do this if I go with amber LED strips.

I would insert a blocking diode in each wire from the lowbeams to each side LED strip.
I would run a wire, with a blocking diode, from the right turn signal positive to the right side LED strip positive, and duplicate for the left side.
I would use a total of 4 blocking diodes. Each LED strip would only have one blocking diode acting in its feed wire at a time, dropping .7V.
 

·
Administrator
2018 2.5i Premium CVT
Joined
·
18,970 Posts
I think the 'always on' aspect of the DRLs would overcome the blink.

I've actually been working on a similar circuit for another member here, but the job got badly sidetracked as a consequence of recently having moved, not being able to locate some of the required components from among my unsorted piles of stuff, interference from other jobs, and the like. I've breadboarded it but never took it any further. It used a 555 timer to give me a 'blinking' signal, and a couple of 4x NOR gates to handle the logic. I'll try to resurrect it. I'll also try to overcome the temptation to ditch the whole circuit and move to using a PIC or similar MPU. I learned early in my career the importance of a design freeze! :icon_wink:
 

·
Registered
2018 XT Touring CVT
Joined
·
1,294 Posts
Discussion Starter #11
You are right, my plan would only work when the headlights were on and DRLs off.
 

·
Administrator
2018 2.5i Premium CVT
Joined
·
18,970 Posts
I've reviewed the diagrams for pre-2003 SF Foresters, early SG Foresters equipped with dual filament H4 bulbs, and the later SG Foresters equipped with separate high and low beam bulbs. DRL wiring is pretty much the same in all of them. (And the only difference I noticed H4 SG vs. separate bulbs SG was where the wiring was ultimately broken out to the separate bulbs.) Headlights are all fed in common on the high side. The low side of right and left high beam and low beam is paralleled with its counterpart and switched to ground to control it. Additionally,the low side of the low beams goes to the DRL resistor and then to ground via the DRL switching circuitry.

Then I tested things with a 6-LED strip that pulls about 65 ma, wired directly across the DRL resistor connector (disconnected from the resistor). It worked great. There was no appreciable difference in the voltage across it compared to the voltage at the battery itself. The low beams appeared to be totally dark.

I can't say whether the 6-LED strip would be ample for use as a DRL substitute. That's a pretty subjective decision (as is most everything else about DRLs). Nor did I try artificially increasing the load to see whether there would be any meaningful reduction in available voltage. But a quick run through the math says that this shouldn't be a problem with any practical number of LEDs.

So, an approach that would work great would be:
Disconnect the connector from the DRL resistor
Seal it off in a plastic bag or equivalent
Tap into its two wires
Connect LED strips in parallel across those two wires, observing proper polarity​

Nonetheless, since you would be running wires from the area of the DRL resistor to the front anyway, I'd go one step further and use the common + terminal on the headlights as the source of power, not the low side of the low beam light. Switched ground would still be through the proper terminal of the DRL resistor connector. This will take the low beam filaments totally out of the DRL circuit and allow your LEDs to operate even in the event of a Murphy's Law failure of low beams on both sides.

There was an error in my statement of the order in which components of the DRL circuitry are connected. The common + approach won't work. See this post for correction.

Now, as for the directional thing. I'll get back onto that. Meanwhile, great idea!
 

·
Administrator
2018 2.5i Premium CVT
Joined
·
18,970 Posts

·
Registered
17 Forester Premium
Joined
·
106 Posts
That was also the source of my original concern about blowing out the LEDs. A set of LEDs, appropriately configured for operation at 12 volts, is 'a strip,' 'an assembly,' 'an array,' or words to that effect.
Did anyone try to attach a resistor or regulator to the led strip ?

It seems that these strips don't have resistors in them, which explains why the failure rate is high.


You would do yourself a favor to run a voltage regulator. The fluctuation of voltage, up or down, can cause the diodes to fail. I have never had an LED strip burn out with a regulator.
First off, I'd recommend getting your LED strips from oznium.com. They sell very good quality strips. Secondly, in order to keep your LED strips from burning out, you should put a 12v regulator in place. Go to radioshack and get a 7812 regulator. Connect the car's power to the "in" of the regulator and run the "out" connection to the LED strips.power.
http://www.radioshack.com/product/index.jsp?productId=2062600
http://www.newenglandsubarus.com/forums/showthread.php?t=12167&page=2
 

·
Administrator
2018 2.5i Premium CVT
Joined
·
18,970 Posts
Using a voltage regulator would certainly be prudent, but I'd probably consider it excessive in most situations.

But resistors? LED strips, arrays, assemblies all have to have resistors (or an appropriate driver IC). The resistor limits the current through each LED to the prescribed ~20 ma. and soaks up the voltage not accounted for by the forward voltage drop of the LEDs. The only time you can get away without resistors is in things like flashlights, where the internal resistance of the specified batteries provides the essential current limiting capability. That's how the Chinese do it!

Edit: I just looked at the datasheet for the 7812, and I think there would be a problem using it in a car. It requires an input of at least 14.5 volts before it will regulate. Up to that point, it simply delivers about 2 volts less than the input voltage, unregulated. An adjustable regulator like the LM117 family could be configured to deliver a regulated voltage in the 10-11 volt range, but this would require external components.

oznium.com - looks good! I'll add it to my Sources Thread.

There's some bad misconception going on over at that NES thread. TNPaparazzi is confused about the resistors. Their use with LED replacements for incandescent blinker bulbs has nothing to do with limiting the impact of voltage spikes. They are connected in parallel with the LEDs in order to pull extra current, so as to fool the circuitry that detects a burned out bulb and makes the blinkers go fast as a signal to the driver to replace the bulb.
 

·
Registered
17 Forester Premium
Joined
·
106 Posts
Using a voltage regulator would certainly be prudent, but I'd probably consider it excessive in most situations.

But resistors? LED strips, arrays, assemblies all have to have resistors (or an appropriate driver IC). The resistor limits the current through each LED to the prescribed ~20 ma. and soaks up the voltage not accounted for by the forward voltage drop of the LEDs. The only time you can get away without resistors is in things like flashlights, where the internal resistance of the specified batteries provides the essential current limiting capability. That's how the Chinese do it!

Edit: I just looked at the datasheet for the 7812, and I think there would be a problem using it in a car. It requires an input of at least 14.5 volts before it will regulate. Up to that point, it simply delivers about 2 volts less than the input voltage, unregulated. An adjustable regulator like the LM117 family could be configured to deliver a regulated voltage in the 10-11 volt range, but this would require external components.

oznium.com - looks good! I'll add it to my Sources Thread.

There's some bad misconception going on over at that NES thread. TNPaparazzi is confused about the resistors.
I wasn't sure if these strips had resistors or not , but it seems that most people had their strips dead quickly.
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=290441182981&ssPageName=STRK:MEWAX:IT

Actually I was referring to "The Internet"(post #18) & "apkarian100" (post#23). They Claim that they never had led strips fail after installing voltage regulator.
If 7812 regulator delivers about 2 volts less than the input voltage and kicks in at 14.5, wouldn't it be suitable for led strips due to reduction of voltage to <= 12-12.5V ?
 

·
Administrator
2018 2.5i Premium CVT
Joined
·
18,970 Posts
The problem is that a car doesn't deliver 14.5 volts. Tops is about 14. But yes, it would smooth out a nasty spike should one occur.
 

·
Registered
2016 Forester XT
Joined
·
453 Posts
No resistor is a bad problem on cheap LED strips. But here is why they do it:
Say the forward voltage of your LED is 2.5volts. If your source voltage is 12volts, and you wire 5 LEDs in series your LEDs now need 2.5x5=12.5volts before they will go into overcurrent mode. So these cheaps strips use 6 or so LEDs to be "safe".
They also often are designed to be run off a regulated power supply, specifically a "constant current LED power supply", mostly for use in buildings.

The problem is that it only takes a few milliseconds of overvoltage to blow a LED, they don't have to actually heat up and burn out to die. Even a tiny resistor would help limit their current and make them last longer. You can still buy a cheap LED strip without a resistor, just add your own in-line as if it were a fuse.
 

·
Administrator
2018 2.5i Premium CVT
Joined
·
18,970 Posts
Right! You should never try to run LEDs directly off of a stiff power supply, even if you somehow manage to perfectly match the voltage. As far as current is concerned, a LED is a short circuit. It needs a resistor or a constant current supply to limit the current.

The only reason they work OK without resistors in flashlights and the like is the limiting effect of the internal resistance of the specified small batteries. Swap them out for a stiff supply and the LEDs would go poof.

My $4.99 6-LED strips from V-LEDs (I guess that qualifies as cheap) have two 680 Ohm surface mount resistors clearly visible between each pair of LEDs. Non-destructively reverse engineering them, I'd say that they're constructed as three parallel groups of two, each group comprising 2 LEDs in series with each other and with a parallel pair of the resistors. Assuming Vf of white LEDs in the 3+ volt range, this works out to about the expected 20 ma through each LED pair, which totals out to around the measured 65 ma. for the strip.
 
1 - 20 of 52 Posts
Top