Before I begin, let me start with a disclaimer: This modification most probably will kill your bulbs and headlight housings prematurely. Running two filaments at once will cause the bulbs to run hotter, and lead to heat related side effects. This is the same as running higher wattage bulbs in stock headlight housings. So if you find your headlights retaining moisture and / or bulbs burning out quicker, I will not be held responsible. This DIY is about extracting the most out of a H4 setup to simulate a double barrel setup without adding too much complication or load to the electrical system.
With that out of the way...
I always hated my OHC headlights functionally. No matter what I tried, the lighting always seemed inadequate. I've tried all sorts of wattages, and technologies and every single one of those ended up sucking. I figured this was because:
1. The "open reflector" [for the lack of a more accurate description] design was not meant for high speed night driving.
2. Barrel type housings do a better job of focusing the light output.
3. Most importantly, the H4 design forces the user to choose between spread [low] and throw [high], but doesn't give you both at the same time.
This is where double barrel designs with individual bulbs for low and high beams excel. On Low beam, only one circuit is active, but on High beam, both circuits are active, thereby providing both spread and throw at once.
The same is accomplished to a certain extent in H4 setups by using Aux lights to provide spread while High beam is active.
What I wanted to do was to achieve the double barrel effect without adding Aux lights and stuff, but extracting the most out of what already existed from the factory. I was also certain that this would work nicely because if I held the headlight switch in the "Flash" position, where both circuits are simultaneously active, it would make things much better. The only problem being I that could not lock the switch in that position.
So I decided to create a one way bridge between the High and Low circuits.
Here is a simplistic schematic of the stock setup:
My first thought was to add a diode between the two circuits like so:
The problem here is that when High beam is active, Low beam is powered by the same wire that carries current for High, and the designated Low beam wire does nothing. This is liable to place load on the High beam circuit and probably burn out the wire. Not good.
The next solution was use a relay instead of a diode, with the low power side of the relay being powered by the High beam circuit, and using the high power side pf the relay to power the Low beam circuit. This places negligible additional load on the High wiring while providing a safe path for the Low beam current.
With the choice of solutions decided upon, the fun bit was to implement it elegantly. That meant:
1. Keep it stealthy
2. Keep the footprint small
3. Don't cut any stock wires
This is the result:
I used a relay and coupler from a wiring harness I had lying around. Next, I stripped the ends of the wires, keeping the exposed wire long enough to properly slide into the factory coupler's spade terminal for a proper lock. The rest of the wire was routed through some blank holes in the coupler that seemed to exist for this very purpose. If these blank spots did not exist, It would have been a major pain to get the coupler to lock properly in place. Thank you Honda.
The end result is total stealth.
After this, once I fixed my long pending headlight alignment, the results are absolutely stunning. I did a night drive from Bangalore to Rajahmundry via the dual carriageway via Mulbaggal and Palamaneru, and was able to easily keep speeds above 100 through the twisties without having to squint and strain my eyes. Never before have I had such stress free driving on a dual carriageway at night.
What makes the whole thing sweeter is the fact that 60/55 bulbs are performing better than 100/90 and at par with 120/100 without requiring additional wiring clutter in the engine bay. Technically, this isn't 60/55 anymore, but 115/55, with the difference being that the 115W is distributed between spread and throw, unlike a conventional 120/100 setup where you get 120W of throw and 0 spread.
What remains to be seen is how the additional heat is going to affect the bulbs and housings. Whenever I find out, I will update this thread.