On 3/16/03 Tom Barber (aka Barbarian) from the FJROwners online forum sent in this Web page describing how he modified his FJR such that the "low-beam" switch position actually partially lights both the low-beams and the high-beams, thus providing better nightime lighting without blinding oncoming traffic.

The three pictures following were taken with the FJR on the center stand, which alters the pitch of the bike enough to cause the lights in all three pictures to appear to be aimed noticeably lower than they actually are, and thus simulates the effect of moderate braking.

The picture above is of the high beam pattern. Note that the fire hydrant is brightly lit from bottom to top, and the curbs on both sides of the road are easily visible far beyond the hydrant. Take note also of the depression in the road near the hydrant. Although this depression is actually a utility access cover, it could just as easily be a road hazard such as a deep pothole or patch of sand.

The picture above is the original low beam. The hydrant is barely visible, the curb on the right side vanishes into the darkness just a few meters beyond the hydrant, and the curb on the left side is hardly visible at all. The simulated road hazard is practically invisible. Under hard braking, the downward pitch of the bike would be much greater, and both the hydrant and the road hazard would be in total darkness.

The picture above is the modified low beam, taken with the same or similar exposure as the other two, as evidence by the similar brightness of the fixed lights along the horizon. The apparent difference in color temperature (average wavelength) is real and not the result of some error or change in the camera settings, which you can confirm by comparing the hue of the lights on the horizon. The color temperature is lower in this picture because both filaments are actually being used and they are both operating at a lower temperature. If you compare the hydrant and the curb with the other two pictures, you will observe that the hydrant is lit from bottom to top as it is with the original high beam, but it is lit much less brightly than it is with the high beam. Similarly, both curbs are visible well beyond the hydrant, in stark contrast with the original low beam, but not nearly so brightly as with the high beam. The road hazard is much more visible than it is when using the original low beam.

There is another significant problem with the FJR's low beam pattern, which is not obvious in these pictures. When the bike is leaned hard in a right-hand turn, the pattern rotates clockwise, causing the distance for which the right edge of the road is visible to be substantially shorter yet. Leaning the bike to the right also causes the very bright band of light that extends outward to the left, to shine directly in the eyes of the oncoming motorist. If the oncoming motorist can see your headlight before you initiate the lean to the right, they will observe a sudden and pronounced increase in the brightness of your light, and they will naturally conclude that you have switched your lights to bright.

Part of the reason why I resolved to do something about this is because I have been irritated by the exact same situation on the CBR1100XX, which has a similar low-beam pattern, since 1997. If you should be of the opinion that these problems can be effectively mitigated via simple adjustment of the headlight aim, I assert that this is not the case. To be sure, no matter how high the light is aimed, the low-beam pattern will still end abruptly some distance ahead, and that distance will still be way too short during hard braking. Furthermore, aiming the headlights higher will only exacerbate the problem of blinding the oncoming motorist whenever leaning to the right. This problem is very real notwithstanding that it is not a visibility problem experienced by the motorcyclist.

At this point, I should probably digress and say something about the technique that I use to aim the dual headlights. On a bike with a single headlight, all one has to do is go for a ride on a straight and level road to determine whether the road is lit best when accelerating, decelerating, or riding at constant speed, and then adjust the aim accordingly. The dual design makes it more difficult to gauge exactly how each light is aimed. In order to see exactly how one or the other of the two lights is aimed, you have to disconnect the other light. Doing the simple road test but with the two lights alternately disconnected will reveal whether or not they are each set to the correct height, and will additionally reveal whether or not the bright center spot of each light is properly centered on the road straight ahead, and not aimed off to either side of the road.

Subsequent to the road test, I perform the additional procedure of shining both lights together on the garage wall and making sure that the two narrow bright bands are stacked one above the over. The dim region between the two bright bands should be small but discernable. If I see two separate bright bands, I will cover one headlight and make sure that the higher one is the one on the right. If the left one is the higher one, I will raise the right side and lower the left side so that they change places. If I don't see two separate bright bands, I will raise the one on the right and lower the one on the left until two separate bands appear.

I desired to modify the headlights so that when the headlight control switch is on the low position, the high beam filaments would be partially lit. Because the high beam coverage includes the region that is covered by the low beams, the two low beam filaments would ideally be dimmed somewhat, the left one more so than the right one. This would be straightforward if it were possible to analytically determine what the voltage across the individual filaments should be when the switch is on the low position. However, this analysis is inordinately complex, such that only an experimental approach would likely be successful. The resistors used to divide the voltage would also consume power that is precious on the FJR.

I therefore decided to install dimmer controls, one to be shared by the two high beam filaments, and one each for each of the two low beam filaments. These dimmers would be functional only when the switch is on the low position, and would have no effect when the switch is on the high position. Initially I tried to find an old-fashioned rheostat (a length of resistive wire coiled around a heat-tolerant ceramic core and with a movable contact), but I discovered that they just aren't made anymore. They have been replaced in DC applications by PWM (Pulse Width Modulation) devices that achieve a similar result by opening and closing the circuit at a fixed frequency and varying the proportion, of the cyclic period, for which the circuit is closed.

I sent Mike Coan of Warm-n-Safe an e-mail asking him if he knew of any potential issues with using the Heat-Troller as a headlight dimmer. Its current rating was more than adequate at 16 Amps, but the problem that his engineer discovered was that their latest version operates at 1 Hz, which is very efficient for heated clothing but would produce a strobe effect if used as a light dimmer. Mike found an earlier model that he believed operated at 100 Hz or more, and I purchased it from him. At about that same time I found an electrical supply house that sells a DC dimmer that seemed to be suitable for this application, with operating frequency of 400 Hz and also rated at 16 Amps. This company is Euramtec, which distributes the part number A-9040 made by AEC. I ordered a couple of them and figured that between the HeatTroller and the AEC part, one or the other should work adequately well.

A requirement that applies to both the Heat-Troller and AEC device is that they must be placed between the load and ground, regulating the ground connection for the load. That fact, together with the fact that the two filaments in each H4 bulb are joined inside the bulb to what is by default the common ground, meant that the polarity of the headlight connections would have to be reversed. That is, the common ground would have to be converted into a common +12V supply that arrives via headlight relay #1 (controlled by the ECU), and the wires that by default supply +12V separately to the high and low filaments (via relay #2) would have to be converted into separate ground paths for the high and low filaments. Headlight relay #2 would have to be re-wired so that it would switch the ground path alternately to the high or the low filaments instead of switching +12V as it does normally.

Additionally, the high beam indicator lamp is supplied +12V via the high beam output of relay #2. Since the high beam "output" of that relay would be converted to supply a switched ground for the high beam filaments, the polarity of the indicator lamp would also have to be reversed. Alternatively, relay #2 could be replaced with a DPDT relay that would provide a separate switched circuit for the indicator lamp, or a supplemental SPST relay could be used for the indicator lamp. The cheapest and simplest solution was to make some trivial changes to the meter panel's printed circuit board, so that the side of that bulb that is presently grounded is isolated from the ground, and the ground is replaced by a connection to a location on the board where +12V is present whenever the main switch is on. Although this was the cheapest and simplest solution, it had an undesirable consequence as I learned.

I decided to use all three of the PWM devices that I had. The Heat-Troller would be used to provide a supplemental ground path for both high beams, by-passing relay #2 so that this ground path will be available to the high beams even when the headlight switch is at the low beam position and relay #2 provides a ground path only for the low beams. The two AEC components would be individually inserted into the separate ground paths for the low beam filaments, i.e., each would be placed between one of the low beam filaments and relay #2.

The two pictures above are the schematics for the two supplemental wiring harnesses needed to do this, showing how they must be joined to the stock wire harness, that runs across the front cowling, at the headlight relays and the headlight connectors. The two supplemental wiring harnesses are joined by a 6-conductor connector, permitting the dimmers to be located under the seat using also an extension cable that is inserted between the mated 6-conductor connectors.

The picture above shows the AEC dimmers together with the plastic boxes that I rigged to hold them.

The picture above shows the finished harness section that contains all three PWM devices, ready to be connected to the other half of the supplemental harness via the extension cable.

The picture above shows the supplemental harness #1 joined to the stock wiring harness. I made it longer than was necessary in the middle section, but that is okay because the excess wire fits easily into the recessed area between the right and left headlight. Note the section wrapped in yellow tape toward the left, which is the wire bundle to the auxiliary relay. The auxiliary relay that I got at Radio Shack has a plastic tab with a hole and the perfect location is screwed down to the windshield motor apparatus at one of the four bolts that hold that apparatus to the aluminum stay. This harness section also has, from left to right, battery + and - terminals, a 4-conductor cable that provides redundant + and - connections to the Heat-Trollers used for the heated grips and the vest, the power supply for the V1, the 6-conductor connector that joins to supplemental harness #2, and the connector for the GPS power outlet.

The picture above shows the reverse side of the meter panel's printed circuit board and the simple change that I made to it.

The picture above shows the placement of the dimmer controls under the passenger seat. The boxes for the two AEC dimmers that control the low beam filaments are a perfect in that location, with the knobs even protected by the flat surfaces that support the rubber feet of the passenger seat. The smaller knob on the shaft of the Heat-Troller's potentiometer is mounted on the vertical plastic part, almost directly under the far end of the mystery rubber thingie and near the upper right corner of the picture. The 6-conductor connector is located just off the right edge of the picture, on the other side of the cross member from the Heat-Troller's potentiometer.

There is one minor annoyance. The high beam indicator lamp is dimly lit when the switch is in the low position, just like the high beam filaments themselves. As long as both it and the high beam filaments share the primary ground path through headlight relay #2, the secondary ground path through the Heat-Troller will also be available to the indicator lamp. This can easily be corrected by using a miniature SPST relay to provide a separate circuit for the indicator lamp, switched via the headlight switch. If I had done this initially, I could have wired the miniature relay to switch +12V and left the polarity of the indicator lamp unmodified. When I next have a reason to remove the front cowling, I will add a miniature SPST relay, but I will leave the printed circuit board as it is now and wire the SPST relay so that it closes the ground circuit for the indicator lamp under control of the headlight switch.

Notwithstanding the issue with the high beam indicator lamp, which is truly trivial, this project was 100% successful when viewed as a personal solution. I spent a good deal of time on this project, but I will be rewarded each and every time that I ride the FJR after dark, for as long as I own it. Even though I never had any expectations of this being a solution that other riders might readily adopt, I do have some regret about not having produced a solution that would help anyone else. Had this been something that others could do in an hour or two, there is no doubt in my mind that many others would do so and would be entirely pleased with the result. Oh well. For those of you that actually read this, at least the exercise was at least moderately educational; dare I say, enlightening?

Copyright © 2003, by H. Marc Lewis and Tom Barber.
All rights reserved.