81 XJ650 - Headlight woes

Having some problems with replacement headlights. For the second time now, I’ve had a Sylvania sealed beam go out on me while riding. The high beam light would be stuck on, and I would get no light regardless of the high beam switch setting. Thankfully, I kept my OEM headlight (yes, the literal 1980 bulb that came with the bike). Popping the OEM headlight back in makes everything work.

There’s just one problem. The stock headlight uh… sucks. It appears to be a straight-up incandescent lamp. The light it puts out is completely inadequate for night riding, and I do not feel safe riding after dark unless I can follow a car. I’d love to replace it with a proper halogen sealed beam, but it has a 40/50W power rating, which doesn’t seem to exist anymore. My replacement bulbs have all been 40/60W, which I didn’t think mattered honestly. I even swapped my tail light, plate light, and gauge cluster illumination with LEDs to save some power overhead. I don’t see why the higher high beam wattage would be killing these bulbs in the XJ electrical system.

Any ideas why these 40/60 units don’t work?

 

There shouldn't be any reason at all for the Sylvania bulbs to be popping like that....maybe their supplier had a bad batch? Poor filament quality or support can cause filaments to break due to excess vibration. "Anti-vibratory" bulbs have thicker filaments and more filament support "guidewires" to deal with such issues.

Although I don't get what you mean when you say that "the high beam light would be stuck on"......the only way that could happen is if:
- the bulb shorted internally
- there is something amiss in your control switch headlight selector switch or the wiring to the bulb.

The stock headlight (and really all others of that era/wattage) "suck" is because they aren't getting full voltage from the battery (assuming your charging system and battery are in good shape)...there are voltage losses thru all the wiring, connectors, switch contacts, etc. Light output is determined solely by the voltage available at the bulb ("visibility" is determined by light placement, which depends not only on the light output, but also by the reflector design and quality, and the lense fluting design and clarity).....bulbs have a "rated" output (such as the 40/60W) at 12.8 VDC (13.2 VDC outside of North America) and any drop in this voltage available at the bulb results in an exponential drop in light output (in incandescent bulbs; LED's are a different story). So, for example, a 10% drop in voltages (from 12.8 to 11.5 VDC) results in about a 30% decrease in light output!

Executive summary:
- make sure all the connections in the circuit are clean-clean-clean and tight
- figure out how to get direct battery voltage to the headlight. This is most easily done with a HEADLIGHT RELAY SYSTEM.....besides drawing voltage directly from the battery, it also by-passes the many feet of tiny wiring and relegates the headlight switch function to being just a low-voltage trigger to turn the relays from lo-beam to hi-beam.
- everything you need to know about vehicle lighting: https://www.danielsternlighting.com/tech/relays/relays.html

 

Interesting read. I understand that light wattage is dependent on the listed voltage. That’s why, for home light fixtures, you can always screw in a lower wattage bulb to a high power fixture, since the bulb will draw whatever it’s going to draw at 110V/220V AC. But exceeding the wattage on the fixture poses a risk, since the power rating of the internal wiring may lead to a wire burning out if the bulb’s power draw exceeds it. That was my initial worry about the 40/60W bulb — I don’t want to burn anything out if the harness isn’t rated for a higher load. But based on what you said, it’s likely I’m not drawing anywhere near 60W on the high beam due to low voltage. Worth noting, I did clean all the connectors in the headlight circuit with contact cleaner. Didn’t seem to do anything, but did make those connectors much easier to take on and off. I really should do the same for every connector on the bike.

With regards to the high beam light, I took some resistance readings on the Sylvania light with my multimeter. Here’s what I got between the pins:


No filaments appear to be broken, otherwise I would get no connection between at least one pair of pins. It makes sense to me that two pairs of pins would show similar resistances. Connect one of those pairs and you’ve got low beam, connect the other pair and you have high beam. But the fact that the top and right pins are shorted seems… wrong? That means connecting either of the “good” pairs to power would also connect the third pin. Just can’t be right…

I will say, the old light is still definitely worse than the Sylvania (when it was working). Regardless of the headlight voltage, the modern unit is clearly brighter on both low beam and high beam. The old filaments are probably just worn out. A headlight relay system seems like a good idea anyway — a brighter light would make me feel much safer at night. I have a question though: if I’m always applying full voltage to the light, do I risk draining the battery while the bike idles and the alternator isn’t generating much power?

 

After reading the Daniel Stern page, I think I know the answer to my question. The relay increases the voltage to the headlamp by removing resistive loads from the circuit. So, there shouldn't be any increase in overall power draw. Sounds like a pretty sweet deal.

 

I have an update. I ordered a headlight relay kit from chacal. The install was fairly simple, though it seems that mounting the new relays with the flasher relay under the tank isn’t an option. Thankfully there was enough room in the headlight bucket that I was able to just… cram everything in there.

The result? Immediate, massive improvement in brightness. I’m sure it would be even better with a modern bulb, but for now, that old one ain’t going anywhere.

My only question and concern: is it going to be okay to just stuff those new relays into the headlight bucket? It seemed like the best way to keep them away from engine heat. The wire length in the kit wasn’t long enough to mount them anywhere near the coils, while still giving enough length to reach the headlight with the three-point connector.

 

I wouldn't.......it gets awful hot in that headlight bucket, too. I would do everything possible to mount the relays somewhere "in the air flow" (even if it's above the valve cover area).

You can use zip ties to mount the relays to the frame tubes anywhere along their length where they'll fit........

 

Makes sense, heat was my concern with the headlight bucket strategy. I’ve come up with an alternative. I basically followed the stock wire bundles down to the metal clips which route them through the bottom of the frame. I was able to wedge the relay tabs into there, then zip-tie the relays to the wire bundles to secure the other side. Since it’s out in the weather, I liberally applied dielectric grease to the backside of the connector. Figure it’ll give them a fighting chance against water ingress.



Still feels kinda sketchy, especially what I did with the power connector in the second photo. I wanted to use the YAMAHA badge across the forks to wedge it in there, so that I’d have some redundant attachment in case the relays come loose. When I do the Dyna coil mod and move that flasher relay, I might take another swing at properly securing this.