Need Help Finding a Momentary OFF Switch

[Fred W]

I think it just comes down to too much power draw at start up for the battery. You don't see it on cars with aftermarket HIDs because they have bigger batteries and the lights don't come on right at start up.

Hmmm. On my Saab 9-3 with HIDs the lights fire up at key-on and stay on during cranking. I'm guessing this works OK because the HID firing current required is more than the run voltage. You might be able to do the same thing on an FJR by rewiring the HID relay to a switched 12V source instead of using the headlight connector signal. But the added load of the headlights during cranking may be a problem for the little battery. I assume that is why they designed the lights to come on only when the engine is running.

 

[KnuckleHead]

I assume those are factory HID's you are talking about. I am sure that all these things are taken into account when they are designing the systems. Aftermarket HID's however dont have the luxury of knowing the exact layout of how they are installed.

 

[Fred W]

I assume those are factory HID's you are talking about. I am sure that all these things are taken into account when they are designing the systems. Aftermarket HID's however dont have the luxury of knowing the exact layout of how they are installed.

Yes, factory. My point was that they come on with the key before the engine starts and stay lit just fine through the engine cranking. I'd bet that if you rewired the FJR headlights to come on before pressing the starter button, they would likely stay lit too, but the starter motor might run slow due to the extra load on the battery.

 

[Fred W]

I am resurrecting this old thread because I find myself in the same situation that Knucklehead worked his way through previously with my aftermarket HID headlights. On my recent trip to Sag Harbor, I was informed by one of the other guys I was riding with that on 2 occasions one of my headlights had not fired up. Since my main desire for HIDs is to increase my conspicuousness in the daytime, having an intermittent headlight firing is not ideal.

In looking through the 1st Gen wiring diagram I see that that the main Headlight Relay #1 (#54 in the below diagram) is fed by +12V at ignition switch on. However, the ground side of the relay coil is a yellow/black (Y/B) wire that runs over to the ECU. The ECU must provide the current sink to actuate that relay after the engine is deemed to have started, in order to reduce the total load on the battery during starting. It appears that a switch or time delay relay wired to interrupt that low current signal would be just what is needed here.

I will be looking around for some sort of time delay relay device to stick in line with that circuit and just use the existing headlight relays to handle the current. After I find something suitable I will post up the results.

 

[KnuckleHead]

I did order this time delay relay from amazon which I think will fit the bill.

That looks like it should work. I'd advise you to try and waterproof it the best you can.

Yeah, thanks. I think I will run it under the seat if I can.

I will report back with the results.

You mean like the above?

BTW, I have had it installed for about 4 months and it works perfectly. I have had one time when only one light fired after install and I ride just about every day to work. The fix was to just turn the bike off, restart, and in the 10 seconds before the lights kicked on I rev'ed it up to about 4k. Both fired up.

 

[Fred W]

I did order this time delay relay from amazon which I think will fit the bill.

That looks like it should work. I'd advise you to try and waterproof it the best you can.

Yeah, thanks. I think I will run it under the seat if I can.

I will report back with the results.

You mean like the above?

BTW, I have had it installed for about 4 months and it works perfectly. I have had one time when only one light fired after install and I ride just about every day to work. The fix was to just turn the bike off, restart, and in the 10 seconds before the lights kicked on I rev'ed it up to about 4k. Both fired up.

Hey Knucklehead. Yes, similar to what you found. But I'm looking at possibly designing a little solid state circuit to delay the actuation of the OE headlight relay #1. I should be able to make something up on a small breadboard and then pot it in epoxy or RTV to make it completely weatherproof. Following the lead of Brodie and his relay mod, it should be plug and play, and be 100% reversible.

You mentioned 10 seconds, is that the delay period that you have found to work the best for you so far? Did you mess around with different delay intervals? Once I figure out the optimum time I'll probably make my circuit with fixed component values (non-adjustable delay) for ruggedness.

 

[KnuckleHead]

Ohhhhh. Ok, I get it now.

I sealed mine up with silicon after setting the timer. As I mentioned, it has been working great, but with the one "incident" I think I might turn it up to 15 or so if I ever get bored and need something to do.

 

[Fred W]

M'kay.

Previously in this thread discussion, it has been found by several folks who have upgraded our main headlights to H4 HIDs (both cheep chinese and otherwise), that occasionally one of the headlamps will not fire up when we start the bike.

The theory behind this phenomenon is that the HID ballasts require the full battery voltage during their high current firing of the Xenons, and then after firing the current demands settle down. Since the FJR headlights do not switch on until immediately after the bike starts, and because the battery voltage is at an all time low at that exact moment in time after cranking the starter motor, sometimes one (or maybe more?) don't fire. Delaying the headlights from turning on for a few seconds should allow the HIDs to fire up more reliably with no detrimental side effects.

Knucklehead bought and installed a time delay to accomplish this and reports excellent results. But we don't really need yet another mechanical relay in the headlight circuit. I mean, heck... we already have two relays controlling the headlights. The main headlight relay (relay #1) is the On/Off and Headlight relay #2 is the High/Low relay. So my idea is: Why not just delay the turning on of relay #1?

To that end, I have designed and built a little RC time delayed transistor switch that can be easily installed in series with headlight relay #1. The way the headlight relay is designed, +12v is applied to one side of the headlight relay coil at key switch on. Then the ECU supplies a ground on the other side of the relay when it determines that the engine has started. Here is a schematic of the delay circuit that I have just built:

The components I chose (especially the transistors) are just what I had laying around from other projects. They are not particularly good (or the cheapest) choices. Pretty much any low power and medium power switching transistor would fill the bill for the darlington pair. I used the zener diode to raise the turn-on voltage in the RC bias circuit to something reasonable for easier timing. The 47k ohm resistor and 220uF capacitor are the bias components that determine the actual time delay. I estimated these values should result in about a 10 second time delay from engine start to headlights firing, which should be enough time for the alternator to get up to full voltage and the battery to at least partly recover. The 150k ohm resistor is just a discharge loop for the capacitor on power off.

I built the circuit onto a small piece of breadboard from Radio Shack and attached some short wires for splicing into the bike's circuitry. In the original design there were 3 wires. In the simplified there are only 2 wires. One wire gets 12V switched power from one side of the relay and the other goes to the ECU. The idea is to break the yellow/black wire that currently goes from the ECU to Headlight relay #1 close to the ECU (under the seat) and install the circuit board with a molex type connector on the wires. That way if this circuit were to fail you can just remove the circuit board from the connector and install a jumper plug to restore normal operation.

I'll update this thread with results once I get a chance to fully test the board. After which I will either be be potting or hermetically sealing the circuit board with conformal coating for weather resistance.

[edited to correct drawing]

 

[Fred W]

Update:

I finally had a chance to test my little delay circuit board on my bike and Shazam!!! It works exactly as intended. I was even able to wire it in without cutting into the bike's wiring harness.

I just need to spray the board with conformal coating and then pot it in something to protect it from the elements.

I'll be locating the board under the seat somewhere and connecting it to the bike's wiring via a molex type connector for easy disconnect. Then I'll make an "emergency jumper" to stick in place of the board if it were to fail.

I'll try and get some pictures of how it all wires in and maybe a short video of the delay circuit in action.

 

[Fred W]

OK, this should be my final update on this. I have completed the install of my headlight relay delay circuit.

To recap: when you install a set of HID H4 headlights, because the stock FJR ECU turns the headlights on immediately after the engine starts up, and the starting process may drain down the battery a bit, some folks have noticed that their Xenon capsules are not 100% reliably firing up. The idea here is to wait for a while after the engine starts to allow the alternator to ramp up a little before attempting to turn on the lights.

I designed a little time delay circuit which can be put in series between the ECU and the #1 headlight relay to accomplish that goal. In my prototype circuit I required 3 wires: a +12V (switched) signal supply, the signal from the headlight relay and the control signal from the ECU. After determining that I'd really like to install this circuit under the seat (near the ECU) rather than hidden inaccessibly up in the nose cowling, I reworked the circuit to a 2-wire input circuit such that the Vcc is common to the signal from the relay. This definitely simplifies installation. (I have modified the schematic above to reflect this change).

In my first board (picture above), getting the 3 external signals made the board pretty busy. After redesigning the board for just the two inputs, it simplified the board greatly. Also, in the first iteration I attempted to seal the components with an acrylic conformal coating and then pot them in a silicone compound. It was messy, and a heavy little brick.

Iteration 2, after the circuit changes, I merely coated the board well in acrylic conformal coating and dropped it into a small project box bought at Rat shack. Here you see the simplified board components, in the small project box. The milky looking coating of the board is the acrylic conformal coating before fully cured.

I placed this circuit in series with the headlight relay ground signal coming the ECU, which is the Yellow w/ black tracer wire in the ECU connector. Rather than cutting into the precious FJR wiring harness, I just unlocked the ECU connector pins, as is detailed in the Barbarian jumper mod instructions, and pushed the yellow/ black wire out of the connector. I then stripped a section of wire and shoved as many strands as I could into the small orifice of the female ECU pin, wrapping the remaining strands around the exterior of the pin. After installing shrink-tube around this connection it looks to be a fairly sturdy connection. And reversible. (no damage was done to this FJR for this mod).

After adding some molex connectors to the wires and sealing up the project box with silicone sealing compound I think this thing should be able to brave the elements. I tucked the sealed box in under the seat (along with my AVCC) so if it should fail I will easily be able to replace the circuit board with a molex jumper I have already made up and hanging from the box.

Sorry that this is not a great step by step procedure. I'm not sure how many folks will get any value out of it, so am hesitant to invest the time to do so. Maybe if a bunch more people have a need I'll collect all the info and pictures in a single document for posterity.

 

[DrFaulken]

Even though I am not well-versed in electrical things, I like the ingenuity put forth in solving this problem.

I had my first prolonged ride on the FJR tonight in the dark, and boy do the headlights suck. I thought about upgrading the bulbs to Silverstars or equivalent, but then got to thinking about HiD.

In your opinion, would the problem you described not happen / be less likely to happen if just one light was ignited at startup? What if the right side bulb was activated via a physical switch after the bike had a chance to charge itself up?

 

[RadioHowie]

.....and boy do the headlights suck...

Really?

When I first rode my FJR, I thought the exact same thing until I was riding one night through a forested area and realized the high beams were aimed from the factory for coon hunting.

After re-aiming, I've come to feel the FJR lamps are the finest non-HID units I've seen in 32 years of riding. And that's a LOT of different bikes!

Different strokes for different folks, I guess.

 

[DrFaulken]

.....and boy do the headlights suck...

Really?

When I first rode my FJR, I thought the exact same thing until I was riding one night through a forested area and realized the high beams were aimed from the factory for coon hunting.

After re-aiming, I've come to feel the FJR lamps are the finest non-HID units I've seen in 32 years of riding. And that's a LOT of different bikes!

Different strokes for different folks, I guess.

Yeah, they were so bad during my really really short ride at dusk Wednesday I had the dealer check them today during my 600 mile service. I thought for sure the lights were not aimed correctly, but apparently they are in spec.

The light is also yellow and the depth of field is short in comparison to my last three bikes. I'd put it somewhere in between my 1980 Honda CB400T and my 2002 BMW K1200LTE (before I did an HiD upgrade). I will say the width of the pattern is nice; we get a ton of deer out where I live.

However, you give me hope. I might put some Silverstars in there and re-level the lights myself.

 

[Fred W]

Even though I am not well-versed in electrical things, I like the ingenuity put forth in solving this problem.
I had my first prolonged ride on the FJR tonight in the dark, and boy do the headlights suck. I thought about upgrading the bulbs to Silverstars or equivalent, but then got to thinking about HiD.

In your opinion, would the problem you described not happen / be less likely to happen if just one light was ignited at startup? What if the right side bulb was activated via a physical switch after the bike had a chance to charge itself up?

No, I think you would have the same intermittent problem firing the single headlight too. It appears to depend on how much the battery voltage gets sucked down by the starter.

Yes, you could simply splice a manual switch of some sort into the same yellow/black wire from the ECU to Headlight relay #1. You would then have manually operated headlights and would have to remember to turn them on. Of course they would shut off and on as normal with the keyswitch if the manual switch was left "on".

On the subject of how good or bad the FJR headlamps are, I think they are pretty good actually for a stock motorcycle headlight, which isn't saying much. But my primary reason for wanting HIDs is for conspicuousness during the day time, at which they seem to excel. The low beams are considerably better at night and the high beams are brighter though they still have the piss pooor beam pattern of the factory headlamp bucket design.

My HID install experiences

 

[KnuckleHead]

In your opinion, would the problem you described not happen / be less likely to happen if just one light was ignited at startup? What if the right side bulb was activated via a physical switch after the bike had a chance to charge itself up?

Keep in mind that all the HID kits I have dealt with (3 different brands) all plug into one stock headlight plug. The other stock headlight plug is left hanging. If you really wanted just one light, you would have to redesign the harness (or I guess you could buy two single light kits and run two single harnesses), but that seems like a lot of work when you could either make this relay or, if you are more eclectically challenged (like I am) buy the aftermarket timed relay that I linked above. I have now been using it for nearly 8 months and have had only one instance when one light fired. I ride 4 or 5 days a week to work.

 

[buf2phx]

Not sure where you guys are getting your HIDs, but I've had good luck with www.xenonlink.com. Had a set in my Silverado, and currenly have one in my Seca II. They are completely plug-n-play. No relays or direct battery connections. I haven't burned the bike down yet, it has been about a year&10k miles since install. YMMV.

Mike

 

[worldbound4now]

Not sure where you guys are getting your HIDs, but I've had good luck with www.xenonlink.com. Had a set in my Silverado, and currenly have one in my Seca II. They are completely plug-n-play. No relays or direct battery connections. I haven't burned the bike down yet, it has been about a year&10k miles since install. YMMV.Mike

Scary.... really, no relay? So your HID's draw power through your original wiring harness and original hi/lo beam switch? I'd be interested to see the draw specs for your HID kit vs the gauge wire and amp rating for the switches/relays involved in the stock setup. I like fire, but like to keep it in a pit when camping, not in front of my tank of fuel.

I'm happy to see such a common sense fix to the issue. Great idea with the time delay relay.

 

[buf2phx]

Not sure where you guys are getting your HIDs, but I've had good luck with www.xenonlink.com. Had a set in my Silverado, and currenly have one in my Seca II. They are completely plug-n-play. No relays or direct battery connections. I haven't burned the bike down yet, it has been about a year&10k miles since install. YMMV.Mike

Scary.... really, no relay? So your HID's draw power through your original wiring harness and original hi/lo beam switch? I'd be interested to see the draw specs for your HID kit vs the gauge wire and amp rating for the switches/relays involved in the stock setup. I like fire, but like to keep it in a pit when camping, not in front of my tank of fuel.

I'm happy to see such a common sense fix to the issue. Great idea with the time delay relay.

I don't have the draw specs, but HIDs draw fewer amps than conventional bulbs. That's one of their selling points. The site I referenced states 35% less power consumption. fjrtech.com claims the HIDs draw 36 fewer watts. I'm usually all about over wiring stuff, but I don't see the need here. Again, YMMV.

Just for S&G I looked at some automotive schematics. Not that GM is the ultimate authority, but on the 07 Escalade they power the HIDs with 18 gauge wire from the fuse block to the ballast. I have yet to see one of those catch on fire, and trust me, I see a LOT of Escalades.

Mike

 

[Fred W]

As mentioned above, the HID headlights do draw (a little) less current than halogen lights when running, so the factory wiring harness should be structurally adequate to carry that load. However, during initial firing of the Xenon capsules the inrush current required by the ballast(s) is much higher than the "run current". This thread is about getting the headlights to fire more reliably on kits that are already wired directly to the battery. Adding any additional resistance in series with the lights would be a bad idea. For that reason, having the lights fed directly from the battery via the relay (which is generally included in any upgrade kits) is a good thing.

Also, in the FJR application in particular, the headlight current actually goes through the ignition switch contacts before going to the headlight relay#1 and #2. :blink:

By transferring the headlight load current out of the switch, to a dedicated relay, there is a good likelihood that the life of the ignition switch will be extended, similarly to how a Brodie Relay fix works.

 

[JamesK]

Fred, thanks for designing the circuit, it certainly looks good. It looks like the wiring diagram should be sufficient to enable someone to replicate the design with components from a local Radio Shack. I will likely follow your design so it would be great if you could provide any other pertinent info you might have.

For interest purposes only I wanted to post a link to a configurable time delay relay I found, which whilst being on the spendy side ($53). This series comes in Amp ratings of 10A, 25A & 50A. It's a solid-state relay with Delay Times: 1.6 Sec ~ 133 Sec