signal lamp to indicate headlight failure

[jsob]

I've installed an LED headlight on my motorcycle, and after three months I'm pleased with performance. Unfortunately, reviews of the product are mixed and enough folks have reported premature failure of the light to warrant concern about durability. So, wondering if there were some way of wiring the lamp so that in the event of failure, a warning light would be powered. The headlight draws 30/40 watts on a 12 volt line. I don't want to add significantly to the light's draw as it is a 40 year old bike with limited power generation. Thank you.

[jogri]

Why do you need a light that tells you if the big light illuminating the road in front of you has stopped working?

[jsob]

Most of my riding is during the day. The bright LED gives me greatly (and probably, annoyingly) increased visibility. I don't want to be the victim of someone's "but I didn't see him"...

[SparkyFX]

You could sense the voltage drop in the supply wiring, compare it "on" to "off" or have a photo sensor in the vicinity of the lamp. All methods require to take ambient conditions into consideration.
The headlight itself probably is PWM controlled, so most ways of sensing need to compensate this.

So you should first measure what device you got, what its nominal conditions are, wire resistance between two points and so on.

[calzap]

Forget electronics for this and think optics.  A skinny flexible light pipe will do the job.  Cheap, quick and unobtrusive if done right.

Mike in California

[Manul]

If you want simple oldschool electrical solution, you may try to use reed switch, it should work ok. If your headlight is 30W at 12V, that would be around 2.5A, you may check the current first to be sure. If reed switch is for eg. 20AT (ampere turns), you wind around 10-15 turns of wire to make it switch when the current flows. You may choose SPDT type and use normally closed contact pair to power some indicator LED. If current disapears, indicator will light. Try to keep coil resistance low, do not use very thin wire.

[jsob]

Manul, thank you. I like simple. In your diagram, is line C-D in series with my headlight lead? The headlight wire is 14 or 16 gauge wire (I'll have to verify). The reed switches I've seen are quite small, so I'm not sure how many turns of insulated wire I'll be able to wrap around a switch. Can I over-lap the wiring?

[JustMeHere]

Just look for the reflection off of other vehicles when you're stopped at a traffic light.

You probably need to sense the current and not the voltage for changes. 

[Vovk_Z]

Most of my riding is during the day. The bright LED gives me greatly (and probably, annoyingly) increased visibility. I don't want to be the victim of someone's "but I didn't see him"...

It wan't annoy at day (but may do it at night) - the Sun is too bright.

I use my car headlight during the day too (so in fact - al the time, day and night), it really helps.

[Vovk_Z]

Can I over-lap the wiring?

Yes. But it will have less effect (with thick PVC insulation).
Usually we use thin isolater wires here - laquer-isolated etc. But of cause you can try PVC-isolated wire.

[Manul]

Manul, thank you. I like simple. In your diagram, is line C-D in series with my headlight lead? The headlight wire is 14 or 16 gauge wire (I'll have to verify). The reed switches I've seen are quite small, so I'm not sure how many turns of insulated wire I'll be able to wrap around a switch. Can I over-lap the wiring?

Yes, C-D is in series with headlight. It is simple but still needs some work and patience. You need to find SPDT reed switch which is sensitive enough (low AT numbers), find some enameled wire around 22-24 AWG (for eg. from some old transformer).  Better do one layer, nice and tidy. After winding a coil, attaching wires to contacts and verifying that it works, you may put some epoxy resisn to make it robust. Reed swiches are quite fragile. Strongly recommended to check headlight current first, because I would not trust LED wattages. Then choose amount of windings so reed switch is reliably switching at 1/2 or 2/3 of that current. And you need some bright LED and droping resistor for indicator.

[Brumby]

Manul, thank you. I like simple. In your diagram, is line C-D in series with my headlight lead? The headlight wire is 14 or 16 gauge wire (I'll have to verify). The reed switches I've seen are quite small, so I'm not sure how many turns of insulated wire I'll be able to wrap around a switch. Can I over-lap the wiring?

Yes, C-D is in series with headlight. It is simple but still needs some work and patience. You need to find SPDT reed switch which is sensitive enough (low AT numbers), find some enameled wire around 22-24 AWG (for eg. from some old transformer).  Better do one layer, nice and tidy. After winding a coil, attaching wires to contacts and verifying that it works, you may put some epoxy resisn to make it robust. Reed swiches are quite fragile. Strongly recommended to check headlight current first, because I would not trust LED wattages. Then choose amount of windings so reed switch is reliably switching at 1/2 or 2/3 of that current. And you need some bright LED and droping resistor for indicator.

I like this idea.  ^  ^  ^   

[tkamiya]

Mercedes-Benz automobiles used to have such indicators for all external lights.  An array of small value resisters were placed in series.  When voltage across it drop below pre-determined value, an indicator light in dash came on.

Problem was, aftermarket headlamps often triggered this condition.  Why they didn't simply detect zero volt condition, I have no idea.  Also, it didn't tell the driver which light was out.  So you'd end up checking all.  Often missed were the 3rd stop light on rear.

[james_s]

Mercedes-Benz automobiles used to have such indicators for all external lights.  An array of small value resisters were placed in series.  When voltage across it drop below pre-determined value, an indicator light in dash came on.

Problem was, aftermarket headlamps often triggered this condition.  Why they didn't simply detect zero volt condition, I have no idea.  Also, it didn't tell the driver which light was out.  So you'd end up checking all.  Often missed were the 3rd stop light on rear.

Volvo used this too. The sensors are clever, there are several reed switches, each of which has a pair of coils wound around it with the left and right lamps wired through these coils in opposite directions. When both lights are working the magnetic fields cancel out but if one lamp burns out the field closes the reed switch and illuminates the warning. They are crazy sensitive though, I've had trouble with dirty lamp sockets or slightly mismatched bulbs causing the warning to come on.

[Vovk_Z]

They are crazy sensitive though, I've had trouble with dirty lamp sockets or slightly mismatched bulbs causing the warning to come on.

I have to say that this is rather good. Becouse both dirty socket and different lamps (new and too old) - is bad.
 Dirty socket leads to very hot contact and melted damaged plastic, and old lamp - has lover brightness (so we have always change paired lamps in pairs).

[james_s]

I have to say that this is rather good. Becouse both dirty socket and different lamps (new and too old) - is bad.
 Dirty socket leads to very hot contact and melted damaged plastic, and old lamp - has lover brightness (so we have always change paired lamps in pairs).

It's annoying, often whatever fault is so minor that it causes no visible difference between the sides. More often the car has been plagued by nuisance lamp failure warnings and eventually the owner either ignores the warning lamp or removes the bulb. I fought with it on one car for years and never did fully resolve the bulb failure lamp coming on under certain situations.

[coppice]

Volvo used this too. The sensors are clever, there are several reed switches, each of which has a pair of coils wound around it with the left and right lamps wired through these coils in opposite directions. When both lights are working the magnetic fields cancel out but if one lamp burns out the field closes the reed switch and illuminates the warning. They are crazy sensitive though, I've had trouble with dirty lamp sockets or slightly mismatched bulbs causing the warning to come on.

The 740 had this when it was launched in 1984. When I bought one in 1985, and found it had this feature, I assumed all the new luxury cars would be getting such a feature. It seems they didn't. I never bothered finding out how it worked, but I knew it had to be something electromechanical like you describe. I knew it wouldn't be based on a semiconductor solution at that time.

[3dBi]

Maybe a Audi 80 "relay" box can help you out, VAG part number 443919469B (google it). I'ts a bit old, but newer "relay boxes" have multiple light checkers in one. This one is for pair of 55W halogen H1 headlights, if memory serves. Basically it checks for voltage drop between small loop of PCB trace, to get current reading. If current drops to certain threshold, K output pin goes high and this lights a warning in dashboard board computer.

[Jeff1946]

As others have said a reed switch.  Here is a reference from Mouser  https://www.mouser.com/datasheet/2/262/2118021015e-1633402.pdf.  Note the switch is normally open, so you would need to put it in series with an led and appropriate resistor (1000 ohms)  and  connect in parallel to the lamp.  The spec sheet says max of 15AT to operate the switch.  I assume this is amp-turns.  Here is a note on solenoids.  https://web.pa.msu.edu/courses/2000fall/phy232/lectures/ampereslaw/solenoid.html.

[james_s]

The 740 had this when it was launched in 1984. When I bought one in 1985, and found it had this feature, I assumed all the new luxury cars would be getting such a feature. It seems they didn't. I never bothered finding out how it worked, but I knew it had to be something electromechanical like you describe. I knew it wouldn't be based on a semiconductor solution at that time.

I don't know when it first appeared but I had a '77 240 that had it too so it went back at least as far as that.

[jsob]

Evaluating the Standex Electronics ORT 551/10-15 AT, which has an operating range of 10 ~ 15 AT
I have a question about the number of windings I should use in my electromagnet.
In my 12 volt system, with an LED headlight draw of about 2.69 Amps, the switch is effective across the rated AT range.
How do I decide on the number of turns to use?
Does a greater number of windings produce a more intense (and reliable?) magnetic field?
How do I minimize heat: fewer turns?
I found that 24 gau wire at 15 turns became excessively hot, whereas 22 gau at either 10 or 15 turns produces very little heat.
(I'm using the temperature probe on my retail volt-ohm-milliameter. I don't have confidence in accuracy, so I'm referring to relative temperature readings.)
Thank you.

[Manul]

2.7A current with 10 turns is already ~2x the switching field for this reed switch. That is fine. Of course it depends somewhat on how you wind the turns, position and spacing. Also it is important to note, that if you use normally closed contact pair, such a circuit is more or less "fail proof" because if magnetic field is too small you will get indicator light, same as headlight failure. Anyway, you can use lab power supply to "calibrate" the number of turs for something like half (1.35A) of the current if you really wish. But 10 or 12 looks correct.

I'm a little confused about your wire getting hot, because for eg. 24 gauge, 10cm length copper wire gives me 0.0084 Ohms, and 0.06W of power dissipation at 2.7A, which should not be felt much at all. And for 22 gauge, 10cm, 0.0053 Ohms and 0.04W at 2.7A. There is not very big difference, both wire dissipation is very small. Maybe your 24 gauge is bad, or not 24. Is it both solid core and made from copper?

Any gauge is ok, which fits nicely in one layer for the number of turns required and has small power dissipation (less than 0.1W or so). I'm not sure about optimal spacing, I wind it directly on glass, or use thin electrical tape layer to stop sliping. Hope it helps.

[jsob]

I apologize for not giving complete information about the 24 gau wire. It is copper and I measured its diameter with a micrometer to determine size. I unwound a choke to obtain the wire, so maybe I damaged/shorted the wire while untwisting it? Also, I applied enamel paint to the coil after winding it. So, maybe the paint on the coil inhibited heat dissipation? I did not apply paint to the 22 gau coil. I used my micrometer to obtain the diameter of the switch, then found a finishing nail of slightly larger diameter to wrap the coil around. I plan to mechanically stabilize the coil to keep it in place. Thank you for the guidance. I'll report back at the end of the riding season to let you know how it worked!

[tooki]

Why do you need a light that tells you if the big light illuminating the road in front of you has stopped working?

Given that you’re in a country that mandates daytime running lights for motorcycles, you should know that headlamps aren’t only for driving in the dark.

[coppice]

Why do you need a light that tells you if the big light illuminating the road in front of you has stopped working?

Given that you’re in a country that mandates daytime running lights for motorcycles, you should know that headlamps aren’t only for driving in the dark.

It has value for this purpose, but more generally lamp failures are only obvious at night. If you drive mostly during daylight you can have a failed stop or turn light for a long time before you notice anything is wrong.