Headlight not working; voltage at terminals, but seems to drop to zero with a bulb

[xwhatsit]

Headlight stopped working last night. How many times has Motu et al told us `NEVER ASSUME' -- yet, I found myself with half the wiring loom for the bike apart before I realised that despite looking perfect, a fuse had succumbed to vibration and the top cap was loose and not touching the fuse itself.

So, replaced the fuse, lights work with a bulb plugged into the socket, yay. Took the opportunity to put my mint SR400 rim/glass on (instead of the munted Bandit 400 item that was there -- there's a tip guys, SR400, maybe SR500 too, the headlight shell and rim are interchangeable with the Bandit 400). Noticed with the headlight rim on, headlight was a lot dimmer -- must be a short somewhere, something touching.

The wiring in the headlight shell is a bloody mess. Bits patched here and there, exposed connectors. A good portion of the responsibility is mine, but previous owners have had a hand in it as well.

I fixed connectors, broke a wire (directly going to the left indicator -- need to sort that). Now no headlight at all.

Interestingly, I'm getting a healthy 14 volts at the headlight terminals themselves. But when I put a bulb there -- nothing. Thinking that the connectors might be duff, I plugged in my multimeter (which is getting 14 volts quite nicely), then touched a test bulb to the two prongs of the multimeter. Nothing comes out from the bulb, and the multimeter goes down to zero volts. Is this expected? I suppose I'm wiring the bulb in parallel with the multimeter.

Resistance between the low beam terminal and the earth is not excessively high when I get a good connection on the multimeter.

But the thing I'm confused about most, is that if I have 14 volts at the terminals, why doesn't the bulb light? I've tried three different bulbs, which were all working about 10 minutes ago and conduct electricity between their terminals (I would think that means they're not blown). EDIT: and yes, the bulbs work fine using the momentary passing switch for high-beam.

I've got half an hour to fix this or I'm on the bus :laugh:

So it's all very frustrating.

[MSTRS]

Sounds to me like the problem is not the connectors etc...but is in the handlebar switch? If it is on the 'dead' side then it is affecting the flow to earth.

[xwhatsit]

I think earth is fine -- the bulb has earth all the time, and then either the normal switch or the passing switch gives it positive voltage. As the passing switch lights it up OK, I reckon earth is fine.

I've noticed that jiggling wires around also stops the parking bulb occsaionally, so I may just hop on the bus in a tick then replace the whole wiring loom tonight.

[MSTRS]

Mr Honda did strange things with wiring over the years. And yours being an 'older' example of his dark art, do not be surprised if the pass flasher is on a different circuit.

[MSTRS]

Just thinking about this...if the power wire has voltage at the bulb end, there must be two power-in wires (high and low beam) so the earth is on the far side of the circuit from the h-bar switch...
This suggests that it is the earth side that is the problem. Can you not just jerry-rig another earth off the bulb socket?

[xwhatsit]

Mr Honda did strange things with wiring over the years. And yours being an 'older' example of his dark art, do not be surprised if the pass flasher is on a different circuit.
It is, yep. Pass flasher takes voltage directly from the reg/rec and feeds it into the highbeam terminal. The other terminals come from the left-hand switchblock (high/low switch), then they go back to fuse box (which was causing me initial trouble), then go into the right-hand switchblock (lights on/park/off).

The problem is, where is the fault? As there is next to no resistance between the earth and low/highbeam terminals with the bulb out and the bike off -- meaning the connection is good -- and with the bike running and the lights switched on (but bulb out), I'm getting between 13.8 and 14.4 volts between earth and low or highbeam terminals. Switching the lights on/off, highbeam/lowbeam produces the appropriate voltages at the terminals (either 0 volts or 14 volts).

However, plug a bulb in, and despite there being voltage there, there's no light. I know the connections and bulb is good because it works with the passing switch.

[xwhatsit]

Just thinking about this...if the power wire has voltage at the bulb end, there must be two power-in wires (high and low beam) so the earth is on the far side of the circuit from the h-bar switch...
This suggests that it is the earth side that is the problem. Can you not just jerry-rig another earth off the bulb socket?
As far as my small brain can comprehend, I don't think that's quite right. I'm pretty sure the earth is fine, as the bulb works with the passing switch. The earth at the bulb socket doesn't switch at all, it's always connected, so passing switch and normal headlight modes use the same earth. The difference is in the source of positive voltage.

If one has voltage at a socket but a bulb will not light, does this mean somewhere along this twisted snaking collection of wires and sockets something cannot handle a load -- in my imagination, a loose bit of wire gets hot, curls up, breaking the connection -- take the load away, wire cools back down and touches again. Obviously my multimeter provides next to no load.

[MSTRS]

...The earth at the bulb socket doesn't switch at all, it's always connected, so passing switch and normal headlight modes use the same earth...
If I was thinking 'fast' enough, I'd have got to that conclusion as well.

For what it's worth, I've about exhausted what I know about electrickery...
I do know that voltage is only the carrier wave for other stuff. What has happened to the Watts? Or Amps?

[CookMySock]

Interestingly, I'm getting a healthy 14 volts at the headlight terminals themselves. But when I put a bulb there -- nothing. Thinking that the connectors might be duff, I plugged in my multimeter (which is getting 14 volts quite nicely), then touched a test bulb to the two prongs of the multimeter. Nothing comes out from the bulb, and the multimeter goes down to zero volts. Is this expected? I suppose I'm wiring the bulb in parallel with the multimeter.

But the thing I'm confused about most, is that if I have 14 volts at the terminals, why doesn't the bulb light?It does this because there is a high resistance in the + line somewhere. The multimeter is high impedance test device, and the test bulb is a low impedance device, so they give conflicting results.

Nevertheless, it does indicate that the wiring is solid, except for perhaps a corroded connector or fuse. Reconnect the headlamp, and follow the wiring with the tester or the multimeter until you find 12V, and theres your fault.

Another way, is to remove the battery completely, and ground the positive and negative terminals to the bikes frame, and then follow the wiring with the Ohm meter.

Good luck. Its all time and trial and error. Once you have tried the possible, begin on the impossible, and you will find something you hadn't thought possible.

Steve

[Katman]

For what it's worth, I've about exhausted what I know about electrickery...


Yes, we noticed.

[MSTRS]

Yes, we noticed.

Thanks for pointing out what I've already made public. :doh:
What I noticed, and feel honour-bound to point out now, is that you made no attempt to be helpful to the OP. :confused:

[MSTRS]

I've had a chat with an auto sparky mate, and he tells me that it will most likely be an earth problem in the high/low switch block. The on/off switch is sending power, but the hi/lo switch controls what happens to it. And yes, the flasher switch is on a different circuit altogether, so all that proves is that your bulb is ok.
He also says that using the volt meter is not going to be helpful, you need to be using a probe light to test the circuit.

[xwhatsit]

I've had a chat with an auto sparky mate, and he tells me that it will most likely be an earth problem in the high/low switch block. The on/off switch is sending power, but the hi/lo switch controls what happens to it. And yes, the flasher switch is on a different circuit altogether, so all that proves is that your bulb is ok.
He also says that using the volt meter is not going to be helpful, you need to be using a probe light to test the circuit.
Thanks for having a chat with your friend. I find this interesting -- I'm next to useless at all of this electrickery magic, but I don't think the switchblocks actually have an earth, I think they're just simple switches that either let voltage through or not. It's all positive voltage going through them, the high/low switch just chooses which wire to send it down, and the on/off switch stops it from getting to the high/low switch in the first place. But then again my grasp of all of this is tenuous at best.

Anyway, after catching the bus tonight, I've decided I'm not doing that again in a hurry! That's a whole week's petrol money blown on one trip to work. So I'm going to rig up a temporary wire directly from the main positive line to the low-beam terminal of the bulb and deal with that until Wednesday, when I can skip lectures and have the night off work. Then I think I'll rig up a test lamp, wire one end to the cylinder head fins and then probe around with the other end down the connector blocks until I find where the fault is lurking.

Spoke to girlfriend's father, who is an electrical engineer (but sometimes fiddles with house electrics as well, even though he's way more qualified than that), and through a ghastly manglage of Cantonese and English he thinks that I might be onto it; the multimeter is too light a load for this bad connection to surface. Suggested the same thing as your auto sparky mate, that I need a probe light or something, preferrably a wattage close to that of my headlight bulb.

Fun, fun, fun!

[xwhatsit]

Well I've run a wire from the magic spare positive terminal in the wiring loom (it's on the wiring diagram too -- black wire, terminates in a little round female connector -- is it there on purpose for emergencies? If so, thanks Honda!) to the lowbeam terminal and we have lights.

The dimming lights when connecting the headlight glass/rim to the shell I mentioned before seems to be a problem with the SR400 glass/rim. It would seem the SR400 rim conducts electricity to somewhere which causes a short and the lights to go dim. Original Bandit 400 item does not exhibit same behaviour, problem solved. Bandit 400 item also has much larger rubber shroud (can't get the parking light in, however) which may help, or may not.

Went for a wee pootle down the end of the farm and back. Quite nice riding around in the middle of the night wearing naught but a helmet, dressing gown and slippers. I added the boxers as a last minute afterthought, after all, there is the fuel tank and the wind which could cause issues for onlookers.

I've packed a fairly comprehensive toolkit for tomorrow just in case, so if anybody sees me around varsity or Parnell with oily hands and a bike spewing spaghetti wiring out of it's blank eye, feel free to come over and yell abuse.

[CookMySock]

Good work bud. You will be a pro at this before long.

If you suspect some part of an electrical circuit is not conducting properly, use your voltmeter to measure across the suspect part while it is under load, and it will show a few volts if there is an resistance there.

Steve

[notme]

Are you able to measure the resistance in ohms of the low beam (troublesome) wire, by connecting one multimeter probe to the bulb end, and then the other probe to various points along the low beam wire at connectors as far back towards the battery as you can?

Or did you find the problemo?

[xwhatsit]

Are you able to measure the resistance in ohms of the low beam (troublesome) wire, by connecting one multimeter probe to the bulb end, and then the other probe to various points along the low beam wire at connectors as far back towards the battery as you can?

Or did you find the problemo?
Not yet -- I have tomorrow free so I'm going to get stuck into it then. I thought about measuring the resistance just between the terminals on the bulb socket, but it's negligible (less than 1Kohm from memory, whatever it is it's on par with the parking light bulb).

[pete376403]

If this is a double filament bulb (ie hi & lo beams) then you're checking between one of the bulb contacts and earth, right? There shouldn't be any voltage between both of the contacts (unless the dip switch is stuffed)

[notme]

Not yet -- I have tomorrow free so I'm going to get stuck into it then. I thought about measuring the resistance just between the terminals on the bulb socket, but it's negligible (less than 1Kohm from memory, whatever it is it's on par with the parking light bulb).

It should be a simple test to check the obvious - does the low beam wire have significant resistance in it, which would allow voltage at the terminals that would then collapse with any significant load (i.e. multimeter is not significant, bulb is).

If you find more than say 1 Ohm along the low beam wire then you have a corroded or broken connector or wire somewhere along the way, so 'd suggest you leave one probe at the buld connection to low beam, and go along the wire as far as possible until you start reading significant resistance. If that point is after say some connector block, tidy up the connections or check for corrosion etc, or if it is after say the headlight switch, open her up and check the state of the contacts.

HTH

[CookMySock]

If you find more than say 1 Ohm along the low beam wire then you have a corroded or broken connector or wire somewhere along the way [...] This is true, but very hard to measure down to the one ohm area without a quality test instrument. It is much easier and quicker to energise the circuit and prod along it with a test lamp until you find a discrepancy. The eye can measure varying testlamp brightness quite well.

HTH
Steve

[xwhatsit]

If this is a double filament bulb (ie hi & lo beams) then you're checking between one of the bulb contacts and earth, right? There shouldn't be any voltage between both of the contacts (unless the dip switch is stuffed)
Yup! Haven't quite lost my marbles yet! :laugh:

This is true, but very hard to measure down to the one ohm area without a quality test instrument. It is much easier and quicker to energise the circuit and prod along it with a test lamp until you find a discrepancy. The eye can measure varying testlamp brightness quite well.

HTH
Steve
Yeah, that's one reason why I didn't bother checking under 1Kohm, the multimeter I'm using is a $10 piece of crap from DSE; hell, even touching the outside of female sockets is enough to change the resistance value markedly (the contacts within are quite clean and conduct well, but the blunt face of the connectors often have a little bit of corrosion).

[notme]

Yup! Haven't quite lost my marbles yet! :laugh:

Yeah, that's one reason why I didn't bother checking under 1Kohm, the multimeter I'm using is a $10 piece of crap from DSE; hell, even touching the outside of female sockets is enough to change the resistance value markedly (the contacts within are quite clean and conduct well, but the blunt face of the connectors often have a little bit of corrosion).

Seriously? Even a $10 POS meter should be able to measure to 1 or 2 ohms...out of interest, connect the probes and measure the resistance of the probes and wires by shorting them together, the reading should be < 1 ohm and shouldn't jump about much.....then try a lightbulb - should be 10's of ohms. If the meter can't read the difference between these 2 things, chuck it, because if it really is that bad it will be useless for checking other things - e.g. your charging system (was that 12v or 2V?).

Back to the original problem....if you must use the test light, try the same trick as I suggested with the meter, i.e check the length of wire that feeds the low beam, but clip one side of the test light to the frame or battery negaive, and then probe back along the low beam wire at each connector or switch or relay or whatever until the light comes on. Where it lights is the last point where you have a good power feed to the low beam wiring, so if for e.g. it lights on one side of some connector but not the other, that connector is corroded or loose or broken.

[Ixion]

I got a couple of them DSE meters. Checked them against a good(ish) one (Fluke) they were reasonably accurate. measure down to 1 ohm OKish. Course, at that level contact is important, and corrosion on exposed wire ends and stuff. But I think the meters are OK. I didn't check high voltages, that is where I would expect to see inaccuracies.

[davereid]

Most cheap meters are perfectly adequate for fixing motorcycles. They will have a degree of accuracy at least 10x what you need.

As commented by others in the thread, the best tool for diagnosing a problem with lights and switchgear is a $10 test bulb.

Apply the rule of halves, and the rule of movement to your testing.

1) Find a good earth for the probe, and get a visual idea of the test lamps brightness on battery POS.

2) Go to the FAR end of the circuit under test - in this case your headlight. Turn the headlight on, and with your test lamp, probe each of the wires in turn.. HIGH / DIP / EARTH chaging the DIP switch as required.

- Bright test lamp = circuit OK
except if any glow on test lamp at all on EARTH - this means a faulty earth.
Dim test lamp ? Hmm.. are you measuring the right wire ? not looking at a feed via a lamp with a dead earth ? Are any of the lamps under test glowing weakly ? - dodgy earth. Could be a high-resistance wire or switch.. easy to find with...

3) Rule of halves/rule of movement.

Find the middle of the circuit, more or less. Exactly the same test as before, with the same answers. But you have now halved the amount of circuit to test.

Something dodgy ? rule of movement - switches fail first, then wires that have to move like steering head, then connectors, then (almost never) wires that don't move.

G'Luck !

[xwhatsit]

Well, Dave, looks like it was your `almost never' failing wires-that-don't-move wot failed. I think. I dunno. It's working now (for now?), so I won't touch it anymore!

This morning fiddled with some connectors, crimped new ones on, lights were working. Woot. Go to uni and back in the afternoon, on the way home, lights fail again. Shit.

I made a little test light (old indicator warning lamp, a metre or so of wire, and two fancy crimped-on connectors), earthed it to the same earth as the headlight, then traced it back. I finally got a light when I connected to the back of the bike at the fuse box, headlight fuse. So that means the wire between the fuse box and the high/low beam switch must be faulty, right? Probed and prodded and measured until I was 100% that was it.

So I spend half an hour separating that same wire from another loom, then installing it in place. Ran it up beside the existing wiring loom, used bits of drain tubing from old airboxes to shield the fragile little wire. Tested it -- lights work! Yay!

Go to ride down the road and back, and it's not working again. Just about in tears :laugh: I've spent how many hours doing this and I thought it was sorted?

Do all the tests -- same wire. How can it be?

I realise that I was missing out one little wire in the middle. It's about 3cm long, simply goes from the fuse to the fusebox connector. Never moves, looks sturdy, fine. Swap fusebox with one from my spares collection, works first go. Stood up to 10km down the road and back. I hope that was it?

WTF?

I'm too tired to examine it right now, but I wonder what on earth could've failed in the fuse box. It seems to be an intermittent thing. Who knows.

I've still got that patched-on replacement wire in the loom. I'll leave that there because it's working now and I don't want to upset it :lol:

Thanks everybody for your help. I've bookmarked this thread for future reference.

[gijoe1313]

Huzzah! :clap: You have made the lekky gremlins run out of places to hide, so they had to jump ship .. probably to your cellphone or laptop now ... :sweatdrop

You needed piccies! :innocent:

[howdamnhard]

One thing I've learned with electricity is that even though the multimeter says the voltage is great doesn't mean if you try draw current it will work.Sounds like a bad switch contact.Eventhough voltage is great as soon as you apply a load the resistance is too great and result no light.Had it happened on my car headlights once,np brights.Turned out the metal contact on the switch needed tweaking to insure a better contact .Anyway thats my 2 cents,goodluck.:Punk:

[Wired1]

Well I'm a bit late on this topic and it looks like you've found the solution, but in case it recurs here are some things that may help with the fault diagnosis.
Firstly, although bikes use a (generally) negative earth this is only used for the ignition and charging circuits. The horn on modern bikes is permanently supplied with positive and the horn switch grounds it to make it honk and the the neutral light is similarly connected to the positive and the neutral switch grounds it to make it glow but the lights all have both positive and negative wires run to them - ie lights do not rely on the frame earth to make them glow.
Secondly, if you can measure a no-load voltage with your meter of 12 or more volts at the headlight, and zero or bugger all volts with the light switched on then there is a high resistance on one side or the other of the headlight. The easiest way to find out which side is to run a lead from the battery negative to the negative side of the headlight (probably the black wire) and see if it goes. If it doesn't, repeat with the fly-lead running from the positive side of the battery to one or other of the other two pin on the headlight. If it is on the negative side then that's the easier one to find as it will be a loose or worn connector, poor fitting fuse or loose battery terminal.
I would put money on a high resistance in the actual switch if the bike is 20 years old or more, they get filled up with crap and the contacts lose their effectiveness.
You also mention the possibility of a fault in the headlight area when you put the headlight unit back in - assuming the bike hasn't been in a fire (been there done that) this might be something as simple as the insulation slipping back from a connector. Visually check them and also look for cracks in the cable insulation.

[xwhatsit]

The funny thing was, is that there was really low resistance; multimeter showed anything between 0 and 5 ohms, provided I got a good surface connection with the probes.

I sorted the headlight rim thing; turned out to be the metal barrel of the ignition (which I've mounted sticking out the back of the headlight) was touching the shell; the SR400 rim to Bandit 400 rim has conductivity or something, which the Bandit 400 rim to Bandit 400 shell does not.

[classic zed]

Its a fairly common problem on cars too, you will have a bad connection somewhere on the positive line, if both dip and main beam are affected the problem is either in the switch or before the switch. When you test the circuit "unloaded" and get a reading everything appears to work, as soon as you add a load i.e. a bulb the bad connection cannot flow enough current and so the voltage drops off.

Everyone must have met this problem.
Think of your car battery with a loose terminal, when you turn the key the dash lights come on, then you try and start it and you hear a "click" and everything dies. Wiggle the battery terminal and everytrhing works again. Its the same thing as any wiring connector :done: