Electrical gurus - surging blowing fuse?

[MarkT]

I changed all of my incandescent bulbs to LEDs on Jade.  Did the flasher indicator mod and replaced the flasher relay with a digital one for LEDs.  Added my super Highlighter which is partly powered by the (same) br/wh circuit. The rear markers are now run/brake/turn 1157 LED arrays - the old school ones, not the surface mounts. The rear light is replaced by the Radiantz - super bright LED array. The instrument lights are high-power LEDs with a pot wired in for dimming.  All of the lights this circuit feeds have been replaced by LEDs except the license light. All of the lights are brighter than the OEM incandescent bulbs but should draw less power. (The br/wh circuit powers all the lights except the headlight & flasher filiments.)

On a ride this weekend, after 2 hours of riding my running light circuit blew the meter/tail fuse.  Today busted out the pro multimeter, took the tail circuit wiring apart - everywhere I worked - tested in several places while isolating segments and can find no short. However it seems to be OK if I step the fuse up to 15 from 10. Before cutting wires to isolate them I measured about 3Ω. Installed a 10a fuse and it blew. Started cutting wires to isolate sections, while taking Ω measurements. Looks like the biggest single power draw is the tungsten tail light. Wiggling wires, could find no change.  No chafing. Isolated the rear tail light assembly not to touch the chassis. Reconnected the br/wh (hot side) wires I cut with jumpers. Now it measures 2.5Ω impedence on the circuit (at the fuse), and doesn't blow a 15a fuse. Per Ohms law, that means the circuit draws 5.6 amps at 14v at steady state.  So I'm at a loss to explain why it blows 10a fuses when it should draw only 60% of that. Wondering if there's a hidden wire chafed & shorting intermittantly. Or if I should anticipate surges that blow it, so step it up to 15a fuse.  Also I didn't examine the entire circuit. Mainly just where I worked. I am confident in the quality of my work but crap happens.

Could this be a surge blowing the fuse?  Or keep digging.

[Firefighter]

I never worked with LED lights, if everything was fine for two hours and now it pops the fuse instantly, I would think somewhere there is a short. You said you isolated wires, did you plug in the ten amp fuse while isolating the different lights? One of them should blow the fuse I would think. I would not leave the 15 amp fuse in if it doesn't call for that size. Wish I could help.

[John Schmidt]

I'm wondering if you might have abraded some wires under the rear fender near the connector. I'd run a separate wire/jumper to the various lights, see if the 10 amp holds when testing each light with just the single wire/jumper. When one blows the fuse, that's the light or circuit you start looking at a bit deeper. Finding the reason for blowing a fuse can be a real bugger sometimes.

[WintrSol]

5.6A at 14 V is ~78W, which is several times the stock filament bulbs; the stock tail/brake bulb is only about 26W with both filaments lit. LEDs usually draw around 1/6 to 1/4 the amount of power, so you have a real leak there. Remove the bulbs and measure; measuring LEDs don't give an honest reading, because they are not linear, and the reading you get will vary with the meter and scale you are using. Besides, if you are getting Ohms that low, it shouldn't be the bulbs, unless they are fried.

[MarkT]

5.6A at 14 V is ~78W, which is several times the stock filament bulbs; the stock tail/brake bulb is only about 26W with both filaments lit. LEDs usually draw around 1/6 to 1/4 the amount of power, so you have a real leak there. Remove the bulbs and measure; measuring LEDs don't give an honest reading, because they are not linear, and the reading you get will vary with the meter and scale you are using. Besides, if you are getting Ohms that low, it shouldn't be the bulbs, unless they are fried.

The br/wh (also bl/br & br/bu) circuit powers, on the OEM schematic, the taillight (not the brake), license, front markers, meter lights, neutral indicator, engine coolant temp indicator unit and powers the tach and it's indicator lights.   It's considerably more than the tail light.  On mine, add rear markers and highlighter, but reduce the draw for them all being LEDs (except the license bulb).  However they are a LOT of LEDs - hundreds of them. Lemme see, 48 for ea rear marker, 56 for  the highlighter, 128 in the Radiantz fixture, 24 in ea front marker and 10 in the 2 meter lights. 338 LEDs. Plus the license tungsten bulb, indicator lights and tach. All of the LEDs & bulb work. Implying either an intermittant short or some kind of surge.  I think with measuring LED circuits, what you are measuring mostly is the resistors in the circuit.  LEDs themselves have pretty much no impedence.  Without a resistor in the circuit, powering an LED by itself - it will pass unlimited current until it explodes.  Instantly. Done that.

[MarkT]

I'm wondering if you might have abraded some wires under the rear fender near the connector. I'd run a separate wire/jumper to the various lights, see if the 10 amp holds when testing each light with just the single wire/jumper. When one blows the fuse, that's the light or circuit you start looking at a bit deeper. Finding the reason for blowing a fuse can be a real bugger sometimes.

Yep that method is what I did and found no short.  Well sorta.  Didn't run separate jumper to each light but broke the circuits apart into groups to isolate them.  And the problem refused to repeat.  I hate intermittant problems.  That implied I moved some wires such that they stopped shorting so I looked for that.

Back to my question.  Is there such a thing as surges in LED circuits. Sufficient to blow a standard fuse that is rated about 66% more than the load as calculated from impedance measurement and then Ohm's law.  I know there is surges in motors, and also in analog circuits when the circuit is made / broken.  Kinda like water hammer in plumbing.

[hubcapsc]

Read the "Automotive Lighting" of this web page (http://www.mouser.com/applications/lighting-circuit-protection/)
to see people talking about surging... surging might not be your problem, but you're not
whacky to think it might be, it is a real issue that people address...

-Mike

[pancho]

The short answer to your question is  NO , you need to keep looking for the source of the problem.

[Gryphon Rider]

FYI, the following PDF contains a time-current graph for Bussmann ATM fuses.  It looks like a 10A fuse will blow at about 17A after 1 second, and will hold up to 12-13A nearly indefinitely.

http://www.cooperindustries.com/content/dam/public/bussmann/Electrical/Resources/product-datasheets-a/Bus_Ele_DS_2048_ATM_Series.pdf

[pancho]

Hey Mark,, a couple of things,, one, in doing your calculations on current draw, instead of doing a resistance reading to ground and feeding that into ohms law, try starting with the wattage rating of each individual element and going from there.  Two, your multimeter amps/current function will give you a direct look at current draw in each leg or through the fuse itself,, replace the fuse with the meter in the 20 amp range. Without seeing a wiring diagrame of your circuitry, one place I would suspect is the dimming circuit for the guage lights, depending on the rating and how it is wired, a lot of current can flow through the resistive element. Hope you get it sorted soon,, nothing like an electrical problem to keep you tied up.

[WintrSol]

The br/wh (also bl/br & br/bu) circuit powers, on the OEM schematic, the taillight (not the brake), license, front markers, meter lights, neutral indicator, engine coolant temp indicator unit and powers the tach and it's indicator lights.   It's considerably more than the tail light.  On mine, add rear markers and highlighter, but reduce the draw for them all being LEDs (except the license bulb).  However they are a LOT of LEDs - hundreds of them. Lemme see, 48 for ea rear marker, 56 for  the highlighter, 128 in the Radiantz fixture, 24 in ea front marker and 10 in the 2 meter lights. 338 LEDs. Plus the license tungsten bulb, indicator lights and tach. All of the LEDs & bulb work. Implying either an intermittant short or some kind of surge.  I think with measuring LED circuits, what you are measuring mostly is the resistors in the circuit.  LEDs themselves have pretty much no impedence.  Without a resistor in the circuit, powering an LED by itself - it will pass unlimited current until it explodes.  Instantly. Done that.

Still, counting bulbs, I get 3-5W, 1-7W. 2-1.7W, and the tach and indicators, which are about 1.5W each, plus whatever the temp unit draws. Total less than 30W of lamps, so, with LEDs, it should be less than 6W (unless you added a LOT for show), because LEDs at the roughly equivalent output draw about 1/5 the power of filament bulbs. On the IS, add 2-3W and 1-5W at the back, and maybe a few more at the front, so still less than 40W of filaments.

There may be than many LEDs total, but in the brake/tail and running light/signal bulbs, not all are on, or they are at reduced power, to get the dual brightness effect. Also, a typical individual (old-style) LED draws somewhere around 0.1W, so it would take 400 of them to get to the draw of the filament bulbs.

When you measure a LED (with most, not all), you are measuring the diode junction in series with the resistors; diode junctions are very non-linear, and you will get a value that varies, depending on how much voltage and current your meter outputs, which, in turn, depends on the meter and chosen scale, because junctions can start to conduct at less than .5V, but can rise higher, and LEDs can be as high as 2V forward voltage (or more) at full current. So, its not very useful to use a typical Ohmmeter. Plus, diodes don't surge, because of the series resistance, unless you have a defective one. You have an intermittent short, and to search for them with an Ohmmeter, you need to unplug the bulbs, the connector that leads to the tach and other indicators, and the temp unit, and start bending wires. The failure is likely where the cables get flexed a low, IOW, between the chassis and the steering head.

Failing that, if your meter has an Ampere scale that is high enough, connect it in place of the fuse, and start plugging things back in. Since the tach and temp units are unknowns, you should see about 1/3A plus whatever they draw.

[MarkT]

Good discussion & tips guys.  Gives me more to work with.  I'm also gonna save this in my threads library on the PC.  Likely useful in the future, after I get this one fixed.

[MarkT]

OK, thought I might have found it but no joy. There was a sharp edge inside the marker pods I modified but it had not cut the wire.  I eliminated it anyway. Putting the meter on the fuse location, measuring load and manipulating wires and adding bulbs one at a time revealed nothing. So now I'm going to make the short reveal itself.  I broke the circuit into segments and put several inline fuses in, one for ea part.  Put 5a fuses in ea of those, and increased the whole circuit fuse to 15.  So any short will blow the local fuse.  If it never shows up, OK not gonna lose sleep.  Or if it does, it will fail only the segment not the whole circuit.  And I can work on finding it there.

[WintrSol]

Other than where the cable gets twisted at the front, and any changes you've made, the next most common spot is under the seat. Sometimes the wires get shifted so that the seat can move them; not a real likely condition on the Valk, but something to look at.

[MarkT]

Other than where the cable gets twisted at the front, and any changes you've made, the next most common spot is under the seat. Sometimes the wires get shifted so that the seat can move them; not a real likely condition on the Valk, but something to look at.

Thanks.  Not an issue in my case.  The wires under the seat are closely managed and they are also heavily involved in my circuit mods - first place along with the wires to and at the back in my diagnoses.  In this view the wires are obscuring the relay panel.