The K100 scrambler I built and have been maintaining for my customer is having a new issue that is stumping me. I just cleaned the fuel injectors and installed a new hall sensor. He was having issues with the bike not wanting to start after riding. After installing those I encountered a new issue. I took the bike for a test ride around the block and and it cut off on me. I started it back up to ride back home and cut out again like the kill switch keeps getting hit. Then a puff of smoke from under the tank.

The green wire to the ignition switch was torched. Cleaned the switch, cleaned the start button and kill switch, and checked/switched the load shed relay. Don't see any other wires or anything that might be fried. Everything seemed to operate as normal. Put it all back together and it started up and idles just fine, tail light didn't come back on for some reason. Then as soon as I applied throttle it cut off and the green wire to the switch started getting hot again.

What could be causing this? Is there something that sends power through the circuit from the ignition system that might keep frying it?

If there's a short between the kill switch and the load shed relay at pin 86, the green wire could get hot and the headlight wouldn't cut out while cranking. 

You want to know if the light cuts out while cranking. If it does, you want to know if the parking light above the headlight works and if not, if it's blown. Also if the tail light is blown.

I've had a hall sensor wire that had exposed wires from years of rubbing against the frame, which made me imagine this super long shot; if the accelerator cable flexes when you twist the throttle, it could push an exposed wire against the frame, or something like that.

The Hall sensor is brand new with new wires. I didn't see any wires with exposed sections between. Checked between the ignition switch, right hand control, and the load shed relay. The light wasn't cutting off completely before when cranking. I swapped the relay and now the headlight cuts off when hitting the start button. The tail light is LED. It works when the key is turned but didn't come back on after cranking it.

I'll have to take my volt meter and check every possible point for a short maybe?

Have you done other mods? Ks don't like led tail lights unless you put in a resistor. Did you remove the bulb monitoring unit?

Dunno the year and type of brick so I'll guess RS 1985. Also guessing the green wire doesn't get hot with ignition on/ kill switch off, since the load shed relay now works. (the previous load shed relay sounds susso, if now the horn relay, the horn might not work) 

Guessing one of the green/yellow wires (from the kill switch) also gets hot when the green one does, if not with ignition on/kill switch on/engine off, then with ignition on/kill switch on/engine on.

If none of the green/yellow wires get hot, it could be the kill switch shorting out which might explain why it cuts out when you twist the throttle as they're both on the right handlebar. If you put an Ammeter between the battery earth post and it's lead, with ignition on, if it increases a lot with a throttle twist, that might be it. You want to include a fuse a bit lower than the maximum rating of the meter, in case of a dead short.

daveyson wrote:Dunno the year and type of brick so I'll guess RS 1985. 

Megaton's profile states it's a 1987 K100 but doesn't state the model.

Goodo, so it's not one of those later model 4V bricks with their way different wiring.

92KK 84WW Olaf wrote:Have you done other mods? Ks don't like led tail lights unless you put in a resistor. Did you remove the bulb monitoring unit?

 
Yes, all the lights on the K are LED. The blinkers are mBlaze pin turn signals. Gauge cluster was swapped to a small single digital unit using a BEP 3.0 box. I think this one still has the Bulb Monitoring unit. I modified the stock flasher by breaking the one connection so it works with the LEDs.

It is a 2V one. Don't remember which trim it was, though. Been using the wiring diagram 2 for early K100s in the Haynes manual which seems to be the right one.

I checked the Right hand control wires and they doesn't seem to get hot. I thought the kill switch might be shorting out. I disassembled it and clean it. The contacts didn't look burnt and everything was tight. That switch was also replaced with a fairly clean also NIB one.

Do you think the bulb monitoring unit is responsible for the tail light not staying on after it's in run? Would that cause the key green wire to get hot? I built the bike a couple years ago and it was fine for the first couple thousand miles he put on it. Then it's just been coming back to me for maintenance. This most recent time is because it wouldn't start anymore. I had already changed the plugs, wires, and replaced the fuel pump and what not. This time I cleaned the injectors and replaced the hall sensor. It fired right up. idled for 10 minutes. I rode it down the block and then it sputtered and died. I started it back up and rode it back. Right before making the last corner before my house the power cut off like I hit the kill switch and that green wire melted the first time. I soldered all new wires to the ignition switch after cleaning it and wired it back together. Everything was working but the load shed wasn't fully cutting the lights when I hit the button. One of the times I hit it the wire started getting hot in my hand so I cut it. Soldered a 13 amp 16 gauge wire to the green wire instead. Switched the relays and it worked fine and fired up. Let it idle and warm up for a few minutes and seemed fine minus the tail light being off. Blipped the throttle and it cut off. Wire started getting hot again.

Do I need to eliminate the bulb monitoring unit? Is it possible that it's an ignition or timing problem with the new hall sensor? Not sure why it would melt that wire instead of just stalling. Just trying to figure it out and get it reliable again for him to not have to bring it back to me except for his routine maintenance.

An Amp meter would be the go. 

If you don't have one, take notice if the green wire is good at idle for a bit, and if it only gets hot after twisting the grip. 

If the handlebar switches aren't original look at what's on the right side with the twist grip. If the accelerator cable flexes with a twist of the grip, look if it's pushing anything.

I wonder if the owner dropped it and reconnected a handlebar switch wrong. I'd ask if he did any work to it before it stopped working.

"The tail light works when the key is turned, but not after cranking" If that means with the key in the park position then the power comes from a different circuit. You want to know if the parking light above the headlight is good or blown. You want to know if the horn works since you switched the relays.

Here's another long shot. I've never seen an American version but I think there is a jumper wire in the right handlebar plug (under the tank) cause they don't usually have a headlight switch on the right side, like they do most other places. Spose it's come loose and touching the green wire, or the owner hooked it up wrong, or something.

The engine and frame numbers will tell you more about the model and date, at least here anyway.

"do you think the bulb monitoring unit is responsible. .."

Don't think so. You have a serious short, probably best to fix that first. Guessing the tail light is a different problem. Guessing you have tail light with ignition on, the load shed relay cuts it out while cranking, as it should, then you don't get it back, sounds like something funny with the load shed relay. If the indicators don't work properly, that might point to the load shed relay too.

daveyson wrote:Don't think so. You have a serious short, probably best to fix that first. Guessing the tail light is a different problem. Guessing you have tail light with ignition on, the load shed relay cuts it out while cranking, as it should, then you don't get it back, sounds like something funny with the load shed relay. If the indicators don't work properly, that might point to the load shed relay too.

I unwrapped the wires and unplugged all the harnesses to check for shorts. Green wire doesn't trigger as a short on my meter but I'm getting 85 ohms resistant from the green wire to ground the ground point under the tank. Looking at the relays I noticed the diagram on the side shows it has a resistor between 85 and 86 pins so it's never actually disconnected. That resistor is registering right around 85 ohms. Both relays tested the same. I tested the same blue relay type on the harness of one of the K75 in my garage and it reads the same. Another K75 I have in the garage has orange relays that read 100 ohms on that resistor.

Is it possible the short is coming from this resistor always being in contact and the bike being mostly LEDs using less power. Maybe extra amps are getting caught up in the circuit that would normally be absorbed by the incandescent bulbs. Can the K bikes run using regular relays without the internal resistor on the horn and load shed circuits?

There should be no current flowing through that resistor unless the start button is pressed.
That resistor is there to prevent spikes of voltage when the relay releases, don't disconnect it or get a relay without it as the voltage spikes can damage the sensitive electrical components of the ECU and ICU.
Regards the green wire melting, the only way that happens is when there is about 3x current it is meant to carry so go looking for a direct short to eth.
P.S. check the top shock mount bolt, I have seen the wiring harness get caught under the washer on the inside and that creates the sort of havoc or worse that you are getting.