I don't think so cause my test bike doesn't have electrical issues. For a long time I thought Dave had a power problem, I was surprised by the results from removing pin 13 from the circuit. By the way, removing pin 23 (another earth) had the same results, so that maybe wants a look see too.

The tacho remained at 0 for about six seconds. Now it then goes from there to 6,000rpm then settles down to about 3,000. Since Dave's problem often doesn't last that long, I guess the tacho symptom only happens sometimes.

With my brick I got the following lights on the cluster:

Pin 13 removed, Key on, light switch off; low fuel, neutral, high beam, oil, battery, drunk tacho. 

Pin 13 removed, key on, light switch on; high beam, oil, battery, top back lights (so no low fuel light even if the fuel is low) 

The choke light worked in both cases, but not with pin 6 (power) removed, which is to be expected.

Some of these symptoms sound a little like what happens when the battery runs low and the starter relay fries itself.
You could have some low voltages due to a poor battery or maybe your alternator is not outputting the 14v the system needs. Check the charge voltage across the battery terminals with the motor running at 2k rpm. Sometimes the voltage regulator/brushes just needs a clean but if the voltage is low then it's an easy replacement.
The battery voltage after a good charge and a 12hr rest should be reading close to 13v.

Maybe this is better than my earlier idea. Next time it cuts out, turn on the choke switch. If the choke light comes on, points to bad earth. If it remains off, then comes on with the neutral light, points to a power fail.

daveyson wrote:I don't think so cause my test bike doesn't have electrical issues. For a long time I thought Dave had a power problem, I was surprised by the results from removing pin 13 from the circuit. By the way, removing pin 23 (another earth) had the same results, so that maybe wants a look see too.

The tacho remained at 0 for about six seconds. Now it then goes from there to 6,000rpm then settles down to about 3,000. Since Dave's problem often doesn't last that long, I guess the tacho symptom only happens sometimes.

With my brick I got the following lights on the cluster:

Pin 13 removed, Key on, light switch off; low fuel, neutral, high beam, oil, battery, drunk tacho. 

Pin 13 removed, key on, light switch on; high beam, oil, battery, top back lights (so no low fuel light even if the fuel is low) 

The choke light worked in both cases, but not with pin 6 (power) removed, which is to be expected.

How do I “remove Pin 13”....or any pin from the instrument cluster?

Re instrument cluster: I cleaned and DeoxIT’d the male pins (unaware of how I can clean the female terminals so just DeoxIT’d them)

After reinstalling the cluster, I turned the the key and I got the failing instrument lights and drunken tach for about 2 seconds and then - without any intervention on my part - the lights came back on, all became right in the world, and the bike promptly started.

Then I pulled apart the ignition switch and found it in near pristine condition: contacts glowed like gold. In any case I sanded the contact surfaces and reinstalled the switch.

At this point the bike is behaving perfectly. Of course the bike behaved perfectly for the past five months - until it didn’t.

I am carrying around a list of what lights and conditions should exist (or not) so the next time the bike quits, I will have a checklist and will be able to report back (and maybe even work toward resolving the issue).

Thanks to each of you for your input.
Dave

Here's hoping you finally found your problem.

The ignition switch needs clean contacts and a bit of pressure between them to function properly. In my opinion, the pressure applied by those little springs leaves much to be desired. Giving the switch a clean reduces the pressure needed and frees up the springs to work a little better.

Clean looking contacts can still be a problem. They don't need to be black and heavily burned to cause problems. Because of the low force pushing them together, the slightest bit of corrosion will mess up the operation.

I polished the ignition contacts.

Two additional pieces of information (this is like a puzzle, huh?) are as follow:

When bike died, the 4-way flashers would not work.

When the bike died, I was on a slight downhill. I thought I would bump start the beast and coasting along about 5 mph I attempted a bump start. Engine just turned over as if the ignition was off.

Dave

I removed pin 13 from the circuit by cutting a piece of thin glossy cardboard the width of the pin, and a little more than twice its length. 


And pushed it in its socket. 
 
 Yeah it's the lazy way. 

I only did it to see if it would reproduce your symptoms but the drunk tacho seems to have more to do with the light switch on, than earth. 

So the lights (plural)  cut out, so that points to a power fail, as does the hazards fail. 

So if all the fuses and sockets are clean and tight, I'd also look that the wires are all pushed in the same length,behind the fuses and sockets.

Your hazard problem is probably related to your previous turn signal problem thread, which I guess you still have. The flasher unit gets power from fuse three, and also from the load shed relay through fuse seven.

The choke light was also off, so points to a power fail, probably from the ignition switch or kill switch.

If the kill switch is failing, the horn, indicators, hazards and head light will still work (but the brake light won't) 

Since the flashers don't work, it's probably the ignition switch.

Having too many keys on the key ring might be a problem. If wiggling the key makes the power cut in and out, that pretty much wraps up the cause.

Hello again dear friends, I am back after six months of flawless operation with the SAME EXACT PROBLEM. No, we were never able to identify exactly what was causing the bike to quit, but I have learned a bit about corrosion, battery voltage, grounding, and maybe a little about karma along the way. Today, the bike died at a stop sign - same symptoms as in episodes I, II, III, IV and V, but today the bike sat for an hour and a half and refused to come back to life. Of course after being towed home, it started just fine.

I ride the bike daily. The bike is treated pretty well, it doesn’t get ridden in the rain (hasn’t rained here in more than a year) and it sits in the garage when not being ridden. I mention this to negate any/most environmental challenges.

What I am thinking is it may be a broken or frayed wire. And here is where I need your help. Since the bike only quits when the bike is at or near idle, what components should I be focused on which play a role of keeping the bike running at or near idle but not at speed? At speed, the problem does not exist.

Also, how can I test a wire’s capacity to carry current? Testing for continuity will only identify a severed wire. If the wire is frayed I will see continuity and voltage but under a load it may fail; this is my latest guess.

Dave

This is good that the problem now lasts for an hour, that's more time for testing.

Fuse 1 depends on the kill switch for power. Fuse 7 depends on the ignition switch for power. If you don't have power to one of these, that confirms its a power issue, not an earth problem.

If no power at fuse 1, I'd look at the kill switch. If no power at both fuses, I'd look at the ignition switch.

If it was my bike, the first thing would be to clean the ignition switch again.  My K75RT needs it attended to on a regular basis to stay reliable.  The design of that switch is disappointing to say the least.  I have had experience with two of those switches being a bit quirky at 30+ years old.

Something else I would try is unplug the flasher unit and see if it then starts. If no luck maybe unplug the instrument cluster and see if it starts with the clutch pulled in.

Did you ever investigate the kill switch, davemadsen?

daveyson wrote:Something else I would try is unplug the flasher unit and see if it then starts. If no luck maybe unplug the instrument cluster and see if it starts with the clutch pulled in.

Interesting idea. I didn’t realize the bike will start without the instrument cluster connected. This is one of the few things I can do without removing the tank to get to a particular connector. One of the challenges of sorting out this gremlin is the bike comes back to life - usually within 15-90 seconds - without intervention from me making troubleshooting a challenge. Yesterday’s event was a perfect example: after the tow, the bike happily started as if never a problem. How the hell do you fix something that isn’t broke!

Laitch wrote:Did you ever investigate the kill switch, davemadsen?

Not recently. The kill switch was a component I cleaned (along with the ignition switch) in April when this gremlin last disrupted my life. Both switches’ contacts were bright and shiny.

Yesterday when the bike would not start, the speedometer needle wavered around 20 mph and the oil pressure and alternator lights burned dimly (this behavior was noted on earlier episodes). Engaging the kill switch kills the errant speedometer antics and the warning lights turn off - as you would expect. Switching the kill switch to the run position brings back the errant speedometer and warning lights. This is not to say the kill switch is not the problem, but the switch seems to be working fine. Furthermore if it was the kill switch, I would think the bike would die at speed, not just at or near idle.

daveyson wrote:Something else I would try is unplug the flasher unit and see if it then starts. If no luck maybe unplug the instrument cluster and see if it starts with the clutch pulled in.

I swapped out flasher unit in the past. I rode the bike without a flasher for a couple of weeks without a problem, reinstalled the flasher and continued to ride without a problem…until I did. In a perfect world I would push the bike to the side of the road, remove the faring and gas tank and flasher unit, put the tank back on and see if it would start. The reality is that given the 30 minutes it would take to do this (best case scenario), the bike would come back to life without any intervention on my part and I would not be any closer to understanding the problem. If there was a way to remove the flasher unit quickly (20 seconds?), removing it would give me good information.

davemadsen wrote:What I am thinking is it may be a broken or frayed wire.

That's likely.

When I encounter an intermittent condition like this and testing doesn't reveal the defect, I don't draw any conclusions unless I can reproduce the problem. A defective kill switch acts like a defective wire, and my experience has shown me a defective ignition switch also acts like that. If cleaned components are making weak contact from lack of sufficient spring pressure or alignment, current flow will drop off but not necessarily entirely.

If you get the engine running, take the switch between thumb and forefinger and try to stall the engine by wiggling it. Try wiggling the ignition switch key while the engine is running to determine if it's a problem.

You're closing in on it.

Normally, when a hot engine quits and then comes back when it has a chance to cool the problem is usually something electronic that is failing.  In the old days, it was often the condenser in the distributor.  Today, it can be any of a couple hundred or more semiconductor devices scattered around the engine and bike.

Another clue, is that your engine quits at idle, and the instruments get a bit wonky.  The alternators on my bikes seem to be regulated to drop the output voltage at about 1500 to 2000 rpm.  A bad regulator device may not be holding the voltage high enough at idle to support all the stuff that is running on the bike. 

Check the battery voltage with the engine idling.  I think it should be at least 12.7V.  You might want to hit the voltage regulator with a hair dryer to warm it up to about 190-200F and see if that makes the voltage drop and the engine shut down.

Another possibility is that the temperature control relay may be failing.  One of it's functions is to shut down the engine when the temperature gets above some maximum temperature(I think it's something like 110C or 230F).  If that circuit is malfunctioning, the normal momentary temperature bump when you stop after a run might force a shut down. 

Does the fan cycle properly?  If not, it may be letting the engine get too hot, or it may be an indication that the relay has problems.  Even if the fan is okay, the safety shut down circuit may be failing.  If you haven't done so already, you might want to swap the temperature relay.

Another source of electronic problems may be the engine control unit.  I believe there is a transistor in it that turns on the injectors.  A combination of temperature and low voltage may not be letting that transistor work properly.  You may be able to force a failure by heating the internals of the injection control unit with a hair dryer, or maybe just try replacing it.  On the other hand, I can't see where a wonky control unit would cause the instrument cluster to act strange.

Good points, point seven five. Does this problem ever happen shortly after start up, or only when warm? And strongly agree with the many previous suggestions at trying to reproduce the problem, rather than waiting for it to happen. In addition to wiggling the kill switch and ignition switch, try the others too, including the blinkers, their cancel, automatic cancelling, and hazards, not just for cut off, but that they don't alter the revs. And wiggling the wires. And wiggling the fuses. 

If you want quick access you can remove the fairing and relay box cover, then you only have to lift the rear of the tank enough to remove the relay, OK it's rough but doable. 

The speedo, alternator light, and oil pressure light all went wonky. What they have in common is power from the green/black wire, which comes from fuse 1, which gets power from the kill switch, but the kill switch gets power from the ignition switch. The horn gets power from the ignition switch through fuse 7. If the horn works when you have this problem, you want to look at the kill switch. If the horn also fails you want to look at the ignition switch, maybe replace it even if it looks good.

OK I'm not good at this but giving it a go anyway. I've read the thread again looking for the fulcrum that everything else is pivoting around. 

You said the hazards failed but came back on when the gear indicator did. The first thing they have in common, in my summary, is the ignition switch, so I'll guess the horn won't work next time it fails.

Maybe the Voltage is too low, maybe it's OK but something is robbing excessive power. 

You say it typically fails at a stop sign.
Does any of this ring true:
It only happens when the engine is warm.
It only happens on hot days. 
It only happens with the brake light on 
It only happens with the indicators on (or when you're stopped and the indicators are on for longer than normal)
It only happens with brake and indicators on (I've had this problem)
The flash to pass light also fails. 
The clock remains on.
The headlight remains on.

One thought - and I need you EEs to weigh in. Given my very limited understanding of Ohm’s Law, if I have a frayed wire or a corroded switch (ie increased resistance) AND voltage drops (as is the case at idle), this means that the current also drops. So…what components are sensitive to a drop in current to the point of allowing the bike to die? My first thought is a relay - but then again I don’t understand relays.

Since most of us have no way of reliably measuring the current in a circuit or wire, voltage is the measurement that matters the most.  Most automotive electronics will start to get a little wonky when their power supply voltage gets down into the 9-10 volt range. 

That is why there are load shed relays on most modern vehicles.  By getting rid of the headlight and accessory electrical loads on the battery, the starter has the full capacity of the battery available, and the voltage drop when cranking  is minimized.

Dirty connections are one of the man causes of voltage drop, and on our old machines you can count on lots of dirt and corrosion.  That's why everyone is always talking about cleaning the connectors with Deoxit.

Battery condition and alternator output are also important.  A weak battery will not provide enough power to run the engine controls, and a bad alternator/voltage regulator will not keep the battery charged enough to do the job. 

I had mentioned checking the voltage regulator earlier because I had a problem with it on one of my bikes.  I wasn't sure if indeed it was the regulator or the brushes because the brushes were well in spec.  However, replacing the regulator corrected the problem with stalling when coming to a stop.

"how can I test a wires capacity to carry current"

One way, use a headlight like a test light and wiggle the wire. 

Does your brick have a clock? If the clock also fails, or resets, the problem is upstream of the ignition switch. I'm thinking about the red wire from the battery to the starter relay which then goes eventually to the ignition switch.

The clock does not require a reset. The problem happens primarily with the bike has warmed up.

The suggestion to attempt to recreate the problem rather than react once the bike stalls is an excellent suggestion. To this end, I will wiggle wires and connectors to see if I can recreate the problem. I will also try to keep my head about me if the bike stalls and record what work and doesn’t work (although I find this difficult when the traffic is backing up behind me (the bike never seems to stall at opportune times).

OK, I'd look close at the red wire from the ignition switch to where they seem to join. Maybe follow it back to the starter relay too, since the lights are dim there might still be enough power for the clock. Maybe turning the handle bar to the extremes might reproduce the problem by stretching the wires.

If no luck reproducing the problem, you could run a switched jumper wire from the battery to the ignition switch, when it stalls turn on the jumper switch, if it starts with your start button, ride on (you won't have to wait five minutes or an hour) the problem is upstream of the ignition switch.

If bulb fail light comes on then its ignition issue. The electrical part is readily available for the R bikes and an exact fit.


If bulb fail light comes on AND clock resets to 00:00 then its upstream of the iginition.

If neither come on then its in fuel/ignition system, Hall sensors etc. Possibly the plug into FICU or ICU


Odd occurrences....check earths and starter brushes

I'm finding this tricky. The clock doesn't reset, but the lights don't cut out, they go dim, so the problem could be upstream of the ignition switch, there might still be enough power for the clock. Some other stuff I'm thinking, if it was power to the fuel relay that wouldn't effect the gear indicator. If two of the four back lights went out, or dim, that might point to an earth problem. If a jumper wire across the ignition switch solved the problem, Dave could ride away without waiting and then the ignition switch could be replaced later.

Or a light installed before the ignition switch, if the light cuts out or goes dim, the problem is upstream of the ignition switch.

Have a look at starter brushes, easy on non ABS K.....also check your earth terminal.

Is the battery good or is it old? Load tested? Just because it may show correct voltage does not mean it is good.

Faulty load shed relay can do this, clock does not go through it. 

Is starter relay a genuine one?

Ignition fail is usually a complete loss of electrical power, you are not experiencing that.

davemadsen wrote:One thought - and I need you EEs to weigh in. Given my very limited understanding of Ohm’s Law, if I have a frayed wire or a corroded switch (ie increased resistance) AND voltage drops (as is the case at idle), this means that the current also drops. So…what components are sensitive to a drop in current to the point of allowing the bike to die? My first thought is a relay - but then again I don’t understand relays.

Current is drawn from a source of energy (battery) and the amount of current drawn is proportional to the load.  So, if all is correct (no bad contacts, wire size is large enough, good ground returns), the load will demand current from the source and that current will be delivered.  For example, a headlamp is 55W.  From a nominal 12V system, it will demand 55/12 = 4.6A.  if the source jumps up to 14V when the alternator is working, the current demand will drop to 55/14 = 3.9A.  So your statement a drop in voltage results in a drop in current is wrong.

When corrosion happens, then the source voltage will be dropped across the corroded part and the remainder across the load (assuming a good ground).  Thus the voltage the load sees is less that what is available.  In our example, the headlamp demands 4.6A from a 12V source.  But lets say, there is a corroded headlight switch where 2V gets dropped across it.  Thus the headlamp sees 10V, not 12V. It will still demand 55W of power so to satisfy that demand the demand current rises - 55/10 = 5.5A.  This increases current draw causes heating of the corroded contact which can melt the switch.  We see this happen on the K-bike.  When a battery get old it's internal resistance increases, which causes two conditions.  First, the battery terminal voltage drops significantly as the current demand causes greater voltage drop across the internal resistance.  Second, the chemical reactions in the battery are worn out and they cannot generate enough energy to deliver the current being demanded.

Semiconductors like diodes, transistors, FETs, SCRs, etc. are current sensitive.  If they see less source voltage than what they are designed to work with, they behave very badly.  Parts like the ICU, ECU, MAF will begin to act up if the voltage goes below a value (I'm not sure but I'd guess 10V).

Relays have pick-up voltages and drop-out voltages.  They are not current senstitive as much as they are voltage sensitive.

A lot of good information and ideas, thanks! 

I will check the alternator brushes and the contacts to the rectifier. I have ordered an analog voltage meter per Clymer’s rectifier testing procedure (only have a DMVM). If the rectifier tests good, my next step will be running a wire between the battery and the ignition switch per your recommendation.

I haven’t had the battery load tested recently. I did check battery voltage after the bike was towed home at it was 12.65v which would seem to indicate it wasn’t discharged when the bike died.

Best case scenario is being able to replicate the problem in the comfort of my garage thus having the time and opportunity to better troubleshoot things.

Dave

I pulled the voltage regulator and found one of the two brushes were outside the service limit. I opted to not replace just the brushes (although a YouTube video almost convinced me soldering in new brushes was going to be a cake walk). The new unit will be in my doorstep within a week and hopefully it’s installation will bring a close to this drama. If not…we will surely meet again. 

Let me sign off with a huge THANK YOU to each of you. Your willingness to help coupled with your expertise are appreciated.
Dave

mike d wrote:I pulled the voltage regulator and found one of the two brushes were outside the service limit.


Do you mean the starter motor brushes?


Mike

I guess I wasn’t clear. The brushes out of spec are on the voltage regulator/rectifier. The voltage regulator has a pair of brushes - very similar to those in the starter motor. There is also a piece of spring steel which provides a current path (or maybe a path to ground) and a grounded grommet which needs to provide a low resistance to the case (ground). I was not aware of these connection/ground points and others who may deal with electrical issues at or near idle should ensure these contact points are included in their maintenance routine.