TOPIC:

The only thing that tells the regulator to start shunting is the voltage on the sense line, which can't really be above the voltage of the battery. But regardless, it still drains even with the sense line disconnected. The charging system just isn't keeping up with the draw. The question is why.

Either the charging system is not producing enough power, or the bike is drawing too much, or there is a bad connection.

It seemed to work with the capacitor in place of the battery, but the capacitor doesn't take any net current once it's charged, which happens almost immediately. So the battery uses enough current to make a difference. If the battery can be fully charged to 14v at less than .5 amps (should check that on the charger), then we know the battery has to be using less than .5 amps.

This defies logic so either the system will eventually lose voltage when running a cap, or the battery is using more current than it should (this means 2 defective batteries in a row... unlikely).

(Just thinking out loud here, not really coming up with anything new.)

Without going back through this whole thread,, .
Have you done a simple amperage draw test on the battery without the engine running???? Hook up your ammeter, turn everything on and see how many amps it draws that way. You could leave it on for a while to see if hot wires are causing the draw to go higher. Just hook up you charger to the battery while you are doing it. The charger will have no affect on the draw amperage reading.
I think you are over analyzing your problem. We all do it at times.
Ride safe.

[110W LOAD @ 14V, NO SENSE]
1000 rpm: 11.3V (implied 88W output)
1500 rpm: 13.5V (implied 105W output)
2000 rpm: 15.5V (implied 121W output)


[110W LOAD @ 14V, SENSED]
1000 rpm: 11.2 V
1500 rpm: 13.3 V
1800 rpm: 13.5 V --- stopped here no matter higher rpm

Sorry if i've missed something coming into this late, but I too am curious if you have a weak phase from the alternator or a dead diode in the rectifier that's losing one of them.

The regulator should shunt based on the instantaneous voltage (maybe slightly damped) it sees. The battery will accept current when the instantaneous voltage is over the 13.?V threshold and send it out when its below the battery voltage.

With all three phases, the peak and trough of the reg/rec output should be high enough that the battery is always charging. If you're missing a phase, the trough will be below battery voltage, and the battery or cap will try to pull it back up. Thus, even though the alternator can supply enough _average_ power to run all loads, the battery still has to provide power for some fraction of a revolution. Caps are very good at this sort of short-term storage, but batteries are less so.

If you were able to measure current into and out of the battery seprarately, that would be interesting, though I don't know if an extra diode drop would confuse the picture even more.
Unfortunatly its too close to bedtime for me to figure out if 2 phases clamped to 14.4V would produce a DC reading of 13.5V.

If this is the case and the caps are enough to run the bike, I suppose one could use caps and battery in parallel ?

I'm also suspicious of the 1V voltage drop across the sense wire - if it's just a sense line there should be just a tiny current on it, right ?

Oops, double.

Post edited by: Electric Monk, at: 2007/06/14 00:43

ALERT -- MEANINGFUL TEST RESULTS!!!

I came home and charged the battery up real good. Battery voltage before doing anything was 13.10V. Dropped the battery into bike with only main fuse installed. There was only the coils powered, the voltage droppe dto 13.08V. Nice.

Start the bike up, run at 2,000 rpm. Bike runs nicely at 15.40V (due to high sense line voltage drop). Put the flasher fuse in and run left turn. Voltage stays at 15.40V whole time. Nice.

Took flasher fuse out and put the fuse to the running lamps in, voltage stays at 15.45V with lights on. Nice.

Added the fuse for the tachometer and it had no change in voltage. Nice.

Added the fuse for flasher and ran it. still 15.4 V WOOT! Only fuse not in was headlight.

Pulled all fuses but main and ran headlight. Still good above 15.3.

So put all fuses in, voltage drops quickly to 13.1. Pull flasher fuse. Still low. Pull tach fuse, still low. Only fuses in are headlight and running light. If I run only one of either, it goes right back up above 15.

Now both these fuses ( along with the tach fuse) are fed by one feed line from the ignition switch. My multimeter can do temperature measure measurements. The flasher fuse and main fuse when running were at 27 C. The headlight fuse when both headlight and runnnig lamp fuse was in was at 40 C. I think I found it.

Since the tach, headlight, and running light fuses are served on a bus, it looks like that line from teh ignition switch can't handle the load of both headlight and running light at same time. I have a replacement fuse box (blade style). I'm going to install that. Also will probably run some 12AWG wire from ignition switch to the fuse bus to make sure that is good.

Right now, I'm going out to run pull those three fuses and connect the cold side of all three directly to the battery, bypassing the switch and fuse bus. If the voltage stays up on that test I know where teh problem is.

==========================

CONFIRMATION!!!!!!

I ran a bypass wire from the main fuse output the the cold side of the the three subcircuit fuses (effectively bypassing the wire to ignition switch, switch itself, wire back to fuse block, and fuse block for those three fuses) and teh system runs beautifully. At 2,000 the voltage is steady at 15.4 V.

I;m going to repalce fuse block and supply lines to and from ignition switch. If problem persists, I'm going to replace the ignition switch.

I know my confidence may be premature, but I'm very very excited to see it steady at 15.4 with everything running for the first time in a year.

Post edited by: dnpurdy, at: 2007/06/14 02:06

WAY COOL!!
Look at the pig tail connector off the ignition switch, I would bet that its melted inside and that is where the problem is.
Glad you got it figured out just in time to make some rides;)

What was so frustrating is apparently it only bogs down when at least TWO subciruits are running on the fuse bus.

Testing the circuits one by one wasn't enough to trigger it.

WooHoo!
It's 7am, but I think I'm gonna crack a beer for you!

Finally, something makes sense (no pun intended)!

Actually, it only makes sense partially. The fuse would reduce the voltage (and current) to the ignition, explaining the rough running,but it wouldn't explain the low voltage on the battery. On the diagram I'm using for your bike, the reg/rec connects directly to the battery without going through a fuse. So the fuse wouldn't affect the voltage (and current ) to the battery.

Some KZ's used a fuse between the reg/rec and the battery. Is that the case on your bike?

I think what was happening is when all fuses are on the subcircuit bus, the load of about 10~11 amps caused a wire or connector between the main fuse and subcircuit fuse bus to heat up or otherwise draw more current.

The earlier calcs I had the generator putting out about 185W at 4000 rpm. Since the bad connectors are after the 30A fuse but before the 10A, the main fuse wouldn't blow unless I was using over 405W.

My regular running load is about 115W. If the connection between main fuse through ignition to fuse bus is bad and using 5 amps to heat, that would consume 65 W, wouldn't trip main fuse, wouldn't trip 10A sub fuses, would drag voltage down to just below break even, which is what i was seeing.

If it is a bad connection, the amp usage would be linear with production? Therefore it would always basically use up anything available above break even? The bike ran, so if wasn't continuous discharging. Although 65 W sounds a lot, at 4,000 rpm I was going 60 mph so the cooling effect of wind could have countered some ofthe heating effects on rides.

Plausible explanation?

A heated connection reduces overall current through the connection and to the down-stream components, but doesn't use any extra power. It actually reduces overall power consumption.

The bad connection is in series with the loads, so it gets the same current draw as all of the down-stream loads, combined. The connection's power consumption comes from the fact that there is a small voltage drop at the connection and the power is turned into heat. It robs power from the down-stream loads, but the overall power consumption by the loads and the heated connection, combined, is actually lower than if the connection was good. This bad connection would actually reduce the load on the alternator, but would make the bike run poorly since the ignition would be getting a much reduced voltage and current.