To the electrically minded

This is the same issue that has stumped Kawasaki mechanics everywhere. If you do everything in the book, It must be right!. But it's not.

Thats why you need to resort to the amp-draw test to make sure the bike is operating above the break-even point.

You should do a amp-draw test on the starter too. Permanent magnets would draw 50 amps and electrified field magnets would draw 100 amps.

To test a stator hot, use a heat gun. Ohm readings between leads will go down from cold readings, but should still be in spec.

The Kawasakis were meant to be run about a half hour to recharge the battery after being started. Frequent starting and short distances led to problems.

In the end I would recomend you to get a clamp on Amp-meter and duct tape it to your tank and go for a ride or too (Extend a wire if you need to, preferably get a Amp-clamp adapter for an exsiting meter)

One other thing to do is check for a discharge while the key is off (Your'e not leaving on the parking light are you?(most likely not)). With the key off, disconnect the negative wire, if you see a small spark, there is a discharge. The most likely discharge is the regulator, remove it and retry.

Hope this helps out

Bill

loudhvx

As a mechanic, you often have to go beyond the service manual or use another manufactures techniques and keep lots of notes.

Post edited by: Capt. Quirk, at: 2007/06/06 01:59

Capt. Quirk wrote:

...
loudhvx
As a mechanic, you often have to go beyond the service manual or use another manufactures techniques and keep lots of notes.<br><br>Post edited by: Capt. Quirk, at: 2007/06/06 01:59

I don't use manuals for most of this stuff. I usually refer to my notes from when I was designing custom electrical systems for motorcycles.

Here are some of my notes:
www.geocities.com/loudgpz

I admire people that Share information on the web.

Alright everyone. Another day, more confusion.

So, I broke down...bought an electrex stator hoping to god it could just kick out more watts and gloss over whatever the real problem is. I'm getting impatient. So I charged the battery up on the counter while I put the new stator in the cover. Surprisingly, the battery charged with 20mins on my smart charger (cue suspious eyebrow).

Put battery in, crossed fingers on stator, and fired it up --- no different behavior on the voltage across battery.

Interestingly, riding home from work today, I noticed my fuel and oil lights (those normally flickered on at start by the self-checker) flicker on everytime I hit a bump....hmmm... after the ride on lake shore, the fuse on that circuit had blown. Clearly, there is some bad wiring there. However, the low voltage readings with new stator were with that fuse BLOWN, so that can't really be the issue.

Getting creative, I did the "start the bike then pull the negative cable" trick. With the sense wire disconnected, The postive batt terminal and negative cable showed 11.6 V with all normal running pieces on. Very slight revving, say to 2,200 rpms put the voltage near 15.9 V across the terminals.

With the sense wire connected, it would get up to 13, then drop to 9. Same thing early -- lacking a buffer, I assume the regular switches on bad AC averages, not real voltage.

So I'm left to believe perhaps the problem is the battery? Since the voltage with full load more than easily gets over 15 just above idle, the only thing I've removed is the battery.

Saturday I can take it back to the pepboys I bought it from a month ago, complain, try to get a replacement.

How's this sound???

Also, any ideas on the checker, oil light circuit issue? I'll run through the schematic on saturday checking wires and such, but common issues to check first might be good. Is there any way to check the black box of the self-checker itself?

Well I'm sorry to hear the Electrex didn't fix the problem, but at least it verifies your earlier findings were in line. It's possible one of your cells has a broken plate and occasionally shorts out with vibration.

Any realiable way to test the battery?

The open voltage seems good. I can do a spec. grav test this weekend, as well as ask the pepboys to do a load test.

Is there anything I can do to test the internal resistance/inductance of the battery?

I think what you want to do is measure the open voltage on the battery and see if it drops when you tap on the battery.

You could also replace the battery with the cap you bought. If you can get the bike started, put the cap in parallel with the battery then remove the battery. Do this with the reg/rec connected with the sense line connected. Then just measure the voltage on the cap as you rev the bike (with all fuses in, and normal lighting load). We know the cap's average current is zero (as long as it's not broken). So all of the alternator power is going into powering the bike and/or being shunted. This will eliminate the battery as a variable.

I would prefer a bigger cap, though, like 6000 or even 12,000 uF. Putting two 4700 uf caps in parallel would make a 9400 uF cap.

Here's the plan. I picked up a second 4700 uF cap and I will start the bike with them in parallel over the battery cables and remove the neagative cable from the battery. Last night's test leads me to believe it will function flawlessly (here's to hoping).

So I found online another way to test the battery. Along with the open voltage test (general charge) and spec. grav test (acid condition), I found a way to test the internal impedance of the battery. Taking two wirewound resistors at 1 ohm each in series, connect them across a fully charged battery briefly while noting the post to post voltage -- cal this V_1.

Then after a minute of two, measure the open voltage of the battery (surface charge now gone due to resistor load). This is V_2. Current A during test was V_1 / 2. Voltage drop across laod was V = V_2 - V_1. Therefore, the internal resistance would be V / A. The site indicated that a healthy fully charged battery should have an internal inductance of 10-30 milliOhms.

Any opinions on if this test, combined with the two others, would successfully diagnose the batter?

Well, that battery test sounds very interesting. I was taught a similar theory, but was never given specific numbers.

Basically, you can consider REAL battery as an IDEAL battery in series with a resistance. As the REAL battery discharges, the IDEAL battery stays the same, but the resistance increases.

Your battery test is measuring the resistance as a way of determining the condition of the battery. The resistors will have to be large, at least 50 watts each, mounted to a heatsink or have a fan on them.

The only problem I see would be precision in the readings. Also, you have to make sure you're not losing much voltage on the connections etc.

I forgot to mention, you can buy a battery load tester pretty cheap (under $50) at many tool stores or automotive parts stores.
www.harborfreight.com/cpi/ctaf/displayitem.taf?Itemnumber=90636

---

So you are going to do the battery-eliminator test. Whenever the voltage reaches about 13.5v should be considered the break-even point. Make sure the sense line is connected and the lights are on. If it runs ok, make an AC voltage reading on the cap as well. I'd just be curious how much AC you get there.

Post edited by: loudhvx, at: 2007/06/08 14:36

Post edited by: loudhvx, at: 2007/06/08 14:39

Across the battery cables/capacitor leads, why would you expect any AC voltage? Post rectifier, it should all be DC, yes?

dnpurdy wrote:

Across the battery cables/capacitor leads, why would you expect any AC voltage? Post rectifier, it should all be DC, yes?

Yes, but with AC content.
AC voltmeters block the DC and measure the ripple as AC voltage. The amount of ripple will indicate how much difference there is from the DC average voltage (which is what the DC meter is reading) and the total RMS voltage (which is a better indication of energy). They are not necessarily the same, but can be close as long as the ripple is small. Large AC ripple messes up voltage regulators and makes your lights burn more power than the DC meter will indicate. That's why we want to reduce the ripple with the capacitors as much as possible.

Also note, I edited my post above with a link to a cheap, battery-load-tester.

I'm off to work, I'll check in later tonight.