To the electrically minded

So if I understand teh theory correctly, the AC ripple is caused by the on-off shunting in the regulator. Above the cutoff voltage, the regulator shunts to ground, and the output voltage drops. This dropping occurs until the sense circuit switches off, and then the scr's are turned off, and the votlage increases again.

AC voltmeters looking for variable voltage will ride the DC voltage as a float zero and measure the variable voltage around this constant float above ground.

So, the larger the AC I measure at when at cutoff voltage, the slower the sensing switching is occuring, allowing more wide swings between shunted min and unshunted max. Normally the battery would act as a current damper preventing this ripples from being particularly noticable.

With 9800 uF across the output, is there any particular AC voltage that you'd expect to be good/bad?

I am a Kawasaki Mechanic with over 25 yrs profesional experince, I can't believe what I have been reading.

Later Girls

dnpurdy wrote:

So if I understand teh theory correctly, the AC ripple is caused by the on-off shunting in the regulator. Above the cutoff voltage, the regulator shunts to ground, and the output voltage drops. This dropping occurs until the sense circuit switches off, and then the scr's are turned off, and the votlage increases again.

AC voltmeters looking for variable voltage will ride the DC voltage as a float zero and measure the variable voltage around this constant float above ground.

So, the larger the AC I measure at when at cutoff voltage, the slower the sensing switching is occuring, allowing more wide swings between shunted min and unshunted max. Normally the battery would act as a current damper preventing this ripples from being particularly noticable.

With 9800 uF across the output, is there any particular AC voltage that you'd expect to be good/bad?

You seem to have a good idea of what's happening. However, there are several other factors. First, the action of the alternator contributes a small amount of high frequency ripple. Second, the action of the ignition contributes a low frequency, squarish ripple. Then there is an intermittent ripple from the shunting action. All of this adds up to a complicated voltage signal on the battery. If you hook up an oscilloscope, it's all pretty identifiable. The peaks of the 14.5v DC voltage actually reach about 16v and the troughs reach down to about 12.5 or 13v.

Because of the complex AC components, it would be very difficult to estimate the AC meter's reading. Since the p-p of the AC ripple is around 3 or 4v, and in general, the overall shape is somewhat square, I would expect a reading in the 2 to 4v range, but this is a very rough estimate.

Post edited by: loudhvx, at: 2007/06/09 01:41

Interesting tests, guys. I'm enjoying following this thread.

I would do two things:

1. idle - the regulator should be wide open at this point and you should check the voltage to make sure the bike is not pulling the voltage way down. You wouldn't want to be too far below 12V here.

2. 4000rpm - the regulator (without a battery to charge) should be on a very short duty cycle (you can pull some fuses to really shorten the duty cycle). This means that for every AC cycle, the regulator might only be operating for a small fraction of the cycle.

What I would do here is measure the DC voltage, and then the AC and be sure that the AC swing doesn't let the voltage dip too much below 10.5V or too far over 15.5V. This is the range where most 12V systems will operate.



On Battery tests - a load test is always a good idea for these starting type batteries. Though if you suspect the battery has been drawing too much power from the charging system it has to turn that power into something, and normally batteries turn it into heat. This results in boiling off the electrolyte and you will see your fluid levels fall off quickly.


(but really, I'm just replying because I just enjoy seeing Capt. Quirk squirm)

Good advice Inline.
The voltage does swing a lot more with a capacitor in place of a battery so I would expect the instantaneous voltage to swing above 16v routinely with the capacitor (and indeed it does when viewed on an o-scope). But these are just short transient spikes and don't affect the average voltage by very much.

The AC meter reading won't be very informative, unfortunately, since the exact shape of the AC is unknown.

You right on with the regulator shunting early in every cycle with the lights off at 4000 RPM. The O-scope shows a very high frequency sawtooth signal under that circumstance.

Post edited by: loudhvx, at: 2007/06/09 02:08

Hepcats and Hepkittens:

Here we go. Yesterday, I went to pepboys just to see if they'd exchange the battery on my word (as I know they can't load test motorcycle batteries due to their clamps being too large) and a receipt that had gone through the washing machine. 10 minutes later, walked out with new battery.

Got home, filled it up and charged it good. Put it in bike, no difference. 12.5 at idle, raised to 12.9 by 4,000.

Did an amp draw on it. Negative 1 amp at idle, by 1,500 break even, +2 amps at 2,000. Strange thing though, as I continued to rev, the amps continued to go up, but the voltage, as we've seen before, stayed low. All the watts are going into amps....

Switched out the batt for the capacitor pair after starting. 11.8 at idle, by 1,500 it was around 13.7, and by 2,000 at 15. Due to the 1V drop on the sense line, the voltage stopped exactly at 15.48V and stayed there. The bike sounded smoother than it ever has!!. The AC across the caps was .49VAC at all rpm ranges.

If I had a kickstarted, I'd ditch that battery, construct a cap matrix and be on my merry way.

So, battery out, caps in, everything else connected and running normal, it performs great. Battery in, the amps go up, but voltage goes nowhere. It barely ever breaks the electrochemical voltage of 12.8 acorss the terminals, so I doubt it really charges. I don't want to sulfate this battery fast due to chronic undercharging.

AHHH

Are the paths identical from the rec/reg to the battery as they are from reg/rec to the caps? In other words, are the caps being connected to the battery cables in place of the battery? If the pathway is not the same, there may be a chance you are losing voltage at the solenoid connection or somewhere else between the reg/rec and the battery.

If the battery is getting two full amps, and it's fully charged, it's voltage should be going up after a few seconds of sustained 4000 RPM. And if it's breaking even at 1500 RPM, it can't be discharging on a steady ride (without a lot of stops).

The caps are being put in as a direct replacement for battery, exactly same spot. I'm going to ride with voltmeter attached to see if I can see it improve over a ride, and my testing still has been the problem.

At the beginning of the thread the first post was:


dnpurdy wrote:

I have a quick electrical question to the EE's out there who know more than me. I have a kz750n-1 spectre with spotty charging. I'm looking to test the stator rotor output. No load I get about 33 VAC output at idle between any two stator leads.

:laugh: :laugh:

Interesting thread.
Ride safe

I feel like this thread could be turned into a small book about the theory of and how to isoloate and test every componenet of motorcycle charging system. Whew.

Haven't been able to that voltmeter ride yet. I am concerned that I must be missing something crazy simple for the caps to let the voltage go up to 15.5 immediately at 2,000 rpm, but connect the battery in, and it never breaks 12.9. I was told to make extra sure my battery cable to battery connection was good so that the amps were actually going INTO the battery, and not just to it. I've sanded the battery and connectors shiny, wiped them clean and reattached with no improvement.

Post edited by: dnpurdy, at: 2007/06/11 09:58

is it possible that the caps send a signal needing a full charge and the battery only sends a partial signal?

BSKZ650 wrote:

is it possible that the caps send a signal needing a full charge and the battery only sends a partial signal?

There are two differences between caps and a battery, as far as how they affect the charging system.
1) The battery can take a small net current continuously without a continuous increase in voltage. That is because it tends to act like a voltage regulator (some bikes use just a battery as a voltage regulator). If the current is too big, it will, however show a voltage increase, and will eventually boil.

A capacitor's voltage will increase steadily if it gets any net positive current. The cap has no tendency to want to stabilize at a particular voltage. That is why it is good for showing what a voltage regulator is doing. Any limitation on the voltage can only be the result of the regulator.

2) The other major difference is that caps allow more ripple in the voltage. This tends to affect the voltage-detection circuit of typical regulators. I, like dnpurdy, have found the voltage regulators tend to regulate a little higher when connected to a capacitor instead of a battery. I remember it being about 1/2 a volt higher, or so.