Sterling,

The SG input battery is providing a pulsed input (non-sinusoidal) to
the coil as the transistor switches it on & off, so the reading you
get from your meter is most likely not accurate. The waveform loosely
resembles a half-wave rectified triangle wave.

How are you measuring the input current? Voltage readings across a 1
ohm resistor? An ammeter in series with the input battery?

If you are using your meter in AC mode, it should give you an RMS
reading, but since the waveform is not sinusoidal, there is an
additional margin of error. For obvious reasons, using the meter in DC
mode would not be accurate at all.

Most affordable multimeters cannot perform the task at hand. Some
meters can digitally analyze the waveform and do the calculation for
you, but they are much more expensive.

The readings you get from your meter are useful for relative
comparisons when attempting to minimize current draw for a particular
configuration. But it is not useful to compare this measurement to a
DC current measurement.

This is most likely the discrepancy you see.

Additionally, the Peukert Effect comes into play again, becuase the
battery is providing a pulsed current, you are providing a blast of
current, then a rest, blast/rest/blast/rest, which will extend the
capacity. I don't how much this differs from a true RMS current
measurement over time, but thought I'd bring up another varaible that
could affect capacity measurements.

For better proof of this concept, it's probably just best to compare
apples to apples and run all capacity discharge tests with DC loads.

Jim


--- In Bedini_SG@yahoogroups.com, "Sterling D. Allan" <sterlingda@p...
> wrote:
> I'm in the middle of (1) my control load test and (2) my Bedini
charge test.
>
> The input battery and the control were trickle charged for 24 hours
together in parallel, and had the same voltage at the beginning of the
charge.
>
> After sitting for about six hours, I then ran a battery capacity
test on them. One measured 93%, and the other measured 94%.
>
> I put the stronger one on the control, to give it the advantage.
>
> The other I put on the Bedini SG circuit to run the motor-energizer.
>
> I started the control on the bulb load about 20 minutes after
starting the other on the Bedini.
>
> The bulb load is pulling in the range of 0.21 amps. (right at the
C20)
>
> The Bedini is pulling in the range of 0.25 amps. (a little below
C20). It is at the first optimum by way of input v. output amps v
rpm.
>
> Here is the interesting thing:
>
> Electronics would say that a load of 0.25 amps would discharge a
battery faster than a load of 0.21 amps.
>
> However, that is not what I am seeing.
>
> What I am seeing is that the drop in voltage between the Bedini
input battery and the control battery running the bulb is nearly the
same, and that actually the Bedini input battery is dropping in
voltage slightly slower than the control. The Bedini battery started
(after the first 10 minutes when things stabilize) 0.02 volts below
the control, but is presently just 0.01 volts behind. [Update. Four
hours in, they are now at the same voltage: 12.52v and 12.52v.]
>
> There is one possible established electrical explanation, and it
could have to do with the length and gauge of wires I'm using.
>
> On the Bedini, I have 12 gauge, short wires connecting things up.
>
> On the bulb load, I'm using around 16 gauge wire of ~24 inch
lengths.
>
> One from negative to one terminal of the bulb. one from positive to
positive lead of amp meter, whose cables are around 12-20 gauge, three
feet long. Then from there through one more ~16-gauge cable to the
positive terminal of the bulb.
>
> In all, that is quite a bit of resistance I am introducing. And
more alligator clip connections versus the male-female connections on
the Bedini.
>
> Could that be the reason for this effect I am observing?
>
> Please note that when I put the batteries now being charged onto the
load test that I plan to use the same configuration of wires on each.
>
> One way I could ascertain the difference is to (in a separate
experiment) use thick gauge, short wires on one load, and my present
configuration on another, and compare.
>
> Until I run that test, the present observation about near analogous
rates of discharge, despite different currents, doesn't have as much
meaning.
>
> Perhaps those of you with more electronics knowledge are chuckling
to yourselves at what could be a very basic principle that most
experienced in the field know as second nature.
>
> Please advise.
>
> Sterling
> http://peswiki.com/index.php/Directory:Bedini_SG:Replications:PES:
Sterling_Allan:Data