Here are some results from another preliminary lower power battery pulsing test I am currently running.
Same type of 5Ah battery. Start voltage for this test was about 12.98V because I had
run a few short tests with a different setup with the same battery just before starting this new test.
(This SLA battery's unloaded terminal voltage when fully charged sits around 13.10V or so).
The battery voltage rose a bit to about 13.03V/13.04V after starting the test.
I started out pulsing the battery with 8ms width pulses which pulled current pulses from the
battery at about 340mA peak as measured with my scope current probe. However, the pulsing frequency
is 1Hz, so the average current draw from the battery is in the low mA range, such as roughly maybe
50mA average or so, or maybe even less. I am pulsing the current from the battery through the primary
winding of an AC transformer (to take the place of the winding for a DC motor or pulse motor).
In this test I am also pulsing a small 0.6W neon bulb as a small load with each pulse. It is interesting
that although in the first hour I was pulling current pulses from the 5Ah battery at about 340mA peak,
once per second, the battery voltage stayed right around 13.03V to 13.04V for the full first hour of running.
The only current that I am feeding back to the battery is through the neon bulb load, making use of the
switching spikes to flash the neon bulb. I am not charging and discharging a capacitor in this test.
After a few hours I reduced the pulse width to about 5ms (from 8ms) as I noticed the battery voltage was
starting to drop a little bit (down to about 13.01V/13.02V). This setup has now been running for over 5 hours
and the battery voltage is now at around 13.00V, so it has only dropped about 0.03V or 0.04V after pulsing for more than
5 hours. Not too surprising I suppose since the average current draw from the battery is not very high
compared to the battery's current capacity, but I think it shows some potential. With some further fiddling
to improve the energy recovery I may be able to get it better. Once I have it as good as I can get it, I may try to scale
it up to higher power with a larger battery to see how the battery voltage holds up while drawing much higher current pulses.
I am still working on ways to improve the energy recovery, so this is still preliminary.
Here's a very short video clip showing the transformer I am using for the coil, and showing the
pulsing of the small neon bulb load. It is not flashing quite as bright now since I lowered the the
pulse width to about 5ms after a few hours of running, but it is still flashing fairly bright. I will leave
it running until tomorrow to see how the battery voltage holds up. It seems to be very slowly discharging
the battery though, so it will probably continue slowly discharging over night. No magic happening yet.
Short video clip:
https://www.youtube.com/watch?v=aBt5Qmcuh7Q(The analog ammeter seen in the video can't respond fast enough to show the true current peaks).
I have attached a scope shot showing the current pulses from the battery when
the pulse width was set to 8ms. With the pulse width set to 5ms, the current pulses from
the battery are peaking at about 250mA.
P.S. I am driving the gate of a MOSFET from a signal generator, so the switching circuit is not
being powered from the battery in this test. If it was, the battery voltage would no doubt be
dropping more quickly. Charging caps draws more energy from the battery as well, so you are
going to need to find a fairly sizable energy gain somewhere if you want to run the control circuitry
from the battery and keep the battery charged up all the time as well. Maybe the big flywheel
with a motor/generator setup and the alternating pulsing while coasting with the flywheel can
add something special into the mix.
All the best...