Recharging Batteries using only voltage spikes. A Results Log Thread

[nul-points]

hi Jeanna

you asked for people's experience of charging batteries from pulse circuits...

i hope the following info is appropriate (text & diagrams following in 3 sections)


my tests were with a switched-charge circuit, transferring charge from an input capacitor (or battery) via a coil and load, to charge an output capacitor with repetitive short bursts of pulses (eg 20 pulses in a burst, say)

after a number of cycles, the output capacitor charge was then also discharged thro' the load

in one of my tests, i used 1.2V NiMH AAAs as the loads - one battery received the cap charging pulses and the other received the cap discharge pulse
(i used an 8.4V 150mAh NiMH to supply the circuit)

you can see on the scope shot, the blue trace shows a burst of charging current pulses for one of the AAA NiMHs and the red trace is the combined voltage on the charging cap and 1st battery - the charged cap gets discharged into the 2nd AAA NiMH after each burst of pulses to the 1st battery

on a JT the 'pulses' would be the +ve & -ve half-cycles of the secondary output waveform

with the JT you could either charge each battery in series with a single diode across the secondary (two parallel branches, opposite polarities); or you could charge one battery inside a full-wave bridge rectifier

the first way is possibly more efficient - only 1 diode drop per battery - but divides the energy between two batteries

with the FWBR approach all the output energy is in a single battery - but there are 2 diodes in the path, so IF the final battery voltage turns out to be lower, then the stored energy is not quite so high because the power which the battery can deliver is proportional to Volts-squared

(End of part 1)

[nul-points]

...Contd...

my general approach was 'a controlled switching of energy', rather than the JT's 'tapping of free-running energy-oscillation'

i had to supply the circuit with timing signals to switch the MOSFETs on and off, whereas on the JT, the natural oscillation of the coil feedback makes the Transistor 'switch' smoothly from On-to-Off-to-On etc in its sine wave operation - so the JT is certainly the simpler of the two circuits

i didn't need to use a transformer to step-up the output voltages because i was already using higher input voltages - i don't think this difference is important because both our circuits provide sufficient peak-to-peak output voltage to charge 1.2V batteries

if you look at the general switched-charge circuit (and i've shown the MOSFET switches as manual switches to simplify the view) you'll see that the pulses of energy are coupled to the load batteries via inductor and either capacitor or diode - it's about as simple as the JT output setup

(End of part 2)

[nul-points]

... Contd. ...

ok, so that's the context - here's the data:

first, i needed to find the energy capacity of the input battery

i gave the input battery a full charge on a commercial charger

then i connected that battery across a load resistor which produced a current draw approximately the same as my test circuit  (around 20mA)

the resistor value is not critical but it does need to be measured

i connected the load resistor across the battery and datalogged the voltage 'til the battery had discharged to 6V (the 'knee' in the voltage curve)

this give me data for the 8.4V battery graph below

using a spreadsheet on the data i was able to calculate the total input energy (in Joules, or Watt-seconds) supplied by the battery when discharging from full down to 6V

i recharged the 8.4V NiMH on the commercial charger again & connected the test circuit - the test was run until the 8.4V battery discharged down to 6V

after charging the two NiMH 1.2V AAAs i used the same method (using different load resistors) to find the energy which had been stored in them

the results were as follows:

energy supplied by 8.4V 150mA NiMH: approx 3910 Joules

total combined energy recovered from the two 1.2V NiMH AAAs: approx 1800 Joules

this was a reasonably close match to the the advertised regular efficiency of charging NiMHs (ie. around 50%)

all the best
sandy

[jeanna]

Thank you Sandy,
I doubt I could repeat it back to you, but it is clear enough and I will study it a little more.

The only question I have after this is could you have missed capturing more?
In very many tests I did and still do find that it is possible to miss much, and not even know I am missing it.

In my opinion, this is because we aren't able to see the spikes that we miss.
They just aren't recorded in our meters most of the time.

My joule thief secondary is throwing spikes as high as 70 volts at a rate of 45Khz.
I am planning to put this onto a 12v gelcell, but later.

I wonder what would happen if you were to charge 4 batteries?
Would they charge as high as the 2AAA and in the same amount of time?

What if you were to reverse your set-up?
Would your NiMH AAA be able to produce high enough spikes to charge your NiMH 8.4v cell?

I saw a video of John Bedini where he said, what we're really dealing with is time.
It took doing these tests to really get what he meant by that.

Another q is about the difference it would make if you were to recharge these in parallel?
Do you suppose the spikes could be caught better in a parallel arrangement?

I really should buy John's dvd. I don't even know how he recharges a bank of 12v batteries, but I do know that I am getting higher and more frequent voltage spikes from my jtc's than he gets from his SG wheel, so I should be able to recharge a lot of batteries... right?(no free torque, though!  )

Thanks sandy,
This is terrific research you have done.

If you are interested in these q's I pose, I hope you will see what you get for answers, and post them

thank you,

jeanna

[crowclaw]

Hi Sandy

Good research and well documented. Did you find the switching frequency important to the timed rate of charge, i have already tried a similar arrangement to yours which works but used a triac for controlled pulsing. My set up as posted yesterday is the method I'm currently working on, my 8.4 volt battery will charge from approx 0.99v (fully discharged) to 10 volts in approx 3 mins. But I still have to log discharge rates before we get to excited... I have been pretty cruel with it though. As i have so far discovered the charging time depends on a critical frequency spot being found, and this for my current trial is so critical it's unbelievable. A fraction either way on a precision 15 turn pott makes all the difference!! I recall JB mentioning something similar in one of his experiments?
Nice work Sandy     Kind Regards.