C.o.p. >1

[Inquorate]

battery charging

@Groundloop - thanks, I saw your posts there, it is hard for me to post there as the site rarely loads on my mobile phone. It only works occasionally and only with one of my browsers. The reply I posted to you took six tries.

Good to hear your circuit is charging; it's an interesting point which no-one has thus far raised, and one I was uncertain of - would charging the battery with both positive AND negative energy work? And it seems it does work. Also, with 20min runs and 5min rests while (3 x 1.5AH) batteries were rotated, I found that the 'charged' batteries were able to 'run' the circuit well, without the loss of performance sometimes seen with bedini circuits.

Thankyou everyone for your interest, I look forward to seeing (or reading about) your results soon, and will try to help out where I can.

[Groundloop]

@ Inquorate,

Thank you for taking time to answer my post. I know you must be
busy with your new baby. Congat. :-)

I will build a 555 pulser with a H11D1 opto and do a test to see if
I can improve my COP.

That said, the circuit performs very well as an battery charger.
Since this is an emitter follower then the base current adds to
the overall performance. The inductive kickbak also adds to the
charging current.

I guess I need very short duty cycle on the 555 circuit?

Groundloop.

[Inquorate]

PIC chip ?

I just posted this to Groundloop at overunity.com

here's the 555 and opto circuit i used;



youtube.com/watch?&gl=AU&warned=True&client=mv-google&hl=en-GB&v=txJaQPHfKqA&nomobile=1

Pulse width is separately controlled to the off time. That way, the pulse width can be chosen to get the inductor to max magnetic field and no longer, then the pulse is cut and the spike harvested.

The off time can be extended or shortened; not enough 'off' time, and the charge battery will boil. Too much off time and the ions in the charge battery will come to rest before the next pulse, and we lose gain.

One second of pulses and rest and another second of pulses and rest etc may also work well, getting the best of the battery's ions' inertia, as well as avoiding damage to the battery by cutting off before things start to boil.

I want to do these tests but cannot just now as I have a month old infant who's taking up my time atm. But I'll tr
y to he.lp

/// someone who knows how to use a PIC chip may be able to use this technique more easily.

[Inquorate]

pulse width

I had considered a multiwire coil, but haven't gotten around to trying it yet. I put a link to a schematic with bifilar coil on the heretical builders thread.


To find the saturation point for any inductor using a scope,

09052010285 - Vox

You'll see that the voltage from the coil collapse will stop increasing when the pulse width gets beyond a certain point.

Or, without a scope, a voltmeter on a high voltage recovery cap to measure the top voltage will probably suffice. Once the voltage recovered stops increasing, you've found the saturation point.

@Groundloop - take a look at the waveforms linked to here, pulse width would want to be small; timing depends on the size of the inductor; if it's too long, any energy expended beyond max magnetic field for the inductor will reduce efficiency as we can only get so much back when the coil magnetic field collapses.

Thanks for taking the time to play with this, and thanks for the Congrats re the newborn, it'll take a while to train him but he'll be my lab assistant

[Groundloop]

@Inquorate,

Instead of baby books he will get a electronic101 :-)

I will take a look at the posted links, thanks.

Alex.

[baroutologos]

Initial attempt

Hello,

I assembled a draft yet powerful setup that i have from Melinchenko experimenting and tried to implement the concept. Initial impressions is that this is a very efficient topology.

As i see it, there too many parameters to play with as source voltage / battery voltage ratio, inductunce / frequency and duty cycle ration etc.

I am gonna proceed with my arbitrary fixed parameters and report. It should be kept a minimum experimental methodology with batteries involved of C20 discharging and at least C10 charging cycle, more than 5 continuous charging/discharging cycles etc as we all know.

Salutes

[baroutologos]

Lamp instead of Battery

During battery rest time, i replaced the charging battery with incadescence lamps. (10, 20 and 45watts).

By adjusting the lamp brightness to a known brightness level and calculating efficiency, it goes around 90%.

More experiments to do though

[Inquorate]

Hello,

I assembled a draft yet powerful setup that i have from Melinchenko experimenting and tried to implement the concept. Initial impressions is that this is a very efficient topology.

As i see it, there too many parameters to play with as source voltage / battery voltage ratio, inductunce / frequency and duty cycle ration etc.

I am gonna proceed with my arbitrary fixed parameters and report. It should be kept a minimum experimental methodology with batteries involved of C20 discharging and at least C10 charging cycle, more than 5 continuous charging/discharging cycles etc as we all know.

Salutes

@barout - with those values of hertz and duty cycle, your air core's magnetic field will get saturated and then some. Pulses should only build the magnetic field of the coil, and then cut off so the collapse can be harvested.

I cannot stress this enough.

Otherwise, you are right, it is just a very efficient battery charger.

Thankyou for posting your build and progress. I'd still be interested in the efficiency at those values.

Love and light

[baroutologos]

Hey inoquorate,
I am worried about core saturation since it wont give back magnetic field's energy.

By the way i am not using air-core coil. I am using ferrite core transformer (closed flux), Philips ETD39-3C85 model. It has a 500mTesla core saturation value (high) and its is rated up to 100-200Khz operation and can handle 100 watts.

Typical losses at 0.2T flux and at 100Khz and 100 degress C are less than 1,5 watts.

I prefer working with high efficiency ferrite core coil (close flux) rather than air core coil since the larger ohmic resistance of the air core in order to attain same inductuce values, introduces to the system huge losses. 1mm enameled wire at 30-40 watts of setup power litteraly burns that way (because at little duty cycle, current spikes are huge)

ps: My typical resistance is less than half ohm. it must be 0.3-0.4 ohms. Imagine working with a duty cylce of 25% and a smoothed current of 2 amps. 8 amps at "on" state. That means losses 8*8*0.4/4 = 6.4 watts.

[Inquorate]

the capacitor is 0.22uF, the duty cycle pot between pins 7 and 8 is 5k, I had it set on 2.2k for an air coil. The off time pot is 20k, I had it on 16k, to avoid boiling the charge battery.

I had it like this . . . . . . .
But it could also be like this .... .... ....
To avoid boiling as well, but a computer controlled pic chip would be better to investigate that.

@ barout - 90% is really good considering the deviations from my setup, especially considering lead acid batteries give up joules with 60-80% efficiency.
Thankyou for sharing your results, it is very much appreciated.

[Groundloop]

@Inquorate,

First test of my 555 controlled switch:

Input: 6,02 Volt at 0,1896 Amp = 1,141392 Joule
Input was from lab psu and was measured via a 0,25 Ohm 1% non-inductive resistor.
(The resistor is 1% so result may be +/- 1% off for the input.)

Out: 4,13 Volt at 0,187727 Amp = 0,775314 Joule
Output was tested by using a 22 Ohm resistor in parallel with a 10000uF cap.
(The resistor is 5% so result may be +/- 5% off for the output.)

COP = 0,55

At higher input voltage:

Input = 12,03 * 0,440 Amp = 5,293200 Joule
Out = 9,55 * 0,434 Amp = 4,145568 Joule
COP = 0,78

So the cop is improving with higher input voltage.

Input = 15,03 Volt * 0,5568 Amp = 8,368704 Joule
Out = 12,26 Volt * 0,5572 Amp = 6,832163 Joule
COP = 0,81

Even better.

Input = 18,02 Volt * 0,6916 Amp = 12,462632 Joule
Out = 15,00 Volt * 0,6818 Amp = 10,227272 Joule
COP = 0,82

At 18 volt the transistor started to heat up a little so this is the limit for this
circuit version. The coil remained cold.

Groundloop.

[Joit]

Actually you should can compensate all different Input with different Frequency and Duty Cycle.
I play around with a adjustable Supply, and see, that there are anywhere different Sweet spots, when you change Volt,
but you need to readjust Cycle and Frequency.
And still anywhere is a Border, you may need 10-12 V minimum, to can see Results, or switch to thinner Wires,
but more like 40V is maybe not allways healthy for the Transistor.
Maybe it depends of the Strenght of the Saturation from the Coil too.

[Inquorate]

thanks for sharing those results Groundloop

Gyula had asked me if the gains would still be evident using capacitor or if the gains were in the batteries as with other bedini based systems..

Your results show that indeed the gains must be dependent on ionic inertia in the batteries, and differences in the ion currents between the source and charge batteries, IMHO

Thankyou again

[Groundloop]

@Inquorate,

I'm currently working on my oscillator version of your circuit.

I hope to get the efficiency of the circuit above 90%. The solid
state version make sense to me since we do not need a 555 and
extra power to the 555. The higher the efficiency the lower the
loss in the circuit and more power to the battery being charged.

I also agree that the most effective way to charge lead acid batteries
is by pulse charging. You circuit solution does that very good. The
base current adds to the collector current and is not wasted. The
inductive back spike also adds to the current going to the battery.
This is excellent. All in all I'm very satisfied with the emitter follower
idea of yours. Thanks for sharing. :-)

I will post my new "beefed up" darlington version when I have fully
tested the circuit.

Regards,
Alex.

[witsend]

thanks for sharing those results Groundloop

Gyula had asked me if the gains would still be evident using capacitor or if the gains were in the batteries as with other bedini based systems..

Your results show that indeed the gains must be dependent on ionic inertia in the batteries, and differences in the ion currents between the source and charge batteries, IMHO

Thankyou again

I agree with this entirely. I think that the success or otherwise of this rig depends on the applied duty cycle and frequency. Unfortunately there's a wide potential here and some gruelling hours of testing. But the logic in the design is faultless. May I say how impressed I am at the thinking here. It's really elegant. Very well done Inquorate.