Bedini SSG - self sustaining

[mscoffman]

I have EXCELLENT NEWS:

My next step was logically to try to scale this thing up. So I did. On the schematic below you can see I added a few more switches. The frequency now is about 1 mhz.

The sequence of switching goes like this:
SW12 and SW11 are ON. SW4, SW8 and SW9 are OFF. Then I turn ON SW3 and OFF and right after turn ON SW4 and OFF. I do that sequence for about 10 seconds or less, until cap reaches the voltage of both batteries added up (I know strange).

Then I momentarily dump the capacitor into battery B1, for that I turn OFF SW3 and SW4 and turn ON SW8 and SW9 for about 1 second. Voltage of the cap goes to the same voltage of B1 and than I turn OFF SW8 and SW9. Now I go back to to the beginning and I repeat the process to infinitum.

The cap WILL charge to the voltage of both batteries added up, and that is FREE. Strangely enough battery B2 will also charge while the process happens.

Important to understand the SW3 and SW4 are never ON together so there is NEVER a closed loop
except when dumping the cap into B1.

My next step will be using a series of batteries in series and see what happens. Since there is never a current flowing (except when dumping the cap into the battery) it will scale to very high voltages pretty well. The voltage increase is about 10th of a volt per 10 minutes in each battery. I would say pretty fast too.

Fausto.

Now we are getting somewhere!

:S:MarkSCoffman

[gyulasun]

....
The cap WILL charge to the voltage of both batteries added up, and that is FREE. Strangely enough battery B2 will also charge while the process happens.
....
Since there is never a current flowing (except when dumping the cap into the battery) it will scale to very high voltages pretty well. The voltage increase is about 10th of a volt per 10 minutes in each battery. I would say pretty fast too.

Fausto.

Hi Fausto,

Would like to understand why you think the charge up of the capacitor to the batteries' voltage level is FREE? And there is no current flowing into the cap when it is charging?

I mean the well known exponential curve for the voltage -time function when charging up any capacitor:
assuming zero voltage in a cap, the moment you switch say 12V across this cap the current is at the maximum, dictated only by the residual resistance of the battery inner and any other circuit resistance in series, including the ESR of the cap.
Then this maximum peak current (which can even be as high as some Amper in case of some thousand microFarad cap) will exponentially decrease as the voltage in the cap increases.  And I think the charging current comes from the batteries.  IT is ok that in your circuit the cap starts charging up from zero voltage only at the first switch-on then the voltage varies between 12V and 24V (neglecting the diodes forward voltage drop) but to get the 24V it needs current. 

I know Thomas Bearden referred to some article on the lossless charge-up of capacitors in small voltage steps but would it mean that no energy is taken from a source when you load the voltage source by an initially empty capacitor for very small but consecutive time durations??

Thanks, Gyula

[plengo]

@Gyula

Thanks for the comments.

You're correct. I should have qualified a little bit more my statements before. What I mean by FREE is the fact the both the batteries are charging plus the cap, so it is free because it is not being deducted from the batteries.

When I say no current I really mean also the fact that the circuit is a "open-loop" circuit. I never close the loop to allow the "regular" current to charge the cap. It must be a different process to explain the current that goes to the cap since it is not "circulating" via the batteries in a known fashion (since the batteries are not discharging).

I have a new circuit where I hell simplified the previous version and all the batteries charge up while running. No caps, no coils, nothing just the batteries and the switches from my optos. I also show the sequence of the events so all can replicate.

I guess the short duration of pulses from the batteries is causing somehow the batteries to charge.

Sequence is:
turn SW11 ON
turn SW12 ON
turn SW4 ON

turn SW4 OFF
turn SW12 OFF
turn SW11 OFF

turn SW9 ON
wait for 500 ms
turn SW9 OFF

repeat for ever.

The diodes in the diagram represents how the switches are in reality.

And believe me, I tried all sort of combinations and timings and the way I show on the sequence of event was the only one that work so far. Very strange. On the scope it shows as a 100 volts sharp pulse at every 1 seconds or so. Probes accross switch SW9.

Fausto.

[gyulasun]

Hi Fausto,

Thank you for all the information. You surely have a interesting circuit indeed, especially this last one in which you left out the capacitor.

I think one possible explanation why the batteries charge up slowly (and not discharging) is that I assume there is some nanosecond common ON time for the switches, this may come from both the PIC outputs and the signal propagation delay for the optos,  so I mean the two batteries are short circuited for a few nanoseconds even if you did not design it that way.

And considering the high repetition rate of some hundred kHz or the 1Mhz you probably still use, this may be an equivalent treatment for your batteries like a desulfation circuit gives to the batteries.  This could explain the small voltage increase.  And even if there is no any nanosecond overlap at all, maybe the charge can go through the switches' output capacitance, making a route for the charging current.
I understand if you do not think my explanation is correct. 

Thanks, Gyula

[plengo]

@gyulasun

I have been playing with those "nanosecond" switching times and all sort of combinations that leads to your explanation. Certainly I can see when the switching of all combined switches (via their ON state or combination of capacitance and so on) is ON at the same time, I have a huge current going through and off course, discharging the battery, BUT, that is not what I am doing. I am carefully choosing the switching sequence and timing that avoids exactly that.

I can measure at any point in time only a .15 ma (point 15 of a mili-amp) during the pulse where the circuit is absolutely in a open loop (no closed loop in any form or shape, via capacitance, inductance or whatever). That has been my biggest difficulty in achieving that and it is not easy, but once I get that sweet spot where there is not current anywhere flowing only this pulse thats when the batteries start charging.

I think I am correct in stating that based on the scope shots, a 1 miliamp meter in series with the battery and the programming sequence that is very absolute. I am lead to start believing that a single wire is being used as a coil on the OFF state, not on the ON but on the off. That's the moment where the magic happens and the 100v pulse shows up and charges the battery.

Fausto.