Donald Smith Devices too good to be true

[hotrod68r]

you brainy bastard.nice work. i'm guessin you know it's wise to have a bit of a voltage clamp or something in place in case it goes critical mass.

[janost]

This mosfet-JT actually charges the battery it's connected to.

I'm trying on a 12v, 7Ah lead-acid now and the voltage climbs there to.

[janost]

This works.
I have not used any added power other than the batteries tested.

The scope pic is measured over a 0.3ohm resistor from the battery.

[hotrod68r]

care to share your schematic yet? if you want a bit of time to optimize it first no worries.sounds promising.

[janost]

It pulls current from the battery for 2mS and then the kickback puts it right back again.

[Nick_Z]

mr.clean:
After watching your video:
Bi Toroid Technology: Comparison, Tune, DC Testing - YouTube

I see that shorting the secondary does something to my set up, as well.
That is also normally how I have to kick-start the system, at times. But, at first I get the lower stage of resonance, and after a minute or two, the second resonance point kicks in, and the Cfl bulbs get much brighter. I suspect that even though I'm feeling little to no heat on any of the transistors used, that as they warm up, the second resonance point comes up and takes over. I can also make this higher resonance point show up by shorting the primary coil to the secondary. But, if the whole system is still cold, this kick-start of the secondary resonance point does not work until a couple of minutes go by and it all warms up at the lower resonant point. All I'm using is three transistors, and no other components in a combined multi coil circuit. I'll show some pics, soon.

[janost]

care to share your schematic yet? if you want a bit of time to optimize it first no worries.sounds promising.

This is what was on the table in the picture.
There is 200VAC on the secondary but I'm not sure how to use it yet.

It also needs a careful adjustment of the pot to keep the current draw a minimum.

Hm, the 0.47F supercap is only 5.5v and the battery is now at 11.23v
Didn't think of that but it's still holding up.

[citfta]

Question?

Hi janost,

Is there a mistake on the latest drawing you just posted? Is the wire going to the center tap supposed to be connected to the wire going to the top of the pot? That looks like it is shorting out one half of the primary windings. The earlier drawings did not show that connection.

Respectfully,
Carroll

PS: I guess you fixed it while I was writing the post. Sorry.

[hotrod68r]

maybe a supercap is more receptive to the spike.i've got a half arsed theory that instantly on swich off isn't the best time to feed a pulse to a battery.maybe current has some sort of pseudo inertia or something within a battery.just a gut feeling more than anything else.power on.

[janost]

Have to go to bed now but will leave it running.
My guess it will keep climbing.

Have to do this with a ferrite-slug to get the frequency up.

I think the secret is the low on losses in the mosfet.

[janost]

Hi janost,

Is there a mistake on the latest drawing you just posted? Is the wire going to the center tap supposed to be connected to the wire going to the top of the pot? That looks like it is shorting out one half of the primary windings. The earlier drawings did not show that connection.

Respectfully,
Carroll

PS: I guess you fixed it while I was writing the post. Sorry.

Sorry
I saw that and changed it.
It is supposed to pass the wire.

[janost]

maybe a supercap is more receptive to the spike.i've got a half arsed theory that instantly on swich off isn't the best time to feed a pulse to a battery.maybe current has some sort of pseudo inertia or something.just a gut feeling more than anything else.power on.

I think it's the rest-time.
Drain a quick pulse and let the battery idle for recovery and the pushing the pulse back into it.

The rest-time is the trick.

[hotrod68r]

do your scope shots show a rest period?i'm under the impression there wouldn't be one.could be wrong.this circuit is my best so far for return v input with a transistor.i'll get a supercap and try it.maybe power off while your asleep --improved circuit details and schematic (post 9022) on pages 301-302--

[janost]

do your scope shots show a rest period?i'm under the impression there wouldn't be one.could be wrong.this circuit is my best so far for return v input with a transistor.i'll get a supercap and try it.maybe power off while your asleep

Yes, you are right.
There isn't any rest between drain and charge current but there is a 15mS rest inbetween those events and the battery idles.

8:00 this morning the battery was at 11.24v

And when I disconnected the JT it jumped up to 11.43v

The Supercap due to it's ESR cannot deliver a current over 260mA.
But at 12v the mosfet-JT has a peak current of only 133mA in a 2.5mS pulse.

[janost]

Nothing I ever built before made the terminal voltage of a battery to climb.
No matter what the starting voltage was, it has always kept sinking.

There is noway it should climb, how can it?
This is actually very amazing.