Use for the Tesla Switch

[nvisser]

Quote:
"1. Try clip-wire short the negative "input" BJT and test run it on ever lower input voltage?
2. Try clip-wire shorting the negative output BJT and lookup the amount of energy recovered from the switch, while also ever decreasing power supply voltage x2
3. Clip-short both of the two negative BJT (=negative bus topology) and compare?
This request also applies to all other builders be it Nvisser's mod ( ) or my mod ( ) of the TS?

What You find ?
Stevan C.
P.S.
beware of loops on cables[/QUOTE]"


As I do not have a adjustable power supply I had to limit the input with the pulse width of the picaxe.My input is a 20V, 8a notebook power supply and I had to limit the input to 3A on the ammeter with a 60msec pulse from the picaxe. The output pulse is also 60 msec as I found that is adequate time for the caps to discharge into the battery.
I tried what you said but only found that the input current went up and the output stayed about the same. When I shorted both the negative transistors the BD140 on the serial transistor blew up and I stopped the experiment.
I do not plan to blow my huge transistors
I am working on a serial charge, par. discharge circuit that has to be charged up with higher radiant voltage. The pc board is edged and I will probably build it tomorrow.
Here is the circuit I planned. The idea of the diodes that switch the caps from series to par. comes from Nilherob.

[scratchrobot]

scratchrobot,
You are already doing a fairly decent job
only few more "toucheups" and we could deliver the verdict re: LED-capacitors

Stevan C.

QOTD:
"On Thanksgiving Day all over America, families sit down to dinner at the
same moment -- halftime."

I tried to zoom in and made another scope shot, is this what you want to see? If so I do another comparison.

470.jpg

I also tried shorting the input BJT and could charge at much lower input voltage, the same when shorting the output BJT. When I shorten both my output goes over 4 Amp
Does this mean we don't need the negative BJT's

scratchrobot

[StevanC]

Quote:
"1. Try clip-wire short the negative "input" BJT and test run it on ever lower input voltage?
2. Try clip-wire shorting the negative output BJT and lookup the amount of energy recovered from the switch, while also ever decreasing power supply voltage x2
3. Clip-short both of the two negative BJT (=negative bus topology) and compare?
This request also applies to all other builders be it Nvisser's mod ( ) or my mod ( ) of the TS?

What You find ?
Stevan C.
P.S.
beware of loops on cables

"


As I do not have a adjustable power supply I had to limit the input with the pulse width of the picaxe.My input is a 20V, 8a notebook power supply and I had to limit the input to 3A on the ammeter with a 60msec pulse from the picaxe.
[/QUOTE]
While I have in mind all chase the self runner here, i do remind this is a PV panel intended design and most fruits are from PV-like power sources.
Also I do understand on everyone can have a lab power supply, but don't expect fruits hanging "in free air" also.
Only when I was able to build (I own a DIY PSU) a regulated power supply 0-30V 0/8A, i was able to really explore the advanced Bedini devices.

Therefrom when I put the PSU to 2A current limit, I imitate the PV I have (20V/2,5A = 45W).
Then I "imitate" cloud/sun by varying voltage from 16V to 22V and then I draw conclusions therefrom...
It's a impedance mismatch correction device IMHO

The output pulse is also 60 msec as I found that is adequate time for the caps to discharge into the battery.

This is tuning the device, not tuning it's source

I tried what you said but only found that the input current went up and the output stayed about the same.

This is true when shorting the "input" negative BJT
While shorting the "output" negative BJT increases the output to about 1:3 or even 1:4, provided the device tuning is 48-2-48-2 ("fifty/fifty")

When I shorted both the negative transistors the BD140 on the serial transistor blew up and I stopped the experiment.

I'm really sorry for Your loss
I would newer suggest something I didn't try first, and found safe:
A. You shorted only the BJT them self, not one to another?
B. Did You watch the temperature?
C. Did Your LEDs survive?
D. Did You get scopeshots before frying?

I do fry a bd140 on ocassion (upon fiddling with the drive circuitry) but not too often?

I do not plan to blow my huge transistors

Nor should You, it's a marvelous build to waste. Could You hook it up to a PV?

I am working on a serial charge, par. discharge circuit that has to be charged up with higher radiant voltage. The pc board is edged and I will probably build it tomorrow.
Here is the circuit I planned. The idea of the diodes that switch the caps from series to par. comes from Nilherob.

II surely look forward to what You find

Stevan C.

[StevanC]

getting closer

I tried to zoom in and made another scope shot, is this what you want to see? If so I do another comparison.

[ATTACH]5398[/ATTACH]

I also tried shorting the input BJT and could charge at much lower input voltage, the same when shorting the output BJT. When I shorten both my output goes over 4 Amp
Does this mean we don't need the negative BJT's

scratchrobot

We need tools to learn stuff we need those BJT IMHO.

Re scopeshot:
muuch better now , can I have some more?

A. the voltage is there so I see it's a 1.3V swing.
B. The time base is still missing?
C. Could You zoom so we only see the leading edge across the whole screen?
D. Now, across what did You capture this waveform, I captured mine across the analog amp-meter

Stevan C.

[scratchrobot]

We need tools to learn stuff we need those BJT IMHO.

Re scopeshot:
muuch better now , can I have some more?

A. the voltage is there so I see it's a 1.3V swing.
B. The time base is still missing?
C. Could You zoom so we only see the leading edge across the whole screen?
D. Now, across what did You capture this waveform, I captured mine across the analog amp-meter

Stevan C.

I had my scope across the battery
batt.jpg

Here I have it across the Amp meter:
ampmeter.jpg

Which one should I focus on?
I think the time base is in the upper right corner?
Maybe you can show me a scope shot of your setup?

scratchrobot

[StevanC]

scratchrobot,
The batterry shot shows a big "hump", I have tons of questions about.(later)
The "over the amp" I would like expanded until the spike is screenwide like the battery hump is.
We have to discover the timebase yet?

Stevan C.
Ah,
lookup the PDF it's there (photos of my scope with yellow letters

[scratchrobot]

Sorry for overlooking your scope shots
I made some new ones and really zoomed in this time.

10ms:10ms.jpg 10us:10us.jpg 1us:1us.jpg

These were made with 47nf caps, I will try to compare with the other one's again.
Just found out that I can change the spike at the top of the 10us shot by changing the cap on just one of the LED's, doesn't matter which one

scratchrobot

[eternalightwithin]

That microsecond scope shot.

Is that little spike what you're inputting?!?!
And the rest of the hump is the molecules moving in the battery?

Or is it artifact ;.;

[scratchrobot]

That microsecond scope shot.

Is that little spike what you're inputting?!?!
And the rest of the hump is the molecules moving in the battery?

Or is it artifact ;.;

No I think it's all part of the spike I'm inputting but I may be wrong.
That little spike was not always there, just noise i guess

scratchrobot

[scratchrobot]

"
While I have in mind all chase the self runner here, i do remind this is a PV panel intended design and most fruits are from PV-like power sources.
Also I do understand on everyone can have a lab power supply, but don't expect fruits hanging "in free air" also.
Only when I was able to build (I own a DIY PSU) a regulated power supply 0-30V 0/8A, i was able to really explore the advanced Bedini devices.
Therefrom when I put the PSU to 2A current limit, I imitate the PV I have (20V/2,5A = 45W).
Then I "imitate" cloud/sun by varying voltage from 16V to 22V and then I draw conclusions therefrom...
It's a impedance mismatch correction device IMHO

I am back to earth
Just got carried away by that Pulse I guess
Thanks for that info and I can imagine it's a impedance mismatch correction device. I'm thinking about buying a small PV and do some testing with that

scratchrobot

[StevanC]

I have that hump
it exists IMHO

Stevan C.
NOTE:
the attached image was drawn while the Bedini_SG group was still live
The waveforms are off inductor, but I trust they apply here to?

[scratchrobot]

I have that hump
it exists IMHO

Stevan C.

So then we want that hump to last as long as possible?
On the scope shot of John Bedini's device I see a different hump, like It's more divided over time, or am I talking nonsense here
Maybe the pulse is different because he is using a DC power clamp probe.

Batt:_batt.jpg vs Bulp:_lamp.jpg

scratchrobot

[nvisser]

Series to parallel, radiant input charger

I did get it finished today. I blew the BD140's a few times and then switched to larger 2SB1375 of which I do have a lot from crt monitor scrap.
As I drive this circuit with a 8 transistor Bedini sg coil with the trigger coil connected to positive as in the joule thief , the input cap of 220uf charge up to a high voltage quite fast and that proofed to be fatal to the pnp's.
What happened was that I forgot to disconnect power to the radiant coil while I reprogrammed the picaxe during which time the switching stops and the input cap charges up to a high voltage that in turns blows the pnp's as soon as the first pulse appears again!
Now that I know that things runs much smoother.
The input pulse need to be about 1000msec to charge the series caps (5000uf) up to about 80V. The discharge pulse can be 100msec to discharge the then par. caps of 80 000uf charged to 20V each bank, to the battery.
The input amps are a constant 1A that keeps the input cap charged and with each 1 sec pulse it goes up to 5 A
The output pulse of 100msec are something else!
On the normal panel meter it shows only 5A. On the fast reacting Suzuki meter it pulses over 15A!
I want to try and put a 0.1 ohm shunt inline and then measure the voltage amplitude of the 100msec discharge pulse on a scope to try and determent what the current really is using ohm's law.
I don't know if this pulses are good for the battery, but the charging is monstrous!!
I will try and make a bad video tomorrow.
Here is the final circuit I used. Anybody that needs the PCB layout on pcb express, just give a shout.

[Bit's-n-Bytes]

I did get it finished today. I blew the BD140's a few times and then switched to larger 2SB1375 of which I do have a lot from crt monitor scrap.
As I drive this circuit with a 8 transistor Bedini sg coil with the trigger coil connected to positive as in the joule thief , the input cap of 220uf charge up to a high voltage quite fast and that proofed to be fatal to the pnp's.
What happened was that I forgot to disconnect power to the radiant coil while I reprogrammed the picaxe during which time the switching stops and the input cap charges up to a high voltage that in turns blows the pnp's as soon as the first pulse appears again!
Now that I know that things runs much smoother.
The input pulse need to be about 1000msec to charge the series caps (5000uf) up to about 80V. The discharge pulse can be 100msec to discharge the then par. caps of 80 000uf charged to 20V each bank, to the battery.
The input amps are a constant 1A that keeps the input cap charged and with each 1 sec pulse it goes up to 5 A
The output pulse of 100msec are something else!
On the normal panel meter it shows only 5A. On the fast reacting Suzuki meter it pulses over 15A!
I want to try and put a 0.1 ohm shunt inline and then measure the voltage amplitude of the 100msec discharge pulse on a scope to try and determent what the current really is using ohm's law.
I don't know if this pulses are good for the battery, but the charging is monstrous!!
I will try and make a bad video tomorrow.
Here is the final circuit I used. Anybody that needs the PCB layout on pcb express, just give a shout.

Vissie, try changing out the 470 Ohm resistors to 100 Ohm. This should allow the tranny's to turn on more without the need to extend the pulse. Great Job!

Thanks

Jeff

[nvisser]

I will try that but I don't think it is necessary as I use the picaxe outputs to drive transistors that switch 12v to the opto's.
The cap bank also needs time to charge up from the radiant energy.