Bedini Comparator Cap Dump

[Nityesh Schnaderbeck]

"RS_" and "Volty" you have spent a lot of effort on this circuit. And I could imagine that many experimenters have built this.

"RS_" you have modified the original non working circuit to a more functional circuit with many fixes, thankyou for the you time that you have spent on this.

It could be possible that there are more modifications needed for this circuit. I do not know why "John Bedini" has this style of mosfet driver circuit. It really does not matter how you drive the mosfets, as long as you can achieve clean switching.

I have modified this circuit and converted the comparator to an adjustable schmitt trigger with adjustable hysteresis. With only 2 additional components(R13 and C4), it is a very easy mod. I have also added a stabilization resistor(R12) to help stop "Opto 1(H11D1)" from false triggering and to improve switching.


Bedini Inverted Cap Pulser with Schmitt Trigger.pdf


So for those who have built this circuit, all they have to do is add 3 components. So you can have an adjustable charge threshold and an adjustable discharge threshold of the capture capacitor. The value of C4(The anti oscillation capacitor) may need experimenting with. As this op-amp based schmitt trigger can be vulnerable to oscillations if not designed properly.

The scope across the capture capacitor will tell the truth of this mod.

Kind
Regards
Nityesh Schnaderbeck

[RS_]

Nityesh,

the H11D1 Mosfet driver circuit is in John Bedini's patent......
I have built many of these cap pulsers over the last 14 years or so, and have added the resistor to pin 6 as you describe, and can see no difference in operation with it, or with out it......

http://pdfpiw.uspto.gov/.piw?Docid=0...iew+first+page

[Nityesh Schnaderbeck]

Nityesh,

the H11D1 Mosfet driver circuit is in John Bedini's patent......
I have built many of these cap pulsers over the last 14 years or so, and have added the resistor to pin 6 as you describe, and can see no difference in operation with it, or with out it......

http://pdfpiw.uspto.gov/.piw?Docid=0...iew+first+page

Yes, you are right, it does not seem to make much difference, like you I have made lots of them to. But everything seems to change if you are turning on 3 transistors simultaneously like in a tesla switch. I had been working on a tri-symmetrical solid state battery switch, (My own development). that uses 9 transistors 3 coils and 3 batteries. And each transistor had it's own opto-isolator. For some reason I had lots of trouble with the switching, and the 100Kohm stabilizing resistors (from Pin6 of H11D1 to the emitter of the transistor) on all opto-isolators solved this problem, and then the switching was clean. So I just do this by default now, as an extra measure. I think the trouble can occur when driving a high gain transistor. And this seems to pick up other pulsing events in the circuit.

On John Bedini's patent http://pdfpiw.uspto.gov/.piw?Docid=0...iew+first+page the opto-isolators don't have the base pin of the photo transistors. Some opto-isolators don't have the base pin of the internal photo transistor.

"RS_" Thankyou for sharing your expertise and experience with me, very happy to talk to another "free energy" enthusiast like me. I really like your feedback. I also have experimented with many free energy devices, I have a box of old circuits. As well as a box of blown up components, and melted plastic spools. It all part of the fun. And I bet you have old circuits and blown up components too

[Volty]

"RS_" and "Volty" you have spent a lot of effort on this circuit. And I could imagine that many experimenters have built this.

"RS_" you have modified the original non working circuit to a more functional circuit with many fixes, thankyou for the you time that you have spent on this.

It could be possible that there are more modifications needed for this circuit. I do not know why "John Bedini" has this style of mosfet driver circuit. It really does not matter how you drive the mosfets, as long as you can achieve clean switching.

I have modified this circuit and converted the comparator to an adjustable schmitt trigger with adjustable hysteresis. With only 2 additional components(R13 and C4), it is a very easy mod. I have also added a stabilization resistor(R12) to help stop "Opto 1(H11D1)" from false triggering and to improve switching.

[ATTACH=CONFIG]4105[/ATTACH]
[ATTACH]4106[/ATTACH]


So for those who have built this circuit, all they have to do is add 3 components. So you can have an adjustable charge threshold and an adjustable discharge threshold of the capture capacitor. The value of C4(The anti oscillation capacitor) may need experimenting with. As this op-amp based schmitt trigger can be vulnerable to oscillations if not designed properly.

The scope across the capture capacitor will tell the truth of this mod.

Kind
Regards
Nityesh Schnaderbeck

Hello Nityesh and Lman,

Thank you Nityesh for your Linear Amplifier Thread posts. I saved every file and am going to build it.

Here is the hand wired solder-side schematic i used for my Cap Dump Gang Box w Rotary Hysteresis Selector, but no extra Zener in the back end for the FET's. They seem to work OK without it.
EDIT * Cannot attach file some unknown reason *


I cannot attach Express Schematic or Express PCB source file, sorry. I added beefy fine strand Hobby Silicone insulated wire for the paths you see as 18 AWG in older photo here of solder side.

I did not use Low ESR Photo Flash type Cap's, just what i found at salvage store and had from EBay buy. I did not use faster and lower Ohms IRFP260, i did not know they existed, just IRFP250's i saw in pre-resin-potted Cap Dump schematic.

I am charging mostly all LiPo 3 cell, 11.1V, 2200mAh Zagi HP Batteries, for the Zagi HP, the worlds best flying wing: http://www.zagi.com/zagi-electric-wi...esc-or-battery

I am not sure if Lead Acid are better at choice of battery type for Cap Dumping or not. When I charge 4 of these LiPo's in parallel like this, with NO Cap Dump used to weaken those Golden Pulses Raining Down..., LiPo's charge from 11.3V to about 12.7V with this 4-Coiler in about 4.5 hours. faster without Cap Dump. It takes about 1.25 to 1.5 hours each if charged LiPo alone one at a time without Cap Dump.

LiPo's charged with SG do not need Balancing. Cells come up together somehow, always. So SG is my favorite LiPo Charger. Over-Charger too :-) As cool as this thing is to play with, Cap Dumping with LiPo's at least is not faster, it is slower. About 20% slower with LiPo's, which seem to resond to raw SG pulses better.

Still I want to use the Cap Dump to meter Dumps Rates or Joules filled and dumped with 1-Coiler Vs. 4-Coiler, all 8 Trani, 23 AWG, then I will make a 23 AWG 16 Trani Single Coil SG and answer AWG Vs. Max Tranis fitting some AWG without heat "Celebrity Death Match". Patrick says 23AWG not good enough, and i say more Trani's more than compensate for this AWG. We will see if 16 Trani's stray cool enough with rotored operation. Cap Dump seems like nice visual meter for how fast it fills and dumps.

The 4-Coiler shown draws between 1.2A to 2.5A @ 12VDC, the higher Amps if i push some DC with it. I use a 2Watt, 10 Turn, wire-wound Po,t hot-glued to that Amp meter.

With my Cap Dump Box with poor ESR Caps and lesser IRFP250's x2, with slow turn-Off times, it takes 5.5 hours to charge these same 4 LiPO's in Parallel. I have NOT tried every combination of Dump Threshold Voltage and Hysteresis Cap Selection and mapped it out. It takes me one hour to go fly off 5 batteries at full throttle, a cool 87 MPH flat and level If I charge them again right away, it is easier to charge them, and not prepared to say how much, but ~ 25% easier or faster next charge, kinda like

* * * So Nitesh, you appear to be way expert with electronics. Can you help with a PNP with NPN series configuration for faster Turn-Off Times?
1)NPN On (Active Turn-On)
2)PNP Off (Active Turn-Off)
3)NPN Off (Relaxation Turn-Off, when PNP already Off, so slowness no matter)
4)PNP On (Relaxation Turn-On, when NPN already Off, so slowness no matter)
I wish i could think of some way to make this work with a Potentiometer Trigger Adjust like it has now, but slow Off-Times make me want to try this out.




I have also built 3 Trani 2 Cap 1 Battery Tesla Switch, and used this same H1D1 to BD243 to IRFP250 FET's in parallel, and found Turn-Off times SLOW.

Relaxation turn off never as fast as Active Turn-On.


I think the SG could switch off faster in the way i listed above with series PNP and NPN. Please advise. I bet you have mastered this already?

Thank you in advance,

Ward

[Volty]

Here are the extra attachments it would not let me attach last:


The "22 AWG" runs you see were replaced with 12 AWG Fine-Strand Silicone Hobby Wire.

I expected the Amp Meter to hinder things, but the voltage at Battery is always 0.1V higher when passing through Amp Meter.

I have 8 AWG same Fine-Strand Silicone Hobby Wire, and I am really considering changing these runs again, with shorter Umbilical hanging out. My 1-Coiler and 4-Coiler all need direct soldered largest possible AWG runs. I ican find 6 AWG Fine-Strand Silicone Hobby WIre, but ony in 18" lengths from some hobby shop. Fine-Stranded Silicone insulated AWG's lower than this seem not to exist, and they are so awesome and flexible and easy to use, and no melt insulation is best.

I listen and hear and wonder about inertia described in the recent SG Call-In Questions w. Aaron and Peter, about how the real effects happen with very low impedance everything, Especially batteries, of the LiFeP0 type. 150Ah costs $600 each though, hmmm.

[Nityesh Schnaderbeck]

Volty your work is very impressive I love your work

First sorry about what I said in an early post about the "hysteresis capacitor", after seeing your schematic I can see how this works, It is a very genius way of creating a hysteresis function. I never thought of doing it that way.



I think it is better than my "schmitt trigger" method, in some ways and more simple. Now there are 2 ways of doing this, both with their merits.


The 15V zener at the back end limits the "gate to source" junction of the mosfets to 15V. This protects from over voltage. A mosfet "gate to source" junction can only withstand about 20V to 23V, typically before it gets damaged.


I can explain about the slow turn off of your mosfets. Your scope shot is showing you a capacitance discharge curve of a "gate to source" capacitance



This is caused from a parasitic capacitance in the "gate to source" junction of your mosfets. The solution is to discharge this parasitic capacitance as fast as possible for a fast turn off. For a fast "turn on" this capacitance needs to be charged quickly, so the mosfet driver has to be able to provide the necessary current to charge the "gate to source" capacitance.
The more mosfets in parallel the greater this parasitic capacitance, and this will load your mosfet driver circuit more, especially at high frequency switching.

There are a number of different ways to achieve a faster turn off.

1). Passively: Have a resistor from the gate to source of each mosfet. The lower this resistance the faster the turn off. With your schematics you only have one 10K resistor for 2 mosfets. You need to have a 10K resister from gate to source for each mosfet to discharge the parasitic capacitance of each mosfet. You can lower these resistances to 5.6k, but doing so, will waste more current, so this would be a trade off.

2). Actively: This method will give you the fastest turn off. But you will need to modify your mosfet driver circuit. And add a PNP transistor, and have a "PNP and NPN" complimentary pair. Here is a suggested active mosfet driver circuit. You may have to experiment with the value of R1.


For the capacitor pulsar a fast "turn on" is important, and for a coil (SSG coil) a fast "turn off" is important. Because capacitors and inductors(coils) are inverse of each other.

Mosfets "turn on" fast and bipolar transistors "turn off" fast.

I hope this helps

Regards
Nityesh Schnaderbeck

[Nityesh Schnaderbeck]

Hi Nityesh,

Would you care to share the comparator driven mosfet circuit by any chance?

Regards

Hello "Ren" Thankyou for your interest.
Here is a test circuit you can experiment with. I have even made a Mosfet driven SSG with this circuit, which also charged the battery.




Regards
Nityesh Schnaderbeck

[Volty]

Volty your work is very impressive I love your work

First sorry about what I said in an early post about the "hysteresis capacitor", after seeing your schematic I can see how this works, It is a very genius way of creating a hysteresis function. I never thought of doing it that way.

[ATTACH=CONFIG]4115[/ATTACH]

I think it is better than my "schmitt trigger" method, in some ways and more simple. Now there are 2 ways of doing this, both with their merits.


The 15V zener at the back end limits the "gate to source" junction of the mosfets to 15V. This protects from over voltage. A mosfet "gate to source" junction can only withstand about 20V to 23V, typically before it gets damaged.


I can explain about the slow turn off of your mosfets. Your scope shot is showing you a capacitance discharge curve of a "gate to source" capacitance

[ATTACH=CONFIG]4116[/ATTACH]

This is caused from a parasitic capacitance in the "gate to source" junction of your mosfets. The solution is to discharge this parasitic capacitance as fast as possible for a fast turn off. For a fast "turn on" this capacitance needs to be charged quickly, so the mosfet driver has to be able to provide the necessary current to charge the "gate to source" capacitance.
The more mosfets in parallel the greater this parasitic capacitance, and this will load your mosfet driver circuit more, especially at high frequency switching.

There are a number of different ways to achieve a faster turn off.

1). Passively: Have a resistor from the gate to source of each mosfet. The lower this resistance the faster the turn off. With your schematics you only have one 10K resistor for 2 mosfets. You need to have a 10K resister from gate to source for each mosfet to discharge the parasitic capacitance of each mosfet. You can lower these resistances to 5.6k, but doing so, will waste more current, so this would be a trade off.

2). Actively: This method will give you the fastest turn off. But you will need to modify your mosfet driver circuit. And add a PNP transistor, and have a "PNP and NPN" complimentary pair. Here is a suggested active mosfet driver circuit. You may have to experiment with the value of R1.
[ATTACH=CONFIG]4118[/ATTACH]

For the capacitor pulsar a fast "turn on" is important, and for a coil (SSG coil) a fast "turn off" is important. Because capacitors and inductors(coils) are inverse of each other.

Mosfets "turn on" fast and bipolar transistors "turn off" fast.

I hope this helps

Regards
Nityesh Schnaderbeck

*********

Hello Nityesh and Happy Thanksgiving to you and yours ! :-)

Thanks for kudos, but the hysteresis Cap is not my design, it is supposed to be an earlier design of John Bedini's Comparator Cap Dump Circuit, and was shown as a 1uF or 2uF fixed i recall, before he improved his designs and made it so cool he had to pot it in resin to keep pirates from making cheapo copies. It must be way space-aged to cost so much and not be serviceable per potted.

The Zener you show to protect FET's i totally failed to consider. I have driven the Dump Caps up to 45V accidentally many times, and the FET's still work somehow. IRFP250 Max Vgs is 20V :-(
I will try the Zener and check the FET's on Scope compared to some new ones to look for damage signs.

I am tired from making turkey stock, and am still soaking your Active FET driver circuit suggestion and your LM358 direct Trani' drive. The slow turn-off Scope shot i showed looks same for bipolar MJL21194, where in both the Opto turns off sharply in comparison. I hear you stating Opto not needed. I just built as i saw, and it was the Tesla Switch that made me notice the slow turn off with both. So i will try to adapt this non-inverting LM358 to the 741 Op-Amp or try it after the 741. The 741 can go close to the Supply Rail for single supply operation, and the inverting configuration is needed to sink enough current, or so i have heard ;-). It looks like the 358 single large output voltage swing would be 12V for a 15V single supply.


How fast can your bipolar MJL21194's (or favorite FET) switch on and off with this LM358 drive? Have you managed to speed up SG turn off times? Have you considered the PNP in series w. NPN for active turn-on and active-turn-off idea?
1)NPN On (Active Turn-On)
2)PNP Off (Active Turn-Off)
3)NPN Off (Relaxation Turn-Off, when PNP already Off, so slowness no matter)
4)PNP On (Relaxation Turn-On, when NPN already Off, so slowness no matter)

The SG may be a work of art not to be altered, but the Tesla Switch needs fast turn off for non-resonant inductive loads, at least for Matthew Jones Simple Switch with 3-wire XFRMR to take advantage of BEMF ping-ponging. Thanks for your suggestions,

Ward

[lampliter14]

KOODOS to RS_ for the Schematic on the Bedini Mosfet Comparator circuit. I have built the circuit as per the diagram. There seemed to be an issue in the timing of the Op amp. The Mosfets would spill the caps to the battery only if the 100k pot was manually swung from one side to the other. That suggested that the downstream section of the circuit was functioning but it would NOT automatically sense the buildup of capacitor voltage. Capacitor C2 (.1 uf) was substituted with a number of larger values but the one that was settled on was a 100 uf. After making that change the circuit worked beautifully. I am wondering if the .1 suggestion was simply a typo or is there something I'm missing? My Capture Capacitors are twin 10,000 uf Caps in parallel yielding 20,000 uf total value. Even with the analogue meter indication of the pulsing action, I still wanted an LED visual as well. Initially a 510 ohm resistor in series with an LED was installed from the output of the op amp (pin 6) to the negative ground rail and it did function BUT it glowed during the off part of the cycle instead of the on pulse. That idea was removed in favour of an LED simply being installed in series with the pin 6 output of the comparator and R1 (1.2k) which feeds the front end of the opto-coupler and that solution behaves perfectly. I am very, very pleased with this circuit and with those couple of minor suggestions so far it performs flawlessly under test. Thank you RS_ for the original circuit. Any comments are welcome !
lampliter14

[dennis foyil]

Great post. good info.

[AlvaroHN]

How do I match the mosfets???

Finally I have all the parts for the comparator (RS_ snag 2 12 v). But I was wondering what is the right way to match the irfp260 ???

tnx

Happy new year to all!

best

Alvaro

[Tom C]

http://sound.westhost.com/transistor-matching.htm

[AlvaroHN]

http://sound.westhost.com/transistor-matching.htm

Thank you Tom !!! Have a great new year!

best,

Alvaro

[AlvaroHN]

Just finished the comparator circuit. I had the same issue that lampliter14 pointed out with c2 (.1uf).

In my case I changed c2 to 2.2uf , with 100uf I had too much "on time".

My capture cap is small. 35v 3300uf.


Best

Alvaro

Pd: thank you RS for the circuit!!!!

[AlvaroHN]

Brand new doubt: Does a 35v cap charges faster to 24v than a 25v cap??? (in the case that the 35v cap have the same capacity at 24v than the 25v cap at 24v)

I ask this because the cap charge graph starts fast and end slowly... maybe is better to charge the cap until a % of the total charge and it would be faster? or is the same?

best,

Alvaro