What's the best cap dump circuit?

[forelle]

Hi All,

I got this cct working along with John K's recommendations.
My question now is how do I set the voltage at the the end of the dump?

bro d

Hi All,
I have the same question,because i have only 4V difference,eg. from 24V down to 20V,or from 30V down to 26V.
Thanks

[Diverfun]

Hi Diverfun

Do you have any ideas on how to stop the dump from going down to 12v?

Mike

I am not so sure that it is necessary. If you are using an SCR version cap dump, the SCR turns off when the current stops flowing. (ie... cap voltage equals battery voltage). Also the 555 timer dump version stops dumping when the timer tells it too whether the cap has finished dumping or not.

In my circuit, you adjust the lower trip point by changing the value of the resistors. That will start/stop the dump at whatever voltage you want it too, so it will take some experimenting on your part. I tweaked my circuit again to get very close to accurate values using this calculator - http://www.random-science-tools.com/...calculator.htm. Give it a try and let me know what you think.

Dean

[brodonh]

Hi Diverfun,

Can you briefly describe your ssg setup with the battery capacity and charge time.

Thanks,

bro d

[RS_]

Diver Fun,
"The Mosfet gate turns full on (saturates) at around 4v and also has a limit on maximum voltage - usually 20v."

This is NOT Correct...! Only MOSFET's with Logic level gate inputs are 5VDC.... Most MOSFET's have a 15V or 18VDC gate input voltage, and have to be pulled up to those voltages to fully turn on......
on the comparator circuit the gate voltage comes from the 15V zener power for the comparator,

on other mosfet apps, i use a 15V or 18V zener on the gates to protect them.........

[Diverfun]

Hi Diverfun

I am not saying your idea is wrong. What I need to hear is that you can produce this form of radiant power and have actually tested these circuits on it.
All of the parts I have seem to lock up on this stuff.

You got a machine of some kind?

I am sure you have some good points for normal circuits.

Also I see you left out dealing with the way these circuits should stop at 18volts not 14v.

Mike

Sorry - I posted an incorrect circuit. (It has now been corrected.)

When you use a cap dump circuit, John Bedini states that it converts radiant energy to normal, so tuning the circuit to the battery is not needed, so normal rules apply to components. So if you are cap dumping a 80+ volts to a 24v battery, putting that 80+ volts on the gate is a very bad idea when only 4 volts is needed. Also, you can adjust the resistor values to start + stop at whatever voltage you desire - this circuit is flexible.

Yes, I have a ssg/sg machines and solid state devices and yes I use this circuit for my cap dump. I also have a degree in computer electronics and have designed circuits for years, so I know what I am taking about.

[Diverfun]

I wanted to put my 2 cents worth in.

Most of the Mosfet circuits in this thread are driving the gate incorrectly. The Mosfet gate turns full on (saturates) at around 4v and also has a limit on maximum voltage - usually 20v. Most of these circuits are tied to directly to the charging cap through a transistor so they momentarily see the full charge of the cap at the gate which is enough to fry the Mosfet. Instead of using a 10k resistor, try dividing it down using 6K and a 3K resistor to safer voltage levels. Also the gate has a capacitance value which will hold the Mosfet ON longer than you want it too, so the 3k resistor will discharge faster than the 10k.

Also there is allot of wasted power by driving LEDs +Opto couplers and diodes used for voltage reference. Current needed to drive these items can be put to better use (like charging the battery).

I am including a Op Amp Driven Mosfet circuit that has many advantages. It uses a 5v Regulator so the Op Amp can drive the Mosfet directly and is a voltage reference for the schmitt triggered design. Since the Op Amp + Mosfet are voltage driven you can use large value resisters to minimize power usage - Meg Ohm resisters work just as well as K ohm. R1 (2 ohm) resistor is used because there is no current flow at the gate - only voltage. This maximizes the speed of turning on and off the Mosfet.

Theory of Operation: As the cap charges to 30 volts - the voltage is sensed at the Op Amp through the 1 Meg and 100k resistors reducing it down to 1/11th of its value. (33v is about 3v at the Op Amp) The Op Amp is on at this point causing R4 to be in parallel with R2 causing the trip point to be about 2.8v. Once the charge Cap reaches 30v (2.8v at Op Amp) the Op Amp turns off and drives the Mosfet on, also R4 goes in parallel with R3 lowering the trip point to 1.3v. Once the Cap discharges to 14v (1.3v at Op Amp) the Op Amp turns off the Mosfet and R4 is parallel with R2 again raising the trip point back to 2.8v.

Here is the circuit - All resistor values are approximate (except R1) - Many schmitt trigger calculators on the web for finding resistor values. I make no claim that these resistor values are accurate for YOUR circuit - only the concept is accurate. Adjusting the value of R2, R3, + R4 will most likely be needed.

[brodonh]

Shan, Ron,



Here is the correct connection for the opto and BD243C collectors

[ATTACH=CONFIG]1556[/ATTACH]

Hi All,

I got this cct working along with John K's recommendations.
My question now is how do I set the voltage at the the end of the dump?

bro d

[forelle]

Hi Ron,

Ive attached your schematic below, along with a pic of the monopole Im testing with. My "artsy" computer is down at the moment, the one I do image manipulation on. And your schematic is so neat and pretty I'm hesitant to stain it with my changes

So I'll just mention my additions, which John K helped me with.

First, my cap charger is a 3 coil monopole, each coil has 6 x 0.9mm power strands. Input is 12v @ 1 amp, though I have tested it up to 120v @ 1 amp with a variac. Capacitance is 22,000uF on front and back, 75vdc electrolytic. @ 12 watts input the output capacitor dumps just faster than once a second @ approx 22v. Of course this can be tailored with the adjustment of the 25k pot, but John K suggested around 18v for gels and around 24v for lab's. So far working quite well. And there is alot of improvements that could be made with the coils and circuit.

The changes to your schematic are as follows:

D7 is changed to 1N4733 (5.1v Zener). Couldnt get it to work on 12v input with the 15v zener there.

JK suggested C1, 1uF ceramic and C2 2.2uF electrolytic. I used a green cap for C1, couldnt get a ceramic locally.

JK also suggested a 0.47 uF cap over pins 2/3 of the LM741, I have yet to place that there though.

I have a socket installed for the h11d1, but got that far and realized I could just use the output of the 741 to pulse my SSR directly, since it is optically coupled inside the SSR I saw no need for double isolation. So pins 1 and 2 of the h11d1 now connect to my HFS33a (32v, 100A SSR) which I had lying around.

Omitted the Power LED and subsequent 47k resistor once I had it working. Also have a switch to bypass the pulse LED. On a previous build (Tesla Switch) I was using an Arduino to switch the SSR's and I found that with an LED inplace the SSR didnt switch on correctly. Removing that LED fixed the problem and the SSR switched much better. Wasnt sure if by chance it was doing the same thing, so have the option to switch that LED on or bypass it. Doesnt appear to make much difference however.

Id still like to build the full schematic with the FETs however as this relay is very limited in voltage, it just happened to be what I had laying around. If I try some cap discharges with this circuit into motor coils then it will have to be rated much higher.

BTW, I love that wooden Ferris wheel you made. So oldschool and super cool. I cant look at old wagon wheels the same way now!

Regards

Shan





PS: I cant seem to find the post by Mike K, I seem to remember reading it on the old Yahoo group just before it closed down. There were a number of NPN transistors from memory that he tested. I'll keep looking.
[ATTACH]1544[/ATTACH][ATTACH=CONFIG]1545[/ATTACH][ATTACH=CONFIG]1546[/ATTACH]

Hi Ren
did i understand this right,when i want to use this circuit for the SSG with one 8 filiar coil and 12 V i only have to change D7 to 5.1V?
And the resistor after D4 with 2.2K is the only one which has to have 5W,all other are 0.6W?And do i need the FWBR also?

[blackchisel97]

That looks like it takes a little juice to keep going.
this guy had a real innovative method for a cap pulser. you set the dump voltage by choice of transistor breakdown. LINK "tongue-in-cheek"
It dumps the entire cap across the charge battery, very simple. same ckt can be used with SSR, even if you can't get your SSR to dump w/ it's own breakdown.
-KR
Patrick

Hi Patrick.
This circuit works very good, indeed. Pulse can be adjusted by "pause" value and it can run from low current, such as 300mA pulsing 1A to the battery. I'm also experimenting lately with Mosfet cap "dumpers". IC drivers are far superior to transistor based buffers but they will work as well. What I found crucial is using good capacitors with low ESR and fast diodes (beside fast Mosfets, of course). It is good idea to parallel caps with same ESR rather than using single one. Same goes with Mosfets. What I'm trying is inline charger/desulfator and another version - voltage doubler with large capacitor bank. Need more time to iron this out but should work fine. I'm expecting 80A or more pulses being generated. There is no need for hefty inductors either. Quite small toroids will do. Will post more when done with it.

PS. One more thing which isn't really related to this thread but I didn't want to start a new one. This is about protecting transistors from inductive spikes in case of bad battery connection or even high impedance due to the heavy sulfation of charging battery. In case of solid state multifilar built with heavy wire little neon bulb serves no more purpose as Oh s**t! indicator. What I did was to connect transient suppressor diodes (3 x 62V in series) between collector and emitter of each transistor. These diodes come in various voltages and work similar to high voltage Zener. In case of MJL21194 rated at max C-E 200V any combination under such value should be safe. No more burnt trannies

Thanks
Vtech

[min2oly]

This is another way of pulsing cap. This circuit has been designed over two years ago by Bit's when we were working on Tesla Switch idea over EF. It can take the input from rectified SG output as well as solar panel. Cap pulser part is charging two capacitors in series and discharging them in parallel into the battery so they dump about 22V witch each pulse. First part is an inverter boosting voltage to about 55V. PICAXE works great here but you can look at this circuit as two separate modules and substitute microchip with other IC driving pulser. Just an idea.
https://www.youtube.com/watch?v=pmUheLj3kgk

Vtech

That looks like it takes a little juice to keep going.
this guy had a real innovative method for a cap pulser. you set the dump voltage by choice of transistor breakdown. LINK "tongue-in-cheek"
It dumps the entire cap across the charge battery, very simple. same ckt can be used with SSR, even if you can't get your SSR to dump w/ it's own breakdown.
-KR
Patrick

[blackchisel97]

This is another way of pulsing cap. This circuit has been designed over two years ago by Bit's when we were working on Tesla Switch idea over EF. It can take the input from rectified SG output as well as solar panel. Cap pulser part is charging two capacitors in series and discharging them in parallel into the battery so they dump about 22V witch each pulse. First part is an inverter boosting voltage to about 55V. PICAXE works great here but you can look at this circuit as two separate modules and substitute microchip with other IC driving pulser. Just an idea.
https://www.youtube.com/watch?v=pmUheLj3kgk

Vtech

[min2oly]

Anyone try asymmetric dumping?

[min2oly]

Since we know we are aiming to dump at 2 x (two times) the voltage of the battery we are trying to charge...
Has anyone experimented w/ frequency and/or

Joules per dump vs Ah rating(impedance) of the battery we are trying to charge?

What is the draw of this circuit? Should there be a minimum size build before even using it?

We've gone back to the free mechanical dump using a pulley wheel commutator. It seems to work much better for my little mechanical builds than the solid state ones I've built so far.
-KR
Patrick

[hobbyrobotik]

Hi Oliver,

i use English language for my reply for all the other non German speakin members here, but i see you're German speaking Austrian - is this correct?
Because it is much easier for the both of us, let us comunicate in German. If anyone of the others is also intrestet, please let me know.
Thank you!

Oliver, kontaktiere mich bitte unter "info<at>hobbyrobotik.de" das macht alles viel einfacher OK?

Dieter

[forelle]

Hi Dieter,
can you tell me what kind of SCR you use?
Thanks

Its because i have looked around and i saw that the highest Vgt is 5V and Igt 500ma and whe the neon has 90V is the SCR then not damaged?
You have to know that i am not very familiar with electronics.
Thanks for your help.