Cap Dump Circuit

[BroMikey]

Gate resistance danger

Okay Group Members I need to ask this.

Why would someone use 1000 ohms on the gate when 10,000 ohms opens the gate fully? Unless they were using a lower voltage? I mean some guys are using the same circuit and charging the same 36vdc battery packs and yet showing a 220 ohm gate resistance for fet irfp260.

The question I have is if 1000 ohms opens the gate fully and 10,000 ohms opens the gate fully won't the higher ohm resistor be easier on the fet and make it last longer?

Thanks again to all.

Mike

[BroMikey]

pulsing a mosfet hard

(1 or 2) 'fets, 1k from source to base ( each ) 1k to base ( each ) should get ya in the ball park....

...unless yer pulsing "Giga-Amps"

No two 'fets are exactly the same lab/spec and can work out of sinc with each other , thus work against each other.

J. Bedini talked about this in matching transistors, however what's bad for the "goose" can be the same for the "gander".

Also, ( 1 ) 'fet should more than handle what you are doing w/sink. If not, increase the 'fets "wattage" handling capability.

You shouldn't be hearing any sound at if firing correctly...

750 watts will run cool...$$$$

K.I.S.S. formula - "Keep..It..Simple...Stupid.

Hello Stargate

I charged up my 120,000 uF bank just for fun to 65-70vdc and pulsed that through one big mosfet to a battery and that thing only worked once.

Maybe I will end up will 12 instead of 6 AAA?

But 1 or two? it's little legs just can't carry it.

My cap bank is set to deliver massive amounts of inrush to the battery.

By the way I did go down to 10,000 ohms of resistance on the base to get a full On so as to dump the entire load but that fet gave up the Ghost.

Just had to see. That pulse also burnt a tooth off of my gator clip so maybe the shorting out added resistance smoked the fet I do not know these things.

But one thing is sure the Wiley will always be my friend forever after that stunt. Yes big big inrush amps enough to blow the sulfation right of a huge battery plate.

I am going to pick up some 500 ah 4vdc cells in a few hours.

So I need this thing working come hell or high water.

Yeah 10k on the base gave me a "FULL ON" condition and it smoked that fet in a New York mil second.

Mike

[Mario]

Okay Group Members I need to ask this.

Why would someone use 1000 ohms on the gate when 10,000 ohms opens the gate fully? Unless they were using a lower voltage? I mean some guys are using the same circuit and charging the same 36vdc battery packs and yet showing a 220 ohm gate resistance for fet irfp260.

The question I have is if 1000 ohms opens the gate fully and 10,000 ohms opens the gate fully won't the higher ohm resistor be easier on the fet and make it last longer?

Thanks again to all.

Mike

Mike, the higher the gate resistance the lower the turn on/off times will be. Yes it will turn on with high resistance, but instead of emulating a switch it will emulate a dimmer…. not what you want.

If you're using an opto to switch the mosfet it's not going to be as easy as driving a transistor. A mosfet needs at least 10V (better 15V) to fully turn on. You can find logic gate mosfets which turn on at 3V. So keep that in mind, the cap voltage minus the batt voltage will determine how many volts will be at the gate.
This is why I asked if you have a scope to monitor the turn on/off states. Otherwise you're never sure that the mosfet is completely on or completely off (this is important!), and you don't know your rise/fall times which directly depend on the resistors uses.
When I really want to get precise I replace the resistors with pots, I play with them until I see the perfect switching on my scope. Then I measure the pots and replace them with fixed resistors. I would start with about 47 ohms on the gate and 1 Kohm going to the source for a good turn off.

regards,
Mario

[RAMSET]

Being discussed elsewhere

BroMikey
your project is being discussed here also [member Seamonkey started a thread for this]

BroMikey's Capacitor Dump Circuit

Unfortunately the Diagrams you share here cannot be seen unless you are a member of this forum and signed in...[some are denied admission]

Good luck with your research .

Thx
Chet

[BroMikey]

Okay Okay here comes the scope

Mike, the higher the gate resistance the lower the turn on/off times will be. Yes it will turn on with high resistance, but instead of emulating a switch it will emulate a dimmer…. not what you want.

If you're using an opto to switch the mosfet it's not going to be as easy as driving a transistor. A mosfet needs at least 10V (better 15V) to fully turn on. You can find logic gate mosfets which turn on at 3V. So keep that in mind, the cap voltage minus the batt voltage will determine how many volts will be at the gate.
This is why I asked if you have a scope to monitor the turn on/off states. Otherwise you're never sure that the mosfet is completely on or completely off (this is important!), and you don't know your rise/fall times which directly depend on the resistors uses.
When I really want to get precise I replace the resistors with pots, I play with them until I see the perfect switching on my scope. Then I measure the pots and replace them with fixed resistors. I would start with about 47 ohms on the gate and 1 Kohm going to the source for a good turn off.

regards,
Mario

Mario

You are the only guy who has got me digging my $20,000 scope out to see the switch. I knew someone would actually get me into the game and help me see what to do. I have the tools but many are unable to teach.

I have to go right now and will do tests when the scope gets setup in my new shop area. I understand you perfectly, keep talking when you do I go forward.

47-1000 it is.

I am thinking there will be some signs of distortion in the scope shot if....HUM?


Cap voltage minus the switch voltage is like 60- 80volts.

Got to think. There was something I was going to ask you.

Later for now this gives me something to think on and I just love learning new things.


Mike PS you are the only one Mario

[BroMikey]

Big Bank-O Batteries

Here is my new skid of batteries. They are fully charged excellent batteries I got from a power company who buys new ones every 5 years.

Sorry about the dark shots but not bad for no light pics.

They come 3 two volt cells per rack about 24X20X8 I have 4 now or a 24vdc bank.

It says it is a 870ah battery but I can't believe that yet. Maybe.

Will be pulse charging on Friday.



This is my appliance hauling trailer with the pallet of batteries.




I hooked them up fast and put them on my 200 amp roll around auto charger and they are at 14vdc already after 30 minutes.


That Acid has got to go soon. A toilet will take care of the acid, perfect



tonight I do scope shots, thanks Mario, you have no idea how excited I am to have someone coach me a little.

Mike

[Mario]

Hi Mike,

I wouldn't dump 80V into the batteries. It might ruin them. Go for twice the battery voltage. Say you have 12V batts dump the cap at 24V. Also, with voltages that high you risk frying your mosfet gate, unless you calculate the right resistances.

If you have trouble controlling the mosfet with the opto method there's another option. You could use a small dc-dc isolation converter (i'm talking 1 or 2 watts power), say 12V to 12V, powered from your primary batt or from whatever source you want. With that you can run your timing circuit and use proper mosfet drivers so that your mosfet gates will always switch on and off properly, fast and with the correct voltage.

regards,
Mario

[BroMikey]

Scope Setup

Hi Mike,

I wouldn't dump 80V into the batteries. It might ruin them. Go for twice the battery voltage. Say you have 12V batts dump the cap at 24V. Also, with voltages that high you risk frying your mosfet gate, unless you calculate the right resistances.

If you have trouble controlling the mosfet with the opto method there's another option. You could use a small dc-dc isolation converter (i'm talking 1 or 2 watts power), say 12V to 12V, powered from your primary batt or from whatever source you want. With that you can run your timing circuit and use proper mosfet drivers so that your mosfet gates will always switch on and off properly, fast and with the correct voltage.

regards,
Mario

Hey Mario

I dump 30vdc for a 12vdc battery that will rise to 15-16vdc for a full charge and right now my battery array upstairs is a 36vdc system so miuimum 75vdc dump voltage.

Here is my favorite circuit.

pulsing594.JPG

Also I wanted you to know that I setup my juicy scope and am trying to get use to it again. Oh I remember now I said and this scope does not show you so much as it tells you everything.

It is like a running program it is the old HP54100A/D one can store images and print. To bad for me cause I don't know what i am looking at yet.

But I can get my rise and fall down to 3 or 4ns well i got that much tonight out of it just watching the amp meter and using 50ohm 100ohm 1000ohm didn't find the pot yet but I will.

The big bank in the yard is at 14.1 and from what I know about the AGM batteries she is in need. It works out perfect. Buy half dried out brand new industrial batteries and hit them with ALUM and they will and have become better than ever.

Thanks Mario for helping figure this all out. I am in burnout mode the last 3 days and better get back quiet for a few ZZZZ

Let me know what to look for, should i use 2ns or 200ns? The resistor on the bases seem to open the gates but the real story is being told on this super awesome digital scope. This scope practically walks and talks.

Mike

[Mario]

Mike, it's impossible to get into the nanosec switching range with an opto setup. The H11d1 has rise/fall times of 5 microsecs minimum if I recall well, so even if you do really well on matching the resistors to the gates that's the fastest you can switch. With the mosfet driver option I described earlier it's another story, but I wouldn't worry about that too much for now. Get some experience first. And I would work with a small system first..

What I would worry about is getting a control signal that let's you narrow the ON duty cycle down to the 1% range. If you use a simple 555 circuit with 50% On time you discharge the cap but for a much longer time your charging source will go right trough the cap and mosfet and charge your batts directly. You don't want that. You only want a very short pulse long enough to discharge the cap down to almost the batt voltage.
Even better is to make a circuit that shuts the power supply down during the cap dump. But you can worry about that later…

Here you can see a way to control the 555 duty cycle, page 44.
http://www.free-energy-info.co.uk/Chapter12.pdf

regards,
Mario

[BroMikey]

1% On Time

Mike, it's impossible to get into the nanosec switching range with an opto setup. The H11d1 has rise/fall times of 5 microsecs minimum if I recall well, so even if you do really well on matching the resistors to the gates that's the fastest you can switch. With the mosfet driver option I described earlier it's another story, but I wouldn't worry about that too much for now. Get some experience first. And I would work with a small system first..

What I would worry about is getting a control signal that let's you narrow the ON duty cycle down to the 1% range. If you use a simple 555 circuit with 50% On time you discharge the cap but for a much longer time your charging source will go right trough the cap and mosfet and charge your batts directly. You don't want that. You only want a very short pulse long enough to discharge the cap down to almost the batt voltage.
Even better is to make a circuit that shuts the power supply down during the cap dump. But you can worry about that later…

Here you can see a way to control the 555 duty cycle, page 44.
http://www.free-energy-info.co.uk/Chapter12.pdf

regards,
Mario

Hey Mario

This is a great doc for guys to freshen up with, somehow I missed this one.

Dc to dc conversion and shutdown the supply WOW all great proposals and I was thinking of some way to throw a (HUGE) resistor in between the bank and supply to slow the bank charge time alittle but better to let it charge fast and shut her down.

Yes I need brushing up that is for sure. My circuit with the 555 timer I showed you has both the standard 555 timer 50% duty and the extra knob is for cutting off dump time.

This means that my LED will flash like the blinker on a cars turn signal and is adjustable all the way down to an ever hastening (faster) cutoff time than can be seen with the naked eye.

I must have read the scale wrong but I was varying my timer circuit up and down and the scope was showing 5 ms to 50ms or whatever the division.

Okay I know this the button I push is "RISE TIME" "FALL TIME" well all of it can easily be read I just need more time. Last night I was out of it.

Yes all of the things you are saying, right down to the nicer controlling drivers and smooth supplies have lodged in my thinking.

I have not missed anything you have pointed out or I would come right out and ask you about it.

Keep talking

Here is that circuit

timeropto.JPG

Many people miss the fact that a standard timer circuit can easily be altered to control the duty cycle without going the the 556 dual timer with separate sections one for timing and one for duty cycle.

This circuit is working on 2 of my other cap dumps that are in use now and allow me to dump caps down to any voltage I like.

For instance this circuit was running lat night and I charged the caps up to 45 volts and cut off the dump time to 35 volts showing on the meter.

I can dump at 60 and start recharging at 55volts or I can dump at 30v and stop at 25volts if I want to.

I have more to learn about the difference in using a LM741 or a NAND gate but I could start if needed.

Let me play with this scope some more and I will tell you for sure what I am getting.

I have been controlling dump down voltage and charging batteries this way for 1 year as I said with the smaller units.

In fact I am about to take my small bank of 60,000uF out to the trailer so I can get it charged up more, it is just that I wanted to use Mosfet's on a bigger system.

The one I use now has 4 deflection transistors in it and they never blink.

They run cool and are very stable.

Thank Mario for helping me use Mosfet's. Among other things of interest.

It is not like mosfets are the only way of doing this.

Mike

[lotec]

I found this one helpful
https://www.youtube.com/watch?v=10R0Mrqwjuo

I liked the little pnp for draining the gate fast.

[BroMikey]

Great Help

I found this one helpful
https://www.youtube.com/watch?v=10R0Mrqwjuo

I liked the little pnp for draining the gate fast.

Thanks for the help.
I rebuilt my fet bank and this time I used 560ohm on the bases but will need more at 90vdc dumping.

Sg Oscillator to dump to industrial AGM ALUM converted batteries - YouTube

You guys are all helpful and so i will get up to speed.

I spent 4hrs on my scope and can see waveforms now.

Mike Ps the book on this scope is hundreds of pages long

[BroMikey]

Measuring the switch

Mike, it's impossible to get into the nanosec switching range with an opto setup. The H11d1 has rise/fall times of 5 microsecs minimum if I recall well, so even if you do really well on matching the resistors to the gates that's the fastest you can switch. With the mosfet driver option I described earlier it's another story, but I wouldn't worry about that too much for now. Get some experience first. And I would work with a small system first..

What I would worry about is getting a control signal that let's you narrow the ON duty cycle down to the 1% range. If you use a simple 555 circuit with 50% On time you discharge the cap but for a much longer time your charging source will go right trough the cap and mosfet and charge your batts directly. You don't want that. You only want a very short pulse long enough to discharge the cap down to almost the batt voltage.
Even better is to make a circuit that shuts the power supply down during the cap dump. But you can worry about that later…

Here you can see a way to control the 555 duty cycle, page 44.
http://www.free-energy-info.co.uk/Chapter12.pdf

regards,
Mario

Hi Mario

I am finding my head with both hands learning this scope functions.

I had the wrong probes out of 4 probes I own. Now when I sit down it pops right on to the place where I left off.

I can go down to 107ms rise and fall is 33ms

Don't know what that means.

Mike

[Mario]

Hi Mario

I am finding my head with both hands learning this scope functions.

I had the wrong probes out of 4 probes I own. Now when I sit down it pops right on to the place where I left off.

I can go down to 107ms rise and fall is 33ms

Don't know what that means.

Mike

Hi Mike, it means that if you have a square wave the rise time is the time it takes to go from off to on (low to high) and the fall time is the opposite, meaning the time it takes to turn off.
Is that millisecs or microsecs?

regards,
Mario

[BroMikey]

Rise and Fall

Hi Mike, it means that if you have a square wave the rise time is the time it takes to go from off to on (low to high) and the fall time is the opposite, meaning the time it takes to turn off.
Is that millisecs or microsecs?

regards,
Mario

Hi Mario

Yes I realize that it means what you just said but what I should have said was I do not know how 107 ms relates to a possible 1% figure that you mentioned as a target.

Okay the div/ms on a scope humm....... not sure........ all I know is that a 107 ms flash on my led is faster than the blink of my eye.

Eyes blink at 150 ms miliseconds. I think.

Are we looking for micro-seconds?

As I dial up and down the LED will get lazy looking at 900 ms humm.........

I better research this as it really was a question of mine when I read that value.

Well at least I am getting my head on straight a little

Maybe these pulses are in mili-seconds and if so that is very slow.

Is it possible that a 560 ohm base resistor would slow the rise time down that much?

I see data showing Mosfet's switching much faster. I just don't know what a 1% switching means in relation to seconds.

I thought that mili-seconds = (mS) microseconds=(uS) and nano-seconds =(uS)

I better stop being afraid to lower base resistance, it looks like huh?

Of course I am switching at low level voltages 20-25vdc and amp pulses are

at 1-2 amps so I will go up to 90vdc at 20 amp pulses first to see if

switching time is effected dramatically or no.

This was in my mind from the beginning and so far I am studying the function

of the scope and now can see some point of reference.


This is my first day. Today is the first day in years that I wanted to get

myself in the know concerning switching times.

Thank you so much for your continued patience helping me crawl in slow

motion. Patience is a virtue I am told and is probably quite true.

I have heard your instruction on using a pot from 1000 ohms down to 47

ohms till you get what you want.

Now that I can read the scope in a small way I will complete my test at the elevated power levels as mentioned above and them go back to a single FET to test again.

Last night was the first time in my life that I sat down in front of my scope

and the reading was exactly the same as it was the day before.


Just so you know Mario 2 days ago I had the wrong high frequency probes

hooked up and I had power supply hum and you know the drill, I mess of

signals entering my view even from my own hand touching things, so I am

not totally scope illiterate.

I know very little, this is my first day to see that with the high frequency

probes I better use my 2 regulated supplies for a clean view of switching

times.


I have dozens of tiny 12vdc batteries so I hooked them up to get my cap at

24vdc and dumped the power to another battery of 12vdc using a separate

12vdc battery to run the timer/trigger circuit.


All clean voltage, no hum.

Now that I see this difference it has become quite clear that my bench

battery chargers will not work here.


I recently learned how to fool a tl594 chip found in PC switch mode power supplies and have an extra 0-24vdc regulated supply.

I also have a 0-15vdc +- 5vdc+- variable 1 amp dc supply.

My bench needs cleaning and the shelves are going in now.


It is an exciting time for me.

Mike