Cap Dump Circuit

[BroMikey]

Pulsing 100vdc+

Here is a typical example of a commercially grown inverter board that drives a 115vdc 3 phase induction motor. I have lots of these things.

The diagrams show control board, motor and data sheet for the transistors with some ratings.

The whole point is that to pulse 100vdc a 600vdc max rated part is selected so anyone pulse charging a 36vdc battery should never use a 200v device.



This only one of many versions



This motor is used for most and a 14 amp part at 600vdc usually 5 or 6 devices to handle 800watts.



This is 10 years after Maytag went belly up and inverter boards are a dime a dozen. You can get one for $20 now used and these newer boards use common sense design topology, not some 10 percent of the max rating fly by night get rich scheme.

Mike

[BroMikey]

Rise times and false ones

Here is a video where the scope reading for rise times are ????????? or the ????
question mark is on the end of a fictitious number.

How to make a rise-time measurement on an oscilloscope - YouTube

This is the first video I have every seen that discusses rise times.

So even a modern scope does not read these fast times well in all cases.

Mike (PS still studying)

[BroMikey]

Hello Mario

Hi Mike,

If you follow the SOA curve of your devices you should not burn them.

I already wrote this, but if you use an opto the best rise/fall times you will be able to see on your Fets is 4-5 microseconds (not nano). It's simply not possible to go faster because it's the limit of your opto (h11d1).
Only using a driver you will be able to switch faster. So I don't know how you're able to see nanosecond switching if you're using an opto.

One reason for toasting fets can also be this:

When you turn the dump OFF very sharply (talking nanoseconds) and you are not dumping down to the battery voltage (even just 20 or 30V above the battery) you will get a negative spike that can be even a few hundred volts in amplitude. It's like your battery is behaving like a coil, you pulse it, and when you let go you get the flyback spike. This is why it's important to watch the dump pulse on a scope and learn all about it while you change parameters and see what happens.

regards,
Mario

Yes Mario I checked these rise times out as you suggest. The H11DX has a 2.5us rise so I am learning and trying to understand negative times as well as positive times. So far no one has answered any of my questions other than you and a few others.

I stated the rise times positive and negative have occurred with very small adjustment on the dial so who knows. Not me that is for sure.


I read some things and I see it is possible. The duty is throwing readings into the negative and giving faster rise times than 10ms. 10-20ms is showing up as well as I adjust. I noticed I had a bad scratch on my wiper on my 10 turn pot and I think this is throwing me off.

Also I read that the new opto's are faster now and I just bought them. I guess this opto is a 2.5us one.

Also I am reading the wave with a scope at the switch not the opto so when the opto fires the large transistor maybe it amplifies the speed of the signal?

The opto is connected to a pretty powerful transistor and that transistor fires the Mosfet. The Opto does not directly fire the Mosfet.


Is it possible that the transistor is firing at a faster rise time? You see as long as the opto fires the transistor fast enough and hard enough maybe upon firing the transistor it is faster. The transistor is just a medium power switching transistor that I put into the circuit to see what would happen.

The opto is able to fire the the Mosfet without a transistor between them.

I have not thought about rise times long at all and am only guessing about almost everything.

I would like to know why positive and negative rise and fall numbers appear as I adjust my 555 timer connected to my opto connected to a pretty good size transistor.

Mike

[BroMikey]

Heat Sink Build

Here are the pictures of my new heat sink where 12 IRFP460 mosfets place 3 per side.

BroMikey's Science Projects

The fan is a 115vac ball bearing unit. Unlike the pc fan that works on dc the ac fan starts up very slowly and runs quiet for 5 times the air of the inexpensive PC fan.

The sink with the 6 patches of grease with bolts shown is what I originally had using a cheap pc fan just kind of blowing it across the best way possible.

Now the the square unit the air is forced through it like a duct work much more efficiently. The square sink is built from what is shown in the pictures using those 2 short fined sinks then shims to hold 1/16th" plate bolting one sink to the other.

This sink is pounded together very tightly before bolts are used to hold it in place. I am well pleased with it's tightness for a homemade design. The reason it is so tight is that the thin shims are tapered in wedge form and tapped in.

The top and bottom are bolted plates all aluminum build from stuff I had around the shop.

Mike

[BroMikey]

New Block Diagram

Hello Cap Dump enthusiasts.

Here is the direction I am going building a capacitor discharging unit for zapping batteries. First of all I need to point out the main important thing about this diagram and that is the MOSFET NUMBER.

IRFP460

If we are going to follow just anyone around the web who "SAYS" that they now have a working example of a cap dump then let them demonstrate it, not just pass around diagrams that might work.

You will see circuits that promote the use of a IRFP250 for cap pulsing 36vdc battery banks and this is a poor choice for a MOSFET.

Why? And once again the "WHY" relates to the circuits durability and thus it's longevity. The MOSFET is no different than any other part in the industry having a maximum voltage rating and therefore common sense designs need to reflect intelligence.

Anyone following the battery pulsing/pump topology set forth by John Bedini must take into consideration what John has said. First off John and other men in the alternative energy field tell you that their circuits are not designed like conventional ones.

bye the way John Bedini made a statement in one of his videos that his work with cap dumps had nothing to do with what circuits were floating around the web and I now can see why he made this statement.

Lots of folks make diagrams of dumps that are not yet functioning in the practical world and in most cases have flaws to say the very least.

The first circuitry difference is that today's engineer is trained to snub out the ringing in the circuit and the alternative boys state that their circuits do not do that.

The Mosfet is said to be able to operate within 10 percent of it's peak ratings. So according to this a mosfet that have a peak rating of 200 volts can run all day long at 180vdc and as long as there are no other parts such as resistors, diodes, caps deviating from their values the fet will not go above 180vdc, maybe.

However this may be fine for a plastic toy robot, running the upper limits of a Mosfet is not best.

Here is a more sensible design for running a voltage of 100vdc.

The IRFP460 has a decent amp handling capacity yet has a voltage limit that will handle inductive kick backs. This has been found out the hard way with my circuits going up in smoke.

If it works don't fix it! right? Well that is right and I am throwing all of my IRFP250's back in the drawer for lower voltage projects, after all they are good parts when used the right way.

The circuit that is underway.



Like Dave is showing us all here in his animations that the coil kicks hard and a higher voltage type mosfet must be selected for pulsing 100vdc.

Our Computer PC power supplies use this part for pulsing 100vdc through a pass transformer so all I did is to look inside of a commercially produced product that is virtually indestructible to find commonly used design practices.

I don't care what the rating says, I am looking at what the industry does, not what it says. Don't do as I say do as I do or something like that.

See my hand? See my other Hand? See my data sheet and see how I build it?

Yeah the results are that the men inside who are in control of huge electronic companies follow the same rules as they always have and not the max ratings.

Otherwise they will be getting their products returned to the factory and we can't have that.

The general "RULE OF THUMB" or the "GOLDEN RULE" for any design criteria is that the parts Max rating is is multiplied by .7 and depending on who you talk to is 66 percent all of the way up to 75 percent so you get the idea.

Therefore if a Mosfet has a Max voltage rating of 200vdc and we multiply times .7 this is 70 percent and this equals= 140vdc.

Now we have the Mosfets upper voltage rating for building a circuit and in our case if we pulse that mosfet at 100vdc and the inductive kick comes back at three times that our part will soon be burnt completely out.

This is exactly what happened to me all this month long. I used 6 fet's in my line up as I built costume modules onto sinks.

As long as I stayed down to 40-45vdc as the pulsed voltage my dump would run fine for days.

The IRFP460's are in the mail and should show better results than I am getting.

2 days to a week for a set of brand new fet's is unacceptable.

I have to show you my homemade heat sink soon

Mike

[BroMikey]

Great Animation

The flyback from the coil seems to be pos according to its direction through the diode, high voltage.

A coil pulsed with a neg current gives us back a pos polarity,

Will a coil pulsed the pos current gives us back a neg polarity,

Consider

Hello Dave

I need more words from all of the guys here and you Dave to get me up to speed, but I will say this right off the top.

It seems like to have right there must be a left and to have a positive you will always have a negative.


It is very simple to me but maybe you had some other ideas in mind?

Either way I like the way you can put all of this into a picture.

Mike

[Dave45]

The flyback from the coil seems to be pos according to its direction through the diode, high voltage.

A coil pulsed with a neg current gives us back a pos polarity,

Will a coil pulsed the pos current gives us back a neg polarity,

Consider

[BroMikey]

Nice machine

Great example Dave

Mike

[Dave45]

[BroMikey]

Proper circuit 36vdc/90vdc Cap discharge

Here is a proper lay out of what a cap dump with it's ratings for Mosfet's should be.



This circuit can handle the reflected 3X spiking that can occur.

Mike

[BroMikey]

Cap steadying the transistor

Greets Bro,

I like yr style, may i share something here?
Thnx


In my xp i found out that load must be connected to a cap that feeds something when not using it, otherwise it will surge true and burn transistors.

Sofare only my transistors burned all out cause of a surge, and its not the amps

Even if my transistor can handle a socalled 8 amps or 6 forex 2n3055

A voltage or a simple spike can fock it all up, so what i do now if my device is running i am transfering that load to somewhere else, and if i dont distribute that energy i use a dummy load to atleast keep my transistor calm that after loads any transistor.

Can i see some of yr art?


Greets JB

Hey John

Good Idea here is my working circuit, not this mess I am making lately.

BroMikey's Science Projects look for BroMikeys dump

Another thing I must say is that a few years back I got word that MAYTAG lost their company because of Mosfet circuits.

I have done appliance repair 40 years. I rebuild digital circuits if need be.

The new generation of engineering squirts came into change the world with their 59 cent Mosfet to run 3 phase washer motors driven by inverter boards connected to the 120vac mains in our homes today.

Their circuits were and are so incapable of doing what they were designed to do that no one had one over 5 years without needing a replacement board costing 600-1000 dollars.

The inductive spikes coming back from the motor coils kill the fets and they were trying to pinch pennies ya know.

Well they put themselves out of a job and embarrassed the largest most reputable appliance company the world has every known up until that time. Whirlpool bought Maytag a few years ago only to resurrect it from the ashes.

Remember the NEPTUNE? The first front loader washers? Well this is where they failed all of their fets.

There is not a more powerful example in our day that shows the stupidity and arrogance of men and how they could be brainwashed so completely by our universities leaving them incapable of doing their jobs.

The whole electronics business/industry is bent on their own foolishness for money.

However there are still a few great men around in this field that know better than to be sucked into any new designs that abandon common sense.

These companies run the industrial show producing nearly indestructible equipment. Anymore the residential stuff is rated at 2000 watts can only run 400-500watts at a 30 percent duty cycle.

What a mess of red tape and lies when it comes to shifting numbers around.

I hate cheaply done electronics mainly because it stops working duh.

What do ya think boys? Hey John show me a shot of where to add my cap to steady the transistor.

Mike

[JohnnBlade]

Greets Bro,

I like yr style, may i share something here?
Thnx


In my xp i found out that load must be connected to a cap that feeds something when not using it, otherwise it will surge true and burn transistors.

Sofare only my transistors burned all out cause of a surge, and its not the amps

Even if my transistor can handle a socalled 8 amps or 6 forex 2n3055

A voltage or a simple spike can fock it all up, so what i do now if my device is running i am transfering that load to somewhere else, and if i dont distribute that energy i use a dummy load to atleast keep my transistor calm that after loads any transistor.

Can i see some of yr art?


Greets JB

This information is for those who are not engineers yet are electronic masters.

These links help me to refresh why my mosfet's got smoked and how to select the right value. In one of these videos the man shows the difference in spikes from load to load.

If you have an inductor or a battery power can be stored and the reflective reverse polarity kick back will burn out you Fet if the break down voltage is incorrectly used.

For instance in my case I do not want to stop the ringing in my cap dump circuit so when spiking occurs with every pulse a snubber is not used.

Therefore a higher voltage mosfet must be selected than normal electronics would prefer. As an example I have been burning out Mosfet's all week wondering why. The pulsed voltage to an energy storing device in this case a battery was 80vdc and the Mosfet I thought being 200vdc would be plenty.

As it turns out many of the fictitious circuits planted all over the web offer ill advice while lots of other give excellent designs.

My working cap dumps use 500vdc Fets while the 200vdc fet all burnout when pulsed at 80 volts.

If you did not know spiking often generates reverse voltages of three times that of what the power supply if providing.

So remember that there are alot of nice people who are designers on the web who make circuits on paper but never test them in real practices and therefore their circuits must fail.

Just a few basic designer fact can actual weed out which circuits are good ones and which ones may work for awhile but are destined to fail.

Part of the problem with design models and company marketing trade offs is that the sellers have trained all of our engineers to drive a heavy load with as few parts as possible driving down production costs and thereby increasing corporate revenue.

These devices have a shorter lifespan than what I am after and this is why I have chosen to take a look at the industrial design topologies.

Many poorly designed electronic devices exist on the market today for the average residential costumer and are often products that are used only occasionally or intermittently.

With industrial power controls and device a 100 percent duty cycle is more often than not a requirement so entirely opposite approach is taken when circuits are engineered. Or should be.


Don't take any wooden nickles "that's all I am saying"

This "CAP DUMP CIRCUIT" discussion will end in a working circuit other than my beginners design from old scrap parts off of TV sets that are all functioning in the pink.

So far using mosfets rated at 200vdc in the way I want to use them only permit a pulsed voltage of 50-60 volts with spikes of 180vdc.

IRFP460 MOSFET is one answer as it has the right break down voltage rating for a pulsed voltage as high as 100 vdc. This 100 voltages dc is being pulsed to a 36vdc battery bank.

I my studies and experiments using inductive spiking and cap dump pulses to a battery I have discovered from John bedini that for a 12vdc battery I should and have pulsed 30-35vdc to it. Other batteries much large charged better at as high as 40vdc with pulses only 1 per second or less.

For a 36vdc bank this would result in a 120vdc pulse to the larger bank and is not to much to ask (Or shouldn't be) for a mosfet as my other transistor types are doing just fine with these voltages.

It's just the misinfo that ticks me off.

Time to study.

Inductive spiking tutorial - YouTube


Remember that you don't want to eliminate ringing in the battery so snubber circuits are out, snubber diodes or any ring canceling strategies.


Just simple simple turn on turn off figures on the data sheet so if a transistor is rated at 110 amps pulsed and it is the size a a split pea remember that these design criteria more often than not make no practical sense.

Just stick to common sense instead.

Practical MOSFET Tutorial #2 - N Channel, Low Side and the Body Diode - YouTube

Remember pulsed current and drain current are max values and some of that is measured in nano seconds of inrush current while the leg you are soldering to a circuit board is equivalent to a 16awg wire that can only handle 10 amps.

So figure 5 amps per mosfet MAX. If you need a 50amp handling capacity and you want to to last and run cool use 10 mosfets.

If you are driving an RC controlled airplane making it drain a battery in 10 minutes and you need to lite weight then use 2 Mosfets because these circuits are known for only intermittent use and fast burn out.

Not me I want my work to run cool. And when it is time to charge my bank of batteries I don't want to have to put in another row of fets every time like I have done now for the last month.

P-FET Reverse Voltage Polarity Protection Tutorial - YouTube


Preventing Reverse Polarity Damage With Simple Diode Circuits - YouTube

These are my personal views concerning circuit designs and when I am done here this mosfet pulsing circuit will out live me.

Mike

[BroMikey]

Mosfet Failure Study

This information is for those who are not engineers yet are electronic masters.

These links help me to refresh why my mosfet's got smoked and how to select the right value. In one of these videos the man shows the difference in spikes from load to load.

If you have an inductor or a battery power can be stored and the reflective reverse polarity kick back will burn out you Fet if the break down voltage is incorrectly used.

For instance in my case I do not want to stop the ringing in my cap dump circuit so when spiking occurs with every pulse a snubber is not used.

Therefore a higher voltage mosfet must be selected than normal electronics would prefer. As an example I have been burning out Mosfet's all week wondering why. The pulsed voltage to an energy storing device in this case a battery was 80vdc and the Mosfet I thought being 200vdc would be plenty.

As it turns out many of the fictitious circuits planted all over the web offer ill advice while lots of other give excellent designs.

My working cap dumps use 500vdc Fets while the 200vdc fet all burnout when pulsed at 80 volts.

If you did not know spiking often generates reverse voltages of three times that of what the power supply if providing.

So remember that there are alot of nice people who are designers on the web who make circuits on paper but never test them in real practices and therefore their circuits must fail.

Just a few basic designer fact can actual weed out which circuits are good ones and which ones may work for awhile but are destined to fail.

Part of the problem with design models and company marketing trade offs is that the sellers have trained all of our engineers to drive a heavy load with as few parts as possible driving down production costs and thereby increasing corporate revenue.

These devices have a shorter lifespan than what I am after and this is why I have chosen to take a look at the industrial design topologies.

Many poorly designed electronic devices exist on the market today for the average residential costumer and are often products that are used only occasionally or intermittently.

With industrial power controls and device a 100 percent duty cycle is more often than not a requirement so entirely opposite approach is taken when circuits are engineered. Or should be.


Don't take any wooden nickles "that's all I am saying"

This "CAP DUMP CIRCUIT" discussion will end in a working circuit other than my beginners design from old scrap parts off of TV sets that are all functioning in the pink.

So far using mosfets rated at 200vdc in the way I want to use them only permit a pulsed voltage of 50-60 volts with spikes of 180vdc.

IRFP460 MOSFET is one answer as it has the right break down voltage rating(600vdc) for a pulsed voltage as high as 100vdc with 300vdc spikes. This 100 voltages dc is being pulsed to a 36vdc battery bank.

I my studies and experiments using inductive spiking and cap dump pulses to a battery I have discovered from John bedini that for a 12vdc battery I should and have pulsed 30-35vdc to it. Other batteries much large charged better at as high as 40vdc with pulses only 1 per second or less.

For a 36vdc bank this would result in a 120vdc pulse to the larger bank and is not to much to ask (Or shouldn't be) for a mosfet as my other transistor types are doing just fine with these voltages.

It's just the misinfo that ticks me off.

Time to study.

Inductive spiking tutorial - YouTube


Remember that you don't want to eliminate ringing in the battery so snubber circuits are out, snubber diodes or any ring canceling strategies.


Just simple simple turn on turn off figures on the data sheet so if a transistor is rated at 110 amps pulsed and it is the size a a split pea remember that these design criteria more often than not make no practical sense.

Just stick to common sense instead.

Practical MOSFET Tutorial #2 - N Channel, Low Side and the Body Diode - YouTube

Remember pulsed current and drain current are max values and some of that is measured in nano seconds of inrush current while the leg you are soldering to a circuit board is equivalent to a 16awg wire that can only handle 10 amps.

So figure 5 amps per mosfet MAX. If you need a 50amp handling capacity and you want to to last and run cool use 10 mosfets.

If you are driving an RC controlled airplane making it drain a battery in 10 minutes and you need to lite weight then use 2 Mosfets because these circuits are known for only intermittent use and fast burn out.

Not me I want my work to run cool. And when it is time to charge my bank of batteries I don't want to have to put in another row of fets every time like I have done now for the last month.

P-FET Reverse Voltage Polarity Protection Tutorial - YouTube


Preventing Reverse Polarity Damage With Simple Diode Circuits - YouTube

These are my personal views concerning circuit designs and when I am done here this mosfet pulsing circuit will out live me.

Mike

[BroMikey]

Diode

Here is a diagram of what a friend used when he built a driver for 250n Mosfets that power a coil.







I am wondering if these diodes protect the mosfet from spiking? If anyone knows the answer please let me know. The diode number is HFA08TA60CPBF Diode Hexfred 600V

https://www.youtube.com/watch?v=rtsuxU2VxOU

In this video we see mosfets of low power being over driven to a potential 300volt reverse spiking. People seem to over look ratings. a 200volt fet is best suited to work at 50-60 volts under heavy motoring loads.

Mike

[BroMikey]

Getting smarter

Hi Mike,

If you follow the SOA curve of your devices you should not burn them.

I already wrote this, but if you use an opto the best rise/fall times you will be able to see on your Fets is 4-5 microseconds (not nano). It's simply not possible to go faster because it's the limit of your opto (h11d1).
Only using a driver you will be able to switch faster. So I don't know how you're able to see nanosecond switching if you're using an opto.

One reason for toasting fets can also be this:

When you turn the dump OFF very sharply (talking nanoseconds) and you are not dumping down to the battery voltage (even just 20 or 30V above the battery) you will get a negative spike that can be even a few hundred volts in amplitude. It's like your battery is behaving like a coil, you pulse it, and when you let go you get the flyback spike. This is why it's important to watch the dump pulse on a scope and learn all about it while you change parameters and see what happens.

regards,
Mario

Hey Mario
Thanks for the info on back spikes, I had not heard this. However when I first built cap dumps I used 600 v 2sd2499 transistors and still have these units.

I just ordered some 500v IRFP460's for the same dang money as those 200v devices. What a mess of diagrams pasted all over the web

You were right I am looking at the SOA now and unless I am very very careful not to go out of the marked areas I can kiss my fet's bye bye.

The reason I decided to use the 460's is because they are in a standard switching PC supply that operates on a voltages from the line of 120vac, so I figure these parts are rated about where I am running.

See I want to hit 36volt batteries with 90-110 volts DC and like you say with all of the back spikes and then the possible failure on my part to go to high on the duty cycle, I am needing to rethink.

Okay here is what I want. I want a box that you can adjust all around a given set parameter , running up power to 70 percent duty and have the device run cool and never blink.

You said it right in the beginning, you told me to see if my parts can handle this and they can not. A IRFP250N and the 260 break at 200volts.

Then the opto's being slow and sometimes fall times are cut short by me and flash all my parts,humm...

Got the tc4420's and the pile of 594's and need to zero my tantalum caps and high tolerance resistors plus poly caps and will be ready to make something realistic.

I think the way i am playing with times and duty cutting off that it is important that I not run the fine line using these 200v devices.

I have had one dump that can not fail using 600v devices. I use overkill. I used 5 devices 2sd2499 never once had any problems. I didn't even use a fan and let them get hot. Still working great. I use a fan now because it works better.

I see the SOA and a 200v device like they are stop at 50vdc for decent amp usage. The industrial guide at the big semiconductor companies say that for 80v operation and above use 500v devices if you want a box that goes 20years, not 20 minutes like my units last. Now i see why.

I am still going to do it using these under rated stamps, got a bank on already and still drilling the holes and tapping.

Okay the scope is showing 300ns and 500ns and sometimes way less but it is erratic. i think you are right about opto's being a bad choice for gates on a Mosfet. Door bells work fine but this device is running power.

The stuff on the web for beginners is there to make it simple to build and if you want something better you will either need to go to someone who knows how to do it right or learn yourself.

I see opto's used in high speed switching but I don't know how fast the app is, maybe certain set oscillations up in the frec's can settle for 100us.

Can't wait to try this new circuit out. With 12 devices i should do better since the 6 I had didn't. But who knows maybe they just can't handle 80vdc with all of the inrush and back spiking.

75 percent duty is an abuse that these circuits should be able to take or they are under rated in my opinion.

I will have another box that you can beat and it never fails.

This is my goal. Not some cheap version of a device bordering on disaster.

Like you pointed out Mario back spiking and the with the SOA curve hardly anyone is building using and showing the application of an industrial level circuit. Mostly little cheapO timers to a fet and wow wee we are all supos-ta be in heaven cause were are pulsing a battery.

I need more than these toy circuits and thanks to you and few others I am on the right track and I will look at my scope close.

Mike