High voltage power supply help

wish i had a kraken!!! - 10-10-2015 at 14:32

Hi all , I don't know much about electricity , But I need to make a 6300 volts power supply with relatively high Amper , So , I visited "http://danyk.cz/jz2_en.html"
but I want to run my power supply for at least 30 minutes , Could U tell me how I have to calculate characteristics of the ballasts needed for this job?
I want to make a power supply like this :

3*TL = 3 150 W ballasts of sodium lamps

Do U think it will work ?

[Edited on 10-10-2015 by wish i had a kraken!!!]

macckone - 10-10-2015 at 21:47

60 ma at 9000v is plenty dead enough.
If you need higher amperage parallel them and buy more.
If you need really high amperage there are several varieties of multipliers.

aga - 10-10-2015 at 22:35

 Quote: Originally posted by wish i had a kraken!!! I don't know much about electricity , But I need to make a 6300 volts power supply with relatively high Amper

High Voltage + High Current is difficult and lethal if any mistakes are made.

Certainly not a project for a beginner - it's the electrical equivalent of Hydrogen Cyanaide gas.

macckone's suggestion would be the way even a highly experienced experimenter would go.

Sulaiman - 11-10-2015 at 00:03

First, I absolutely agree that your proposed hv psu is potentially lethal
I am a full time electronics engineer and have done some high voltage work
yet for hobby use I treat MOTs with a mixture of caution and fear.
I truly believe that a single MOT can easily electrocute anyone with just one slip of attention.
Unfortunately safer high voltage sources are few, and expensive.

For members here I can recommend for small high voltage dc experiments
you can get a 3v to (5 to 7) kV inverter via eBay for pocket money.
or dis-assemble a cheap tennis racket style bug-zapper
(buy 2 or 3 as they can be killed

An MOT can be run continuously (or at least one hour) with no ballasting,
and the internal magnetic shunts do limit short circuit current enough to prevent fire, but not electrocution.

With normal domestic electrical supplies there is a good chance that nine MOT primaries in parallel will blow breakers/fuses occasionally
(depending mainly upon where in the mains cycle you switch on)

last but not least, what do you need kilowatts at kilovolts for?

WGTR - 11-10-2015 at 00:16

Have you heard the expression in english, "If something doesn't kill you, it makes you stronger"? Well, this potentially falls into the kill you category. Now that we have that unpleasant disclaimer out of the way, how much current does the supply need to deliver? What are you trying to power with it? I need to know the application, and the power that is needed.

I don't understand why the ballasts are connected as shown in the schematic. It looks like they are all in series, with the various parallel transformers tapped at different potentials. At first I thought it was an attempt to introduce a voltage droop at the output, but I think this is probably a misprint. Do you have the link to the page that shows the schematic, so that I can look through it?

Selecting the size of a ballast inductor involves a simple use of Ohm's Law. If a custom ballast is made, it is also necessary to have a meter that measures inductance. Do you have such a meter?

A standard neon sign transformer performs the same functions as this circuit, only better. Current limiting is provided by the magnetic shunt installed between the primary and secondary windings. The shunt partially decouples the primary and secondary, introducing the leakage inductance that limits the current.

Honestly, the whole idea scares me. The schematic design looks haphazard, and uses parts that are not really designed for the task. Most of the transformers have 2-4kV potential between the core and earth ground, and this is a dangerous practice. It's difficult to say that a particular MOT will have the necessary isolation to handle 6kV without knowing the part # for the transformer itself.

Building your own transformer is an option, although it's not a one-weekend project. Figure on learning a lot about magnetics if this approach is taken, and spending the next several months, incrementally building up the unit to something that works properly without zapping you or itself. It will cost more money to design and build a proper custom transformer than just buying an appropriately-sized neon sign transformer. At the same time, sometimes the learning experience is worth the extra cost.

Whatever you do, take a few steps back and approach this slowly and deliberately. If you don't yet have the skill in electronics to understand how this works, I'd really suggest sticking with safer projects for now. Also, never work alone with high voltages. I had a friend-of-a-friend who worked on an old vacuum tube radio transmitter late one night, by himself. They found him slumped over the unit the next morning, still popping and sizzling from the high voltage power supply that was still operating.

Have fun, and be safe.

macckone - 11-10-2015 at 08:31

Some additional notes. The circuits are unbalanced in the drawing provided but not the original.
The grounding as shown won't work.

wish i had a kraken!!! - 11-10-2015 at 10:53

 Quote: Originally posted by macckone Buy a neon sign transformer and forget about it. 60 ma at 9000v is plenty dead enough. If you need higher amperage parallel them and buy more. If you need really high amperage there are several varieties of multipliers.

NSTs are more expensive in my place! while they provide 30 mA (in my place 60 mA are rare )

I need a PSU capable of providing 1500 mA ~ 2000 mA

wish i had a kraken!!! - 11-10-2015 at 11:07

 Quote: Originally posted by Sulaiman yet for hobby use I treat MOTs with a mixture of caution and fear. I truly believe that a single MOT can easily electrocute anyone with just one slip of attention. An MOT can be run continuously (or at least one hour) with no ballasting, and the internal magnetic shunts do limit short circuit current enough to prevent fire, but not electrocution. With normal domestic electrical supplies there is a good chance that nine MOT primaries in parallel will blow breakers/fuses occasionally (depending mainly upon where in the mains cycle you switch on) last but not least, what do you need kilowatts at kilovolts for?

1. I also treat MOTs with mixture of fear and caution
(If U see I have strange smile on my face it's because I wanted to extract a photo from the video with my smile + a Huge ARC )

2.I cannot run My PSU for a long time !

3. I need 9 Kilowatts 6.3 KiloVolts for modifying my Machine
U can see my machine hear :

Sulaiman - 11-10-2015 at 11:21

If you insist on going ahead then here are a few of my thoughts;

1) 6300 V @ 1.5 A to 2.0 A = 9.45 kW to 12.6 kW
= 41 A to 55 A for 230 Vac input and 100% efficiency.
i.e. you need a 60 A 230 Vac supply, at least.

2) IF you have suitable power available,
For circuits like this I find it easier to draw the circuit diagram in parts,
. Power input, fuses, breakers, trips, safety barriers, filters, switch etc.
. All of the primaries in parallel
. All of the secondaries in some arrangement plus output terminals.

At this voltage and power level please pay particular attention to the first of the three.

P.S. you'd probably need an arc-flash 'spacesuit' to work within EU regulations !

[Edited on 11-10-2015 by Sulaiman]

wish i had a kraken!!! - 11-10-2015 at 11:29

 Quote: Originally posted by WGTR Have you heard the expression in english, "If something doesn't kill you, it makes you stronger"? Well, this potentially falls into the kill you category. Now that we have that unpleasant disclaimer out of the way, how much current does the supply need to deliver? What are you trying to power with it? I need to know the application, and the power that is needed.

Dear WGTR , I believe Safety is the first priority :-)
I need 1500 mA ~ 2000 mA for my supply
I want to modify my Machine
U can see it here :

 Quote: Originally posted by WGTR I don't understand why the ballasts are connected as shown in the schematic. It looks like they are all in series, with the various parallel transformers tapped at different potentials. At first I thought it was an attempt to introduce a voltage droop at the output, but I think this is probably a misprint. Do you have the link to the page that shows the schematic, so that I can look through it?

The schematic is misprinted I'll send U new one .
The whole Unit is consist of four smaller units paralleled to each other in order to increase the output Amp.

 Quote: Originally posted by WGTR Honestly, the whole idea scares me. The schematic design looks haphazard, and uses parts that are not really designed for the task. Most of the transformers have 2-4kV potential between the core and earth ground, and this is a dangerous practice. It's difficult to say that a particular MOT will have the necessary isolation to handle 6kV without knowing the part # for the transformer itself.

I thought of it but is there a way to know that ? Beside knowing the part number ?

 Quote: Originally posted by WGTR Building your own transformer is an option, although it's not a one-weekend project. Figure on learning a lot about magnetics if this approach is taken, and spending the next several months, incrementally building up the unit to something that works properly without zapping you or itself. It will cost more money to design and build a proper custom transformer than just buying an appropriately-sized neon sign transformer. At the same time, sometimes the learning experience is worth the extra cost.

I also wish I knew how to build a custom transformer for myself , but The thing is that I read somewhere that the calculations for transformers greater than 3.5 KiloWatt is entirely different . and I don't have access to good sources on this issue. i.e , Once I wanted to design a bombarder transformer (as U know bombarder transformers are used in neon sign industries ) , I couldn't find any good practical sources for that.
I would really appreciate if U help me with that(guide me to the sources).

 Quote: Originally posted by WGTR Whatever you do, take a few steps back and approach this slowly and deliberately. If you don't yet have the skill in electronics to understand how this works, I'd really suggest sticking with safer projects for now. Also, never work alone with high voltages. I had a friend-of-a-friend who worked on an old vacuum tube radio transmitter late one night, by himself. They found him slumped over the unit the next morning, still popping and sizzling from the high voltage power supply that was still operating. Have fun, and be safe.

Sure :-)

[Edited on 11-10-2015 by wish i had a kraken!!!]

wish i had a kraken!!! - 11-10-2015 at 11:44

 Quote: Originally posted by Sulaiman If you insist on going ahead then here are a few of my thoughts; 1) 6300 V @ 1.5 A to 2.0 A = 9.45 kW to 12.6 kW = 41 A to 55 A for 230 Vac input and 100% efficiency. i.e. you need a 60 A 230 Vac supply, at least. 2) IF you have suitable power available, For circuits like this I find it easier to draw the circuit diagram in parts, . Power input, fuses, breakers, trips, safety barriers, filters, switch etc. . All of the primaries in parallel . All of the secondaries in some arrangement plus output terminals. At this voltage and power level please pay particular attention to the first of the three. P.S. you'd probably need an arc-flash 'spacesuit' to work within EU regulations ! [Edited on 11-10-2015 by Sulaiman]

I have a single Phase 10KiloWatt variac (a big one) 40 Amp at 250 Volts.

[Edited on 11-10-2015 by wish i had a kraken!!!]

WGTR - 11-10-2015 at 14:13

 Quote: Originally posted by wish i had a kraken Dear WGTR , I believe Safety is the first priority :-) I need 1500 mA ~ 2000 mA for my supply I want to modify my Machine U can see it here : https://www.youtube.com/watch?v=AGiX8Rh-GGI

Very neat setup! My compliments on a very interesting project and video.

I see now what you're trying to do. At first I thought you were trying to make a huge Jacob's ladder . How much amperage are you providing to the target with your existing set-up? You may have to scale up your water cooling system quite a bit.

If you want to use simple current limiting using a ballast, the voltage from the power supply must be much higher than you really need to maintain a glow discharge in the sputter coater. It looks like commercial units deliver only about 1000V, but they do this because they can provide the rated output current without allowing the voltage to droop.

http://www.cpii.com/product.cfm/7/38/193

A sputtering chamber operates with a "glow" discharge; if the current increases too much the glow transitions to an arc, which can damage the target and the substrate. It looks like commercial units get around this by using a separate arc suppression unit in conjunction with the high power supply. The power supply could deliver 10s or 100s of times the rated current if it was shorted, but when the arc suppression system detects the beginnings of a high-current arc, it interrupts the output current long enough to re-stabilize the glow discharge.

http://www.cpii.com/product.cfm/7/38/192

Unfortunately, this size of a power supply is beyond my abilities. Someone like IrC may have some further input, as he has worked with large transformers and high voltages before.

IrC - 11-10-2015 at 16:12

My input would be the only smart way to such voltage at 2 amps is using a pole transformer or at the minimum a damn big plate transformer (if his local AM station goes off air mysteriously we know why). Also I do not see the need for such currents for the sputter coating setup as in that video. WGTR stated things just fine to me, personally I do not see why an NST would not work fine since a low current plasma is enough unless one is scaling up to coat something very large. I did not watch all the vid nor go searching and studying all parameters that would lead me to believe a 7 to 10 KVA supply is required. Perhaps the OP can provide some links to technical aspects one could study to get an idea of just exactly how much current is needed. One mistake with a supply providing 3.6 KV at 2 amperes is the stuff of nightmares. Even if a high power magnetron is used in the design it does not seem such a supply is needed, although in such case obviously the NST idea would not work. Also great care would be required in shielding the RF in such a design unless cooking ones own head was a goal. Make up a set of detailed plans for what precisely you are wanting to build and post them instead of merely having us watch random vids.

OK I took the time to watch all of your vid and it left me not understanding your statement "I need 9 Kilowatts 6.3 KiloVolts for modifying my Machine". From the vid it is clearly a glow plasma setup and I cannot fathom why you need what you say. An NST is all you need for such a setup. In those pages on the site you linked powerful arcs are the goal but this has nothing to do with sputtering. I was looking at some of the machines using a magnetron and thought maybe you were going to 'modify' (you never specified what mods) it by going from a glow discharge to microwaves. However even then around 2 KV at 1 or 2 amperes maximum is enough for even a 2 or 3 KW magnetron. You need to explain or somehow justify '9 Kilowatts 6.3 KiloVolts' for your questions to make sense to me.

[Edited on 10-12-2015 by IrC]

smaerd - 12-10-2015 at 12:45

Like IrC I do not see the need for such a high voltage.

At work I sputter coat using a 100W DC supply at 30-80mA (depending on the metal and grain size). For most common uses 80mA suffices which gives what a voltage of 1.25K?

Lower potential and lower currents mostly translates to longer sputtering time unless grain size is super important for your home work. In the interest of cost and saftey I'd err on going low and slow. While fixing the DC magnetron sputterer at work I saw a corona discharge. It discharged from a bad power inlet to the chamber and onto the chassis before the PSU turned off(it took maybe 5 seconds before the arc disappeared and my awestruck dumbness flipped the switch). I realized that had I of not moved my hand off of the metal chamber it could have arced directly through my chest and I may have died. This stuff is not a joke.

[Edited on 12-10-2015 by smaerd]

Just watched your video - very professional work! I really like your set-up! Your biggest flaw is probably your gasket sealing of the material. Then again it probably works fine for whatever you use it for at home.

[Edited on 12-10-2015 by smaerd]

wish i had a kraken!!! - 12-10-2015 at 13:06

 Quote: Originally posted by smaerd Like IrC I do not see the need for such a high voltage. At work I sputter coat using a 100W DC supply at 30-80mA (depending on the metal and grain size). For most common uses 80mA suffices which gives what a voltage of 1.25K? Lower potential and lower currents mostly translates to longer sputtering time unless grain size is super important for your home work. In the interest of cost and saftey I'd err on going low and slow. I've seen a corona discharge(discharged from a bad power inlet to the chamber onto the chamber and chassis before the PSU turned off) while fixing the DC magnetron sputterer at work. I realized that had I of not moved my hand off of the metal chamber it could have arced directly through my chest and I may have died. This stuff is not a joke. [Edited on 12-10-2015 by smaerd]

Dear Smaerd , Tanks for your reply & yes you and Irc are both right , yet I want to achieve hot Plasma and maintain it for a long time (let's say 30 minutes) , Beside that I have other Plans for such a PSU (I want to power several magnetrons with it , want to use it for charging some huge high Farad 8000 Volt Capacitors , etc ....)

[Edited on 12-10-2015 by wish i had a kraken!!!]

smaerd - 12-10-2015 at 13:16

If there is any problems with your coatings my opinion would be to first check for leaks in your system. I can already see where a few might be. The other concern is the lower limit of your vacuum system. It's pretty recommended to pump down to 1E-6 torr, then flood with argon (maybe to 1E-3) then sputter. This removes the majority of oxygen, some of the moisture, and other things that off-gas, etc.

Sustaining a plasma for 30 minutes!? You may need to cool your sputter-head in order to do that. Which could be quite dangerous. The polymer container is not ideal for that kind of work-load?

Anyways, keep up the good work, your sputter station is very nice.

wish i had a kraken!!! - 12-10-2015 at 13:42

 Quote: Originally posted by smaerd If there is any problems with your coatings my opinion would be to first check for leaks in your system. I can already see where a few might be. The other concern is the lower limit of your vacuum system. It's pretty recommended to pump down to 1E-6 torr, then flood with argon (maybe to 1E-3) then sputter. This removes the majority of oxygen, some of the moisture, and other things that off-gas, etc. Sustaining a plasma for 30 minutes!? You may need to cool your sputter-head in order to do that. Which could be quite dangerous. The polymer container is not ideal for that kind of work-load? Anyways, keep up the good work, your sputter station is very nice.

Tanx , I know if I am about to sustain Plasma for 30 minutes I have to use other materials or better cooling systems , But U know , It would be part of my personal project In which I want to study the effect of time and Higher voltages on adhesiveness of coated layer on substrate(i.e. I want to coat TiN on steel bar in Nitrogen vacuumed chamber (I did it before the results were awful ) So , my guess is that with higher voltage which helps Ions to accelerate more and higher Amp which will eventually cause the substrate to get Hot I will help atoms of the interface to diffuse more and more ) .

And as I mentioned I may use such power supply for future works .
(I know from view point of engineering it seems ridiculous to use such a PSU for just carrying out simple sputtering task , Yet I feel Having such PSU might be very handy for my future works).

I'm Happy that U liked my sputter station :-)

[Edited on 12-10-2015 by wish i had a kraken!!!]

Marvin - 12-10-2015 at 13:54

Wow. Were you on acid when you designed that?

I really hope you build that exactly. It's the only way I'll get to know if it catches fire or not.

WGTR - 12-10-2015 at 18:19

After thinking about this some more, 10kW at 50Hz is going to be a big hunk of transformer iron, whether it's in one big transformer, or a collection of smaller ones. I think your idea of using a multitude of paralleled microwave oven transformers could possibly work, as long as the voltages in each parallel string were equal. If they were unequal, it would give the same effect as shorted turns in a transformer. It's not optimal, however, due to the low frequency of the power.

To test the high voltage isolation of the transformers a Hipot tester can be used. The desired withstand voltage is applied across the winding and core (or whatever else you want to test), and the leakage current is measured. If there is no significant leakage, then the part can be used at this voltage. It's a good idea to test the part at full operating temperature and other conditions (vibration, moisture, etc). I'm over-simplifying this a bit.

https://en.wikipedia.org/wiki/Dielectric_withstand_test

If I were the one building such a large power supply, I would use a high frequency switching power supply, instead of a large power transformer. It would save money and size. To save effort, I might modify the output transformer in a used inverter TIG welder to supply the voltage that I wanted. Then, I'd modify the feedback loop so that the output current would be adjustable. This would not be a trivial project, though.

Here's an interesting idea that is similar:

http://danyk.cz/igbt_1_en.html

It's ground isn't isolated from the power, though.

[Edited on 10-13-2015 by WGTR]

IrC - 12-10-2015 at 18:32

kracken it appears to me you have never been or been around someone seriously injured toying with these levels of energy. I could be wrong but wanting near 4KV at 2 amperes for a vague wide variety of dangerous projects including charging a 'high farad' 8KV capacitor bank for whatever project leads me to believe I am correct. I spent some time looking through pages in the site you linked and the author is very skilled in electronics. However going through his magnetron pages and watching the vids leads me to think he has the common sense of a lemming so be very careful in what you choose to replicate. For example capacitors in the bank you described charged as you are saying can vaporize themselves with tremendously destructive violence. All I can say is be very careful. If you seriously need such energy levels I would get a 14.4 KV 10KVA pole pig and use a variable core arc welder as an adjustable series reactor instead of trying to combine ever increasing numbers of MOT's. Just a thought.

wish i had a kraken!!! - 20-10-2015 at 12:20

Hello guys I'm back again :-) ,
Behold I present U some of my capacitors & MOTs
Dear IrC , Do U think this type of capacitors would evaporate ? (I mean do U think they can withstand 6300 Volts , 1.5 Amp ? )
each of these capacitors can tolerate 1200 Volts , the capacitance of each of these capacitors is ~ 12.5 Micro Farad , So , I want to connect six of them in series together so I guess this way they can tolerate 6300 Volts , though their capacitance would drop to ~ 2 Micro Farad.

[Edited on 20-10-2015 by wish i had a kraken!!!]

[Edited on 20-10-2015 by wish i had a kraken!!!]

wish i had a kraken!!! - 21-10-2015 at 12:30

 Quote: Originally posted by macckone Some additional notes. The circuits are unbalanced in the drawing provided but not the original. The grounding as shown won't work.

Tanx I'll try to fix that :-)

aga - 21-10-2015 at 12:36

With any high power circuit you will need to add in some Excess to relieve some stress on the individual components.

Maximum Ratings generally indicate what they can stand for a Short time, whereas Working Voltage / Current Ratings indicate what they can stand for a Long time.

Add another capacitor or two or one might fail in some way.

wish i had a kraken!!! - 21-10-2015 at 12:47

 Quote: Originally posted by smaerd Like IrC I do not see the need for such a high voltage.

U and IrC are both right , yet I still believe having such a high voltage would be handy for some other projects.

 Quote: Originally posted by smaerd At work I sputter coat using a 100W DC supply at 30-80mA (depending on the metal and grain size). For most common uses 80mA suffices which gives what a voltage of 1.25K?

Correct , How much is the thickness of your coatings Do U know ? and may I know what type of metals U coat?

 Quote: Originally posted by smaerd Lower potential and lower currents mostly translates to longer sputtering time unless grain size is super important for your home work. In the interest of cost and saftey I'd err on going low and slow. While fixing the DC magnetron sputterer at work I saw a corona discharge. It discharged from a bad power inlet to the chamber and onto the chassis before the PSU turned off(it took maybe 5 seconds before the arc disappeared and my awestruck dumbness flipped the switch). I realized that had I of not moved my hand off of the metal chamber it could have arced directly through my chest and I may have died. This stuff is not a joke.

Oh WOW , I'm glad that U are safe :-) , U use large DC-Magnetron sputter device at work ? My guess is U making coated glasses such as windshields or similar products ;-)

 Quote: Originally posted by smaerd Just watched your video - very professional work! I really like your set-up! Your biggest flaw is probably your gasket sealing of the material. Then again it probably works fine for whatever you use it for at home.

Tanx, I'm glad that U liked my setup ! better sealing is a solution indeed , Yet I really think I'd love to have such a power supply.

wish i had a kraken!!! - 21-10-2015 at 12:50

 Quote: Originally posted by aga With any high power circuit you will need to add in some Excess to relieve some stress on the individual components. Maximum Ratings generally indicate what they can stand for a Short time, whereas Working Voltage / Current Ratings indicate what they can stand for a Long time. Add another capacitor or two or one might fail in some way.

I intend to use 6 of those capacitors (1200 12.5 Micro Farad ) in series .

aga - 21-10-2015 at 13:02

Use 7 and recalculate for the reduced capacitance.

Voltage tolerance of the stack just increased by 1200V.

wish i had a kraken!!! - 21-10-2015 at 13:02

 Quote: Originally posted by IrC kracken it appears to me you have never been or been around someone seriously injured toying with these levels of energy. I could be wrong but wanting near 4KV at 2 amperes for a vague wide variety of dangerous projects including charging a 'high farad' 8KV capacitor bank for whatever project leads me to believe I am correct.

Tanx Dear IrC, I don't wanna toy with it, I just want to have such a power supply and I am here asking U (GODS of electronics) cuz I want to discuss my plans for it with U guys before taking any step/s.

 Quote: Originally posted by IrC I spent some time looking through pages in the site you linked and the author is very skilled in electronics. However going through his magnetron pages and watching the vids leads me to think he has the common sense of a lemming so be very careful in what you choose to replicate. For example capacitors in the bank you described charged as you are saying can vaporize themselves with tremendously destructive violence. All I can say is be very careful. If you seriously need such energy levels I would get a 14.4 KV 10KVA pole pig and use a variable core arc welder as an adjustable series reactor instead of trying to combine ever increasing numbers of MOT's. Just a thought.

Your idea is good but U know, I do Have 12 MOTs (900 Watts) and 6 of those Capacitors, and I don't have access to pole pig and a variable core arc welder ! that's why I intend to use them.

wish i had a kraken!!! - 21-10-2015 at 13:08

 Quote: Originally posted by aga Use 7 and recalculate for the reduced capacitance. Voltage tolerance of the stack just increased by 1200V.

Ok , Tanx

But could U explain more why 7 ?
I think 6 * 1200V = 7200 V , which is at least 900 Volts higher than the out put voltage of power supply . Could U explain me in more details why 7 is better to be used ? I mean would U tell me why U think 6 is not sufficient ?

[Edited on 21-10-2015 by wish i had a kraken!!!]

MrHomeScientist - 21-10-2015 at 13:16

Shouldn't you have the poles of those capacitors connected when not in use, so they don't accumulate charge from static electricity?
IrC - 21-10-2015 at 15:27

I figured you were going MOT because as you say pole pigs are next to impossible to source unless your pockets are deep. Post a proposed schematic. I assume you intend to rectify the voltage and are using those caps in series to store a high DC voltage? I do not see any diodes in your pics. Obviously peak voltage is what you need to consider as to the total rating of the string. Your information is confusing and sparse I need to see the actual circuit you intend to build.

WGTR - 21-10-2015 at 15:43

I have a headache today, so I hope this post comes out sounding coherent. The problem that myself (and maybe IrC) have is that we know what kinds of things can go wrong when dealing with high voltage and high energy. It's difficult to cover every possible safety issue with someone who admits not knowing much about electronics just yet. We may overlook something important, or something we say might be misunderstood, and we might end up instructing you to do something that would end up killing or injuring you. Even if we wouldn't be legally liable, it falls under a moral responsibility to not suggest things that could be injurious to someone who's inexperienced in the field. That's why you're getting more safety advise than actual answers to your questions.

Anyway, it's not such a simple matter to connect capacitors in series. You first have to measure the capacitance of each cell, to make sure that they are matched (especially for the type that you have). If they aren't, then they will divide the voltage unevenly, and some capacitors may exceed their voltage rating, which can cause the entire string to fail. If they are within their tolerances, then doing as aga suggests would probably work. You add additional cells in series to give you some more voltage headroom, just in case one or more of them is out of tolerance. Then if one of them shorts out, the rest of the capacitors usually don't automatically fail. These oil-filled capacitors often change capacitance and fail, however.

Also, the capacitors must be mounted in a specific orientation. I don't remember which one off the top of my head, but it may tell you on the side of the can. They are oil-cooled internally, but the manufacturer only puts enough oil in to cover the foil, if it is mounted a certain way. Otherwise parts of the foil are not submerged, and those areas can overheat/melt during use.

This isn't as important for AC use, but for DC use it is important to consider the leakage current of each capacitor. It is generally low, but over time it can cause the voltages across series capacitor strings to become unbalanced. One way to counteract this is to add external resistors across each capacitor, such that the resistor current is at least 10 times greater than the worst leakage of any capacitor in the string.

IrC - 21-10-2015 at 16:22

All that, and it keeps nagging at me from all the posts I have seen I can't quite figure out if he intends to put that cap string across an AC high voltage. Hope not.

wish i had a kraken!!! - 12-11-2015 at 02:23

 Quote: Originally posted by IrC . Post a proposed schematic.Your information is confusing and sparse I need to see the actual circuit you intend to build.

Dear IrC , here is my proposed schematic , I want to build a multi functional
HV Powersupply . there u would see two slightly different schematics , are they the same ? which one would have better operation at the same cost ?
Plz guide me (I don't want to have my MOTs getting hot when I use them for at least 30 minutes so I need to know how to calculate magnitude of current limiter and to put it where to achive what I want )

I am supper sorry for my late reply, I was supper busy with something.
I will let U know about them soon :-)

 Quote: Originally posted by IrC I assume you intend to rectify the voltage and are using those caps in series to store a high DC voltage? I do not see any diodes in your pics. Obviously peak voltage is what you need to consider as to the total rating of the string.

U are 100% right , there is no diodes , I have separated rectifying unit from the Power supply, The intention for using capacitors is to provide more amps So, some huge arcs would be appear when I use the power supply for my Jacob ladder (as I said already , I intend to have multi functional PS ).

[Edited on 12-11-2015 by wish i had a kraken!!!]

[Edited on 12-11-2015 by wish i had a kraken!!!]

wish i had a kraken!!! - 12-11-2015 at 02:58

 Quote: Originally posted by IrC All that, and it keeps nagging at me from all the posts I have seen I can't quite figure out if he intends to put that cap string across an AC high voltage. Hope not.

What is wrong with adding cap string across an AC high voltage ?
(PLZ visit : http://uzzors2k.4hv.org/index.php?page=resonantmots , I used this circuit to make my PS)

But mine is slightly different :

wish i had a kraken!!! - 12-11-2015 at 05:38

 Quote: Originally posted by WGTR Anyway, it's not such a simple matter to connect capacitors in series. You first have to measure the capacitance of each cell, to make sure that they are matched (especially for the type that you have). If they aren't, then they will divide the voltage unevenly, and some capacitors may exceed their voltage rating, which can cause the entire string to fail. If they are within their tolerances, then doing as aga suggests would probably work. You add additional cells in series to give you some more voltage headroom, just in case one or more of them is out of tolerance. Then if one of them shorts out, the rest of the capacitors usually don't automatically fail. These oil-filled capacitors often change capacitance and fail, however. Also, the capacitors must be mounted in a specific orientation. I don't remember which one off the top of my head, but it may tell you on the side of the can. They are oil-cooled internally, but the manufacturer only puts enough oil in to cover the foil, if it is mounted a certain way. Otherwise parts of the foil are not submerged, and those areas can overheat/melt during use. This isn't as important for AC use, but for DC use it is important to consider the leakage current of each capacitor. It is generally low, but over time it can cause the voltages across series capacitor strings to become unbalanced. One way to counteract this is to add external resistors across each capacitor, such that the resistor current is at least 10 times greater than the worst leakage of any capacitor in the string.

Thanks Dear WTGR, I measured the capacitance (~12.5 MicroFarad each slightly different , Like 12.3 or 12.6 Is it going to make trouble ?)
And I want to connect 8 of them in series for AC use.

IrC - 12-11-2015 at 20:29

Quote: Originally posted by wish i had a kraken!!!
 Quote: Originally posted by IrC All that, and it keeps nagging at me from all the posts I have seen I can't quite figure out if he intends to put that cap string across an AC high voltage. Hope not.

What is wrong with adding cap string across an AC high voltage ?
(PLZ visit : http://uzzors2k.4hv.org/index.php?page=resonantmots , I used this circuit to make my PS)

But mine is slightly different :

I cannot see the image in the post above the one quoted here. In this circuit (in quoted post) the caps are not across the high voltage they are in series with it. To say 'across' is to say in parallel. When you are using capacitors of the proper value to resonate at the line frequency with the transformer inductance you are in effect altering the power factor.

wish i had a kraken!!! - 12-11-2015 at 22:40

 Quote: Originally posted by IrC I cannot see the image in the post above the one quoted here.

Dear IrC, did U click on the images U can't see ?
I believe if U do click on them your browser will open a new page in which U can see them.

Sulaiman - 12-11-2015 at 23:11

Ignoring the chokes in your first diagram,
the main difference in the second diagram is that the mot cores and secondaries have been drawn as 'floating'/isolated,
in the first diagram you can see the connection between the inner part of the secondary winding and the core.
(having the inner part of the winding at the same potential as the core makes electrical insulation easier/cheaper)

look at the first diagram and imagine
. what voltage is each core at
. what simple tricks can be used to reduce insulation problems

off to work ....

wish i had a kraken!!! - 13-11-2015 at 00:05

Dear Sulaiman , My intention is to use this PS unit for at least 30 minutes without having MOTs getting hot , So , the purpose of the chokes is to act like a current limiter and prevent MOTs from getting over heated .
So U suggesting me to put the chokes in location number 2?

Sulaiman - 13-11-2015 at 00:14

No, I am referring to the physical construction of an MOT being represented in the first diagram but not in the second.
For cost and safety reasons, the core of an MOT is normally connected to chassis earth and the inner connection to the secondary winding is connected to the core.
Have a look at your transformers.

IrC - 13-11-2015 at 00:14

Quote: Originally posted by wish i had a kraken!!!
 Quote: Originally posted by IrC I cannot see the image in the post above the one quoted here.

Dear IrC, did U click on the images U can't see ?
I believe if U do click on them your browser will open a new page in which U can see them.

Obviously I know how to click on an image link. It is far too large and png displays as black traces on a near black background on my browser.

wish i had a kraken!!! - 13-11-2015 at 00:55

I am so sorry, I'll fix it soon

[Edited on 13-11-2015 by wish i had a kraken!!!]

IrC - 13-11-2015 at 02:38

Your circuit is not 'slightly different' than the link. It is not even close, I do not see how your design will work. Also I do not believe the insulation resistance between the secondary and core is adequate. Running at design voltage each secondary (the ones even wired so as to work) will have a peak voltage of 2,800 volts. I do not see how any house wall outlet is going to provide around 11 KVA for one minute let alone 30 minutes. All I see is a member working for at minimum a near death experience. I do not recall in this thread a diagram of just exactly what this supply will be connected to.

wish i had a kraken!!! - 13-11-2015 at 04:36

 Quote: Originally posted by IrC Your circuit is not 'slightly different' than the link. It is not even close.

Dear IrC , I meant my actual PS (the one that I already built) is using 9 HV capacitors , insted of 4 .(" http://uzzors2k.4hv.org/index.php?page=resonantmots "the guy used 4 capacitors in his work (total capacitance = 1 MicroFarad) , but i feared maybe there would be some problems, so I added 5 more caps to have a total capacitance of 1 MicroFarad )

 Quote: Originally posted by IrC I do not see how your design will work. Also I do not believe the insulation resistance between the secondary and core is adequate.

In my diagram I connected 3 MOTs in series So, I hope the outlet voltage will be 6300 Volts , then I connected 4 of these in parallel so , I hope the outlet Amp will be 4 times bigger.
I just don't know if it is as simple as I think , I hope U help me with that .
I don't know how to become sure that in my design MOTs would be in phase?

And why U think the insulation is not adequate cuz I have seen successful attempts of connecting 3 MOTs in series:

 Quote: Originally posted by IrC Running at design voltage each secondary (the ones even wired so as to work) will have a peak voltage of 2,800 volts. I do not see how any house wall outlet is going to provide around 11 KVA for one minute let alone 30 minutes.

Dear IrC, PLZ explain this part more .

 Quote: Originally posted by IrC I do not recall in this thread a diagram of just exactly what this supply will be connected to.

As I told before It would be a multi functional PS . I want to use it in my investigation of the effect of higher voltages on sputtering process , also investigate the time and other parameters.

[Edited on 13-11-2015 by wish i had a kraken!!!]

IrC - 13-11-2015 at 13:18

"Dear IrC, PLZ explain this part more ."

I am not trying to be difficult here but this question really bothers me. Helping someone build and experiment with such a potentially lethal circuit when it is clear they lack basic understanding of the theory involved does not bode well. An analogy would be one of SCM's energetic gurus helping someone make NCl3 where the reaction is being carried out in a paint shaker. Failure to understand concepts as simple as my 'nearly 11 KW' statement is troublesome to me. 12 MOT's each capable of 900 watts (0.9 KW) yields a total of 12 x 0.9 = 10,800 watts or 10.8 KW. The rated value of each (900 watts) is that which can operate for several minutes nonstop cooking food without overheat melting the primary enameled turns causing a short circuit. In reality during an arc with very high loading on the output far greater than 900 watts can be seen by each MOT so the 10,800 watts could actually be greater. This is far in excess of any normal house wiring. You would have to run your own line to the service box with say a dual 50 amp breaker on the 240 volt circuit. Also if you carefully study the schematic you provided your primaries are not wired the same as any of the vids you linked. In each set of three you only power the center one. If one were to assume you meant dots instead of crossovers then the phasing would be a disaster. As to not knowing the 2800 peak volts I mentioned, magnetrons are the same here or in the Eu so whether or not the MOT primary was designed for 120 or 240 volts the output would be the same or very similar. Study the circuit theory of a typical microwave oven in this link.

http://www.microtechfactoryservice.com/doubler.html

"The voltage across the capacitor will rise to the transformer secondary voltage to the maximum 2800 volts"

This 2800 volts is the peak voltage seen on the secondary winding at the proper moment in time during the cycle, three in series assuming phasing was correct will yield a peak voltage of 8400 volts for the series string. All MOT's are designed and the insulation resistance is considered with the end of the innermost winding being connected to the core, and the oven frame to the green ground wire in the power cord. Which is connected to the building neutral. You will have a situation where the voltage will try to arc to the core, and the insulation of the primary side is not designed to withstand this voltage between the core and the primary winding. Also for sputtering you want high voltage at moderate current for long times whereas the high leakage MOT is designed for very high current for a relatively short time interval. While I am not saying the idea will not work (assuming you corrected the schematic you provided concerning primary wiring with due consideration of phasing), it is a dangerous poorly thought out design suitable for short duration cool looking arcs but not long term operation such as you would need for sputtering. I suggest you design two supplies, one for sputtering and one for playing with short duration arcs.

Marvin - 15-11-2015 at 13:40

 Quote: Originally posted by IrC In each set of three you only power the center one. If one were to assume you meant dots instead of crossovers then the phasing would be a disaster.

I made the same mistake the first time I looked at the new one, and then I made the same assumption. Actually the phasing is fine, the schematic contains 4 way ties.

I'm finding this one much less funny, I see a plain circuit that looks functional and the same total lack of understanding. This is two orders of magnitude greater than anyone really needs for sputtering from what I've read. It's a mistake that could easily kill the first person that tries to use it and the first person that tries to save them, which could be a family member or someone from emergency response.

wish i had a kraken!!! - 16-11-2015 at 11:37

 Quote: Originally posted by Marvin I see a plain circuit that looks functional and the same total lack of understanding.

Dear Marvin , U believe my circuit will work?
Can I draw Arc from this PS for long duration from it ? (30 minutes at least )
IrC made some points which are noticeable and I think I have to change the circuit in order to achive what I want , But I hope I don't have to leave it.
I don't say I am going to use it only for sputtering , I want to use it in especial sputtering (I want to achieve higher temperatures and investigate the effects of it with respect to time on deposited layer.) & I also want to use this PS on carbon nanotube/Fulleren production in underwater AC electric arcing (Some have done this using 40 Volt at approximately 100 Amps but I want to use 6300 Volts and see the difference myself I mean do U think it would be possible to achive longer arcs under water ? Or it wont work ? )

[Edited on 16-11-2015 by wish i had a kraken!!!]

[Edited on 16-11-2015 by wish i had a kraken!!!]