Facebook
Google
GitHub
Linkedin
Can any one provide circuit to make high current up to 100mA 400KV power supply for and experiment?
Mod edit: moved from EHT thread.
Mod edit: moved from EHT thread.
Last edited by a moderator:
400KV / 100ma Power Supply Design
- Thread starter Shatendra Sharma
- Start date
@Shatendra Sharma
Welcome to AAC! Please ask new questions in a new thread.
Welcome to AAC! Please ask new questions in a new thread.
Hypatia's Protege
- Joined Mar 1, 2015
- 3,228
@Shatendra Sharma or @JohnInTX
To the end of avoidance of confusion with our thread bearing the self-same title please consider changing the title of this thread -- many thanks!
HP
To the end of avoidance of confusion with our thread bearing the self-same title please consider changing the title of this thread -- many thanks!
HP
Hypatia's Protege
- Joined Mar 1, 2015
- 3,228
Your question is ambiguous... Please consider the following
-If impulse output is acceptable, your aims are readily and inexpensively achievable via a CW cascade or Marx 'generator'...
-'Flat' continuous DC output at said 'levels' may quite adequately be realized via complementary resonant converters. (note that at Ca. 40kW you're 'looking at' core cross sections ≥ 10cm).
Should your question be in regards to powering an IEC 'fusor', I suggest you contact AAC member @Aleph(0)...
-Line frequency AC is readily achieved via a pair of phased conventional radiotherapy ('DRT') transformers (to 600kv @ 200mA) -- Ditto for half or full-wave unfiltered (line frequency) DC -- In all cases the maximum current (and, hence, loaded EMF) may be 'set' via appropriate feed ballast selection...
!!!Warning!!! If you are inexperienced with --or without comprehensive 'hands on' training on-- high power EHT systems (including arc-flash mitigation training inclusive of active drill) please! Give this project a miss! It is no exaggeration to state that death is among the 'kinder' consequences of the slightest misstep!
Best regards
HP
Last edited:
First, be cautious about Rontgen rays lurking out of this voltage. Anything over 10 kV has this problem, but over 30-40 kV it is hard problem, however 400kV that is giant problem.
Second, You has SMPS of size that is quarter of mine headache today. So there You have a plenty of choices.
Begin from power line side: what will be Your EMI filter (2000-5000 USD) or You may eliminate this problem?
Then what will be Your 6 rectifier diodes? My case I found quite good for less than 20 Eur per piece, that is thyristors. Strange but them are cheaper than pure diodes. Then find the way how to manage the inrush current when someone is pulling the lever "ON". As after stays the rather large capacitor battery, that current for short moment may be as large as hundreds of thousands Amps, and Your energocompany will find a sure bad words to characterize Your job. Therefore ther MUST be the mild start. How to realize it effectively, You may find at my half published article http://www.journalijiar.com/article/653/know-how-and-experience-how-to-build-a->100-kw-power-source-for-laboratory-experiments-in-beam-physics./ (what still changes day to day).
Think twice or more about geometry of capacitor battery, it MUST be noninductive if You not plan a get megavolt size spikes. The absolutely best mount is 1 or 2 mm copper double plate with drillings, but it is damn expensive. So I used welded copper "bifilary line", however the spikes still was a trouble what we depressed by smaller capacitor battery straight on the feed ends of igbt. Dont weight this as "small thing" - it is the main underwater stone will Your machine work or will just burn down.
So, then You go to power igbt, where anything brick-looking is quite well for You, the cheapest is from Fuji (75 USD) but better buy any other, because when You shall burn one (and it will unavoidable happen sooner or later) then You shall realize that "lead-in period" is about year, while for IXYS You may get in very next day.
Then comes driver question on those grand 50000 pF gates. The best is to use a proposed gate drivers from igbt producer, but... sorry, them are each about twice as expensive as all igbt, so see the very effective and stable solution of mine, in the same article.
Diodes - in the article I wrote a solution how to stack them, I think it is sure good even for Your voltage. Apply the due reserve there everywhere, at least of factor of two.
Last thing is the trafo. In the article You may find the producer data of very cheap and good ferrite I cores of 1x1x4 inches. You may glue then as the most perverted fantasy may want. To calculate it is the small thing but take in account that everywhere the voltage is more than kilovolt, will unavoidably happen the micro-discharges, what rather soon will destroy Your insulation materials. So, the ONLY solution giving the rather good results is vacuum-infusion of epoxy. Just enclose the coils You turned on turner-bench into plastic film, close anything well with scotch (or Mc Gyver tape) then glue in the thin plastic pipe in one end, and larger pipe in contrary end, where You shall connect the vacuum pump. Smaller pipe push into mixed epoxy. If it denies to such in, apply the acetone to thin it. In the result the all smallest gaps must be filled with epoxydes what may stand maximum maximorum 10 kV/mm. So, the 400 kV = 4 cm gaps everywhere.
Second, You has SMPS of size that is quarter of mine headache today. So there You have a plenty of choices.
Begin from power line side: what will be Your EMI filter (2000-5000 USD) or You may eliminate this problem?
Then what will be Your 6 rectifier diodes? My case I found quite good for less than 20 Eur per piece, that is thyristors. Strange but them are cheaper than pure diodes. Then find the way how to manage the inrush current when someone is pulling the lever "ON". As after stays the rather large capacitor battery, that current for short moment may be as large as hundreds of thousands Amps, and Your energocompany will find a sure bad words to characterize Your job. Therefore ther MUST be the mild start. How to realize it effectively, You may find at my half published article http://www.journalijiar.com/article/653/know-how-and-experience-how-to-build-a->100-kw-power-source-for-laboratory-experiments-in-beam-physics./ (what still changes day to day).
Think twice or more about geometry of capacitor battery, it MUST be noninductive if You not plan a get megavolt size spikes. The absolutely best mount is 1 or 2 mm copper double plate with drillings, but it is damn expensive. So I used welded copper "bifilary line", however the spikes still was a trouble what we depressed by smaller capacitor battery straight on the feed ends of igbt. Dont weight this as "small thing" - it is the main underwater stone will Your machine work or will just burn down.
So, then You go to power igbt, where anything brick-looking is quite well for You, the cheapest is from Fuji (75 USD) but better buy any other, because when You shall burn one (and it will unavoidable happen sooner or later) then You shall realize that "lead-in period" is about year, while for IXYS You may get in very next day.
Then comes driver question on those grand 50000 pF gates. The best is to use a proposed gate drivers from igbt producer, but... sorry, them are each about twice as expensive as all igbt, so see the very effective and stable solution of mine, in the same article.
Diodes - in the article I wrote a solution how to stack them, I think it is sure good even for Your voltage. Apply the due reserve there everywhere, at least of factor of two.
Last thing is the trafo. In the article You may find the producer data of very cheap and good ferrite I cores of 1x1x4 inches. You may glue then as the most perverted fantasy may want. To calculate it is the small thing but take in account that everywhere the voltage is more than kilovolt, will unavoidably happen the micro-discharges, what rather soon will destroy Your insulation materials. So, the ONLY solution giving the rather good results is vacuum-infusion of epoxy. Just enclose the coils You turned on turner-bench into plastic film, close anything well with scotch (or Mc Gyver tape) then glue in the thin plastic pipe in one end, and larger pipe in contrary end, where You shall connect the vacuum pump. Smaller pipe push into mixed epoxy. If it denies to such in, apply the acetone to thin it. In the result the all smallest gaps must be filled with epoxydes what may stand maximum maximorum 10 kV/mm. So, the 400 kV = 4 cm gaps everywhere.
