In Need of a Little Help in Designing a Versatile High Voltage Power Supply / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

In Need of a Little Help in Designing a Versatile High Voltage Power Supply

previous 1 2 3 4 next

Move Thread

LAN_403

Robert Francis

Sat Sept 12 2009, 09:21AM

Registered Member #2355 Joined: Thu Sept 10 2009, 09:12PM
Location:
Posts: 71

Ok, I think, I understand...

A normal square wave pulsed DC waveform could not be put through a transformer because the vast majority of the waveform is linear and not changing.

A Sine Wave created by the XR2206 chip for example, that oscillates between 0-5V (rather than -5 to +5), fed into a voltage amplifier to get 0-50kV, could then be put through a transformer (say 20:1 turns ratio) to then get 0-1MV because the waveform is constantly changing (not linear) and the voltage would stay positive and not go negative.

Or do I have that wrong and that would still end up outputting a voltage that oscillates from positive to negative?

Antonio

Sat Sept 12 2009, 02:26PM

Registered Member #834 Joined: Tue Jun 12 2007, 10:57PM
Location: Brazil
Posts: 644

ItÂ´s almost impossible to make a linear amplifier with 50 kV output. Remember that itÂ´s difficult to contain high voltages. Sparks in air require just about 10 kV/cm between points. In principle, a step-up transformer can be built to produce high voltage with any waveform that does not contain a DC component, but to build a wideband transformer is not a trivial problem, aggravated by the insulation requirements.

hvguy

Sun Sept 13 2009, 10:12AM

Registered Member #289 Joined: Mon Mar 06 2006, 10:45AM
Location: Conroe, TX
Posts: 154

While there has been some good advice put forward on this subject I think there is still some reality lacking from the picture. First, and I mean no offense here, it is apparent that you lack the knowledge, at least at this point, to build something this complex. Second, what you are asking for could easily turn into a six figure research project spanning several years. I build power supplies, not entirely unlike what you have described, for a living, and I can tell you first hand what you're asking for is on the fringe of current possibility.

If you are really serious about this you should consider revising your requirements, maybe working with a lower max voltage and fewer waveforms, or starting with a low voltage prototype. I know well educated power supply designers who have been around for 30yrs that would not be able to pull something like this off. That being said, I'd take a project like this on if I was approached with an appropriate budget and time line, so I'm sure you could find a way... Good luck

Proud Mary

Sun Sept 13 2009, 01:53PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Why not try mucking about with an ignition coil from a scrap yard? That's how every one else here began, and first rate training it is too.

Do you have factory-sized premises for your megavolt generator? Have a look through Google and see how enormous practical iterations of megavolt devices are in real life. Have a look at them and you'll see there won't be much change left from a million dollars. Anything else is fantasy.

I mean, get with the program, man!

Robert Francis

Sun Sept 13 2009, 03:59PM

Registered Member #2355 Joined: Thu Sept 10 2009, 09:12PM
Location:
Posts: 71

I probably am biting off more than I can chew but what options do I really have?

I can go with Steve McConnor's suggestion which IMO sounds like the best one at this point:

Create a function generator, perhaps one based off of this USB PIC design expanding it to cover a larger frequency range. Problems with this:
- My code writing was never very good despite me having a BS in CS
- I got a C+ in my Microprocessors course showing I am not that good with circuits
- I've never built a function generator or worked with PIC chips.

Create a power amplifier perhaps using a mix of silicon chips to increase current and vacuum valves to increase voltage. (I didn't realize this is basically just an amp like those used in stereo equipment.) Problems with this:
- I have never built an amp before.
- My microprocessors kung fu is weak

Create multiple primary/secondary transformers with each pair designed to cover a different frequency and voltage range.
- I've never built a transformer before

I have been to enough Tesla coil sites to see people have created megavolt tesla coils and it did not cost them a million bucks to do it. More than likely I would build a variety of secondary coils in different shapes and sizes, figure out what frequency ranges and voltages they are capable of and run them within those ranges.

Other options?
I could rectify wall current, leaving the ripple, and build a transformer specifically to raise the voltage of that current at it's frequency either 60hz or 120hz.
Much simpler but much less flexible. No ability to tie it into a PC and use a PC to control the coil's parameters if there were even parameters to control.

This alternative just doesn't seem like a good idea to me.

If I saw a better plan I would use it. I realize I do not possess all the knowledge necessary at this point. How bad is arcing between the turns on a secondary coil. Can the turns be insulated from one another, either by creating a 1-2" thick coat of epoxy around the coil (or will it arc straight down through the wires bypassing the epoxy?) or creating a conical secondary coil or both?

Necessity is the mother of invention.

Proud Mary

Sun Sept 13 2009, 04:08PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

From what you say, I'd guess your talents are so far above those of anyone else on this site, that few seem able to appreciate the grandeur and majesty of your plans.

Steve Conner

Sun Sept 13 2009, 07:28PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Huh? What plans?

If you've never built an amp of any kind before, there's a plan: start with a low voltage one, something like an audio amp. You'll maybe learn about the fundamentals of amplification and feedback that way.

Then build some stuff with microprocessors to strengthen your micro-fu.

After polishing Mr. Miyagi's car for 25 years, and catching a fly with chopsticks, you'll be ready to take on the high-voltage-specific issues in the project, which are every bit as challenging as hvguy says. (Hint, those megavolt tesla coils are not really megavolt, they're probably only about 500kV.)

I don't think the six-figure estimate is that far off, either.

Robert Francis

Sun Sept 13 2009, 08:27PM

Registered Member #2355 Joined: Thu Sept 10 2009, 09:12PM
Location:
Posts: 71

I found a couple cool function generators from Velleman Inc.

The PCSGU250 and the PCGU1000

Ideally I would make my own, release the circuit diagram, parts used, and software written open source but that would likely take me months so I am probably going to buy the signal generator. Neither of these are cheap, about $220 a piece, an expensive paperweight if I accidentally blow it up.

Both models come with software that allow the creation of standard signal waves like e.g. sine, triangle and rectangle as well as waveforms created in their Direct Digital wave Synthesis software.

Some other features include:
100mVpp to 10Vpp amplitude settings
0.01% frequency resolution settings
-5V to +5V offset settings

The PCSGU250 has a max capable frequency of 1 MHz and an oscilloscope probe, the PCGU1000 2 MHz but no oscilloscope. Since the voltage capacity of the probe is 600V it probably wouldn't help me anyway other than to be sure the waveform I was inputting into the amplifier was correct.

I didn't check the PCSGU250 software but the PCGU1000 has a wave sequence tester. It takes a tab-delimited spreadsheet of:
Function: type of wave (sine, triangle, custom, etc)
Frequency
Offset
Output Voltage
Time (how long the waveform is generated)

So I could use OpenOffice to create a spreadsheet with a long list of frequencies in no time and import that into the Wave Sequence program. This would help me out in my experiment tremendously. Other than wishing the Wave Sequence program could timestamp when it runs the various frequencies, all the other software is written for me for how I wanted to use this device.

So if you'd rather buy than build, this looks like a useful option.

One part down, two parts to go.

Proud Mary

Sun Sept 13 2009, 08:27PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

I shan't respond to your importunate private message here, Robert, but whether someone is a total newcomer, or an old timer, then the laws of physics, and the formulae of electronics, are ever our gold standard here at 4HV.

There is no known electronic circuit to which these laws do not apply, so it is often apparent, merely at a glance of the experienced eye, that an idea or configuration just won't work.

I often get things wrong, and there are few here who don't make blunders of reason and logic, but whilst the good student is actually grateful to have their mistakes pointed out, so they may do better on the next occasion, the bad student rushes on from error into denial, and from thence into foolishness and complaint.

Learning, and further learning, and learning beyond that, is our hobby here.

Robert Francis

Sun Sept 13 2009, 09:23PM

Registered Member #2355 Joined: Thu Sept 10 2009, 09:12PM
Location:
Posts: 71

I think you guys are focusing on my desire to get to megavolt levels a little too much. If whatever transformer I build can't do it, it can't do it, its not the end of the world, there is a wide range of voltages and frequencies I want to test.

My only complaint Harry is in destructive criticism, I welcome constructive criticism.

Steve McConnor gave me constructive criticism, that I might need to dial my hopes back a bit but I might be able to make it work. He saw that I wanted to build a device that could make use of multiple waveforms and brought up an amplifier that could make a high voltage replica of a low voltage reference wave.

You told me to muck about with ignition coils (I did that in high school shop class), that my idea was fantasy, to get with the program (exclamation point), a sarcastic statement implying I thought my talents were better than everyone elses, and implying that I am a bad student who has rushed into foolishness.

Do you see the difference?

I sent you a PM because I wanted the thread to focus on the power supply, not bruised egos, mine and yours but instead of replying to the PM you replied in the thread so I have as well.

Steve, the plan was 3 components as I saw it:
the function generator
the power amplifier (10 kV is probably as high as I would need that to go)
the various size and shape primary/secondary transformers

As far as having individual plans for those parts, no, I have none, well except to fork out some cash to save myself some time.

If those 3 components will not work together in that fashion, what's wrong with it, will a transformer not work well transforming a waveform from a power amplifier? Neon sign transformers used to power Tesla coils are usually 10-15 kV.

Also, for the transformer, if I coated a PVC pipe with 2" of epoxy resin all around, that should have a dielectric strength of over 1MV. If I then wired the coil 100:1 turns ratio, maybe 300 - 400 turns of grade 3 insulated polyurethane magnetwire over that epoxy covered PVC pipe, and then covered the magnetwire with 2" epoxy resin that should prevent arcing through to the air or to the PVC.

What would happen regarding arcing between the magnetwire turns? Each turn would only raise the voltage about 2.5 - 3kV. The dielectric strength of polyurethane is about 70 kV per mm. Would it still arc through the turns or would it not arc because the voltage difference isn't enough to break through the polyurethane coating?

previous 1 2 3 4 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob