4hv.org: Forums / High Voltage / Spiral impulse generator = rolled cap Blumlein induction coil!

Forums

4hv.org :: Forums :: High Voltage

<< Previous thread | Next thread >>

Spiral impulse generator = rolled cap Blumlein induction coil!

Go to page 1 [2]

Author

Post

Proud Mary

Tue Apr 22 2008, 05:12AM

Registered Member #543
Joined: Tue Feb 20 2007, 10:26AM
Location: Brighton, UK
Posts: 1636

Dr. Shark wrote ...

Harry wrote ...

Unfortunately, I don't have access to the full text, but the statement suggests the reason why some designs add some extra windings to the active fast line and not the passive slow line.

You do now, check your PMs.

Thank you so much Dr Shark! Its amazing that such a simple device wasn't discovered until the 1960s, and that professional scientists were still trying to explain how it works in this 2003 paper.

I like to have as much understanding as I can before I start on a project, because I've wasted so much time and money in the past on building things which didn't work properly because I didn't understand all the principles involved.

No PMs for any reason thankyou

Dr. Shark

Tue Apr 22 2008, 05:26AM

Registered Member #75
Joined: Thu Feb 09 2006, 03:30AM
Location: Helsinki, Finnland
Posts: 674

Harry, I completely agree with you, especially after I essentially wasted most of last night constructing a bit of a dud. I already edited my above post to reflect my latest ideas, but I'll repeat:
The only reference that claims an improved output with extra secondary turns is the 1984 patent Peter cited in his first post. This looks more like an ordinary transformer to me than a VIG, although the two are of course intimately interrelated. As I understand the VIG, there is a cycle of LC oscillation which reverses the voltage on the capacitor, so the energy has to be stored in the magnetic field as the voltage crosses zero. At this point, a voltage would be induced in the additional turns. Thus the "extended VIG" seems to be quite different from a normal VIG in its operating principle.

All other references I could find about VIGs just use a 1:1 turns ratio, so that is what I will try tonight. Rewinding my VIG is pretty straightforward, so I'll see if that helps and report back.

Back with a vengeance

Proud Mary

Tue Apr 22 2008, 05:55AM

Registered Member #543
Joined: Tue Feb 20 2007, 10:26AM
Location: Brighton, UK
Posts: 1636

Having got the paper thanks to our friend Dr Shark, and seen the information on the D/n ratio in Table 1, I feel able to begin a modest proof of concept model.

I've decided to use 75mm aluminium adhesive tape for the lines, and a double thickness of 100mm adhesive insulating tape as the dielectric, with a bundle of ferrite rods as the core.

I have the ferrite rods, but will have to send off to ebay for the aluminium tape and the extra-wide insulating tape, so won't be able to do anything in a hurry.

I've not made my mind up on the switch, but suppose I will go for a self-igniting spark gap to start with, which should be OK with a 3kV charge voltage from a Bertrand PSU I have to hand. I will charge a reservoir capacitor first, and power the VIG from that after disconnecting the supply, as I see from various patents that the EMP from VIGs can destroy electronic equipment nearby, or to which it is coupled, so it seems to me to need its own separate low Z earth.

No PMs for any reason thankyou

jpsmith123

Tue Apr 22 2008, 06:02AM

Registered Member #1321
Joined: Fri Feb 15 2008, 09:22PM
Location:
Posts: 208

Could someone please send me that paper by Rose et al.? Thanks.

Dago

Tue Apr 22 2008, 07:21AM

Registered Member #538
Joined: Sun Feb 18 2007, 02:33PM
Location: Finland
Posts: 142

I'm interested in the discharge length, does anyone have equipment to measure such short pulses?

Dr. Shark

Tue Apr 22 2008, 08:05AM

Registered Member #75
Joined: Thu Feb 09 2006, 03:30AM
Location: Helsinki, Finnland
Posts: 674

I would not count on the pulses being that short, after all there is considerable inductance in the circuit from coiling up the transmission line. In the papers I looked at the discharge in in the order of a microsecond, which is not what I would call short.

Back with a vengeance

Proud Mary

Tue Apr 22 2008, 08:32AM

Registered Member #543
Joined: Tue Feb 20 2007, 10:26AM
Location: Brighton, UK
Posts: 1636

"The pulsewidth is equal to 2nt τt where nt is the number of turns and τt is the average one way propagation time for a single turn. If t = 0 represents the closing of the switch, the leading edge of the output pulse is delayed by an amount nt τt which for many applications is unimportant."

Source: Fast risetime spiral pulse generator
United States Patent 4140917

For practical purposes, I would reckon on the real life pulse being perhaps twice as long as the theory.

The pulses are triangular, but they could always be sharpened with a spark gap if necessary.

No PMs for any reason thankyou

Tesladownunder

Wed Apr 23 2008, 03:39AM

Joined: Thu Feb 02 2006, 03:45AM
Location: Bunbury, Australia
Posts: 1307

I have built a beefier setup.
Specs
Output... bright 1/2 inch (1.25cm) sparks (perhaps 15kV)
Input..... 2.5kV peak. (ie 6 times voltage stepup) (NST 1/2 wave plus 1MOhm for testing).
Foil........ Equal length copper foils 4.5cm x 243cm plus lead in.
Former... 4.2 x 7cm. Ext diameter 6.2cm.
Turns..... Calculated 14 turns
Spark gap 1mm
Core...... Ferrrite E cores to fit in 4cm gap.
Cap....... 20 nF (prev one was 0.9nF)
Insul...... Polyethylene groundsheet

So a better result with more turns and area.
How long the dielectric lasts is unclear and also what increase in input voltage and power will it tolerate? Probably needs to be in oil for that.

TDU

..... "Power tends to corrupt, and absolute power corrupts absolutely." Lord Acton, 1887.
...................... See absolute corruption at

Proud Mary

Wed Apr 23 2008, 05:04AM

Registered Member #543
Joined: Tue Feb 20 2007, 10:26AM
Location: Brighton, UK
Posts: 1636

Great stuff, TDU.

I have this morning sent off for 75mm adhesive aluminium tape and 100mm insulating tape from ebay suppliers, so I should have a rig of my own to report on in a week or ten days.

I must say how glad I am you dug this up, and gave us all the opportunity to learn something out of the ordinary.

No PMs for any reason thankyou

Tesladownunder

Wed Apr 23 2008, 05:38AM

Joined: Thu Feb 02 2006, 03:45AM
Location: Bunbury, Australia
Posts: 1307

It's interesting to go down a new path. I've often wondered about impulse transformers although I guess this is only just one type.
This new version seems to have scaled in spark brightness consistent with 20 times the capacitance and also scaled in output spark length 8mm to 12 mm consistent with the increased turns from 9 to 14. Construction type was similar and dielectric and input voltage were unchanged.
The capacitor holds its charge well and if discharged with a chicken stick giving a reasonable "zap", will give an output spark as well.
I suspect that for high repetition rates, the use of a multiple segment gap would prevent power arcing.
I also suspect from my experience with TEA nitrogen lasers that the spark gap type and position is critical. The use of a gap with two cylindrical electrodes in continuity with each conductor strip and covering the full width should present the lowest inductance.
As it is, I suspect it could be driven much harder than just a few pulses per second, but I am not ready to break it yet.

TDU

..... "Power tends to corrupt, and absolute power corrupts absolutely." Lord Acton, 1887.
...................... See absolute corruption at

Proud Mary

Wed Apr 23 2008, 06:11AM

Registered Member #543
Joined: Tue Feb 20 2007, 10:26AM
Location: Brighton, UK
Posts: 1636

I have seen references to VIG oscillators, where the lines form part of an LC tuned circuit, so it may be possible to drive them quite hard.

My chief concern is about the 'goodness' of the dielectric, on which the velocity of propagation and voltage efficiency depends, not to mention holding off break downs.

If we could sort out the dielectric, almost anyone could achieve hundreds of kV and maybe more with such a simple device. I've noticed, though, that the seemingly more advanced designs often have more complex segmented winding arrangements.

No PMs for any reason thankyou

Dr. Shark

Wed Apr 23 2008, 05:21PM

Registered Member #75
Joined: Thu Feb 09 2006, 03:30AM
Location: Helsinki, Finnland
Posts: 674

I redid my VIG with a significantly larger diameter, which as Harry pointed out is supposed to significantly increase the voltage efficiency. I also went for a 1:1 turns ratio with 9 turns each. On my 8cm diameter "core", this gives a D/n of nearly one, so the calculated efficiency should be about 50% without a ferrite core. Going for a much larger former would have improved the efficiency more, but what good would that be if it left me with only 3 turns from my 3m long strips, and therefore a maximum step up ratio of three?

I also redesigned the spark gap for super-low inductance, because apparently the inductance of the "fast side" of the generator is dominated by the switch. The two contacts are not separated by air but by 0.2mm of dielectric, so there is essentially zero loop area between them. Of course this does not allow repetitive operation without burning the dielectric.

Testing the thing immediately showed inter-turn arcing, I should have payed more attention winding the thing (It's getting late, and I am loosing interest in this project). As far as I could tell though, the output spark was still quite weak, most energy being burned in the switching spark gap.

Back with a vengeance

Proud Mary

Wed Apr 23 2008, 05:37PM

Registered Member #543
Joined: Tue Feb 20 2007, 10:26AM
Location: Brighton, UK
Posts: 1636

Dr. Shark wrote ...
Testing the thing immediately showed inter-turn arcing, I should have payed more attention winding the thing (It's getting late, and I am loosing interest in this project). As far as I could tell though, the output spark was still quite weak, most energy being burned in the switching spark gap.

Getting the dielectric just right seems to be everything with this device. I'm sure there must be a source of a suitable material that is cheap and easy to get if only we could think what it was.

I've read several papers where the lines are described as metalized tapes, so I know I have that part right. But I have doubts about my plan to use several thicknesses of 100mm wide insulating tape as the dielectric, because of the difficulty of not trapping any air while I'm winding it with five thumbs.

Kraft paper and oil is always an option, I suppose. It has nothing like the dielectric properties of polyethylene or PTFE, but can be laid on thick and cheap!

No PMs for any reason thankyou

Firefox

Wed Apr 23 2008, 08:08PM

Registered Member #1389
Joined: Wed Mar 12 2008, 07:50PM
Location: Pittsburgh, PA
Posts: 346

What about teflon tape? Its not exactly cheap, but it is an excellent insulator, and you can use relatively small amounts of it to insulate the coils.

There are no stupid questions, but there are a LOT of inquisitive idiots.

I . . . confine myself to tyrannizing over electrons and forcing them to obey my will. Unfortunately, the less lazy among them often do not always do what they are told. -Stella

Proud Mary

Thu Apr 24 2008, 04:13AM

Registered Member #543
Joined: Tue Feb 20 2007, 10:26AM
Location: Brighton, UK
Posts: 1636

Firefox wrote ...

What about teflon tape? Its not exactly cheap, but it is an excellent insulator, and you can use relatively small amounts of it to insulate the coils.

Yes, PTFE would be my first choice, but I have no idea where I might find 100mm wide tape of suitable thickness at a price I could afford. I'm sure it is available from specialist plastic stock suppliers, but having bought PTFE tubes and rods before, I have no doubt it would be expensive.

No PMs for any reason thankyou

Dr. Shark

Thu Apr 24 2008, 06:29AM

Registered Member #75
Joined: Thu Feb 09 2006, 03:30AM
Location: Helsinki, Finnland
Posts: 674

PTFE would also be advantageous from a loss tangent point of view, apparently the Mylar that I am using is quite lossy, whereas PTFE is one of the best plastic dielectrics. I have no idea where it could be obtained though.
The insulating tape you plan on using is probably going to be PVC, do you know the dielectric properties of that? I think it is not commonly used for capacitors, but I don't know why.

Back with a vengeance

...

Thu Apr 24 2008, 09:30AM

Registered Member #56
Joined: Wed Feb 08 2006, 11:02PM
Location: Southern Califorina, USA
Posts: 2118

Not sure how think you need the dialectic, but you can get 6" wide .005" for only $2/ft, 6" bu .010" for $3/ft, and many other sizes in between.

A little pricey, but not completely unrealistic...

check out my website!

Proud Mary

Thu Apr 24 2008, 11:07AM

Registered Member #543
Joined: Tue Feb 20 2007, 10:26AM
Location: Brighton, UK
Posts: 1636

Dr. Shark wrote ...

PTFE would also be advantageous from a loss tangent point of view, apparently the Mylar that I am using is quite lossy, whereas PTFE is one of the best plastic dielectrics. I have no idea where it could be obtained though.
The insulating tape you plan on using is probably going to be PVC, do you know the dielectric properties of that? I think it is not commonly used for capacitors, but I don't know why.

PVC has a typical dielectric constant of 3, and a dielectric strength between about 550V - 750V/mm.

Mylar has a typical dielectric constant of 3.2, and a dielectric strength of around 7kV/mm, vastly superior.

PTFE has a dielectric constant of around 2.1, a dielectric strength of 500V - 1kV/mm, but up to 5kV/mm in 'modified' PTFE, (almost as good as Mylar) but the dielectric strength per mm declines with increasing thickness for some reason.

I'm don't think my plan to use four thicknesses of PVC inslulating tape is likely work except at low charge voltages, so I shall have to keep an eye out for cheap sources of plastic shim or tape of suitable thickness.

If I were to charge my VIG with 14kV DC (rectified from an old but good 10kV NST) then the dielectric would need to hold off at least 28kV at the far end of the line. Allowing a 50% safety margin, this would mean the dielectric must be good to 42kV - which would be 6mm of Mylar.

I am beginning to understand why people have written about the use of capacitor winding machines in the construction of VIGs, though there are usually ingenious amateur means of getting almost as good results for almost none of the price.

No PMs for any reason thankyou

Dr. Kilovolt

Thu Apr 24 2008, 11:19AM

Registered Member #152
Joined: Sun Feb 12 2006, 09:36AM
Location: Czech Rep.
Posts: 2073

Harry wrote ...

PVC has a typical dielectric constant of 3, and a dielectric strength between about 550V - 750V/mm.

I really doubt that it has lower strength than air. I assume that mains cables etc. and all the wires with thick rubbery insulation are PVC insulated. Once I tried 40kV DC on a wire with some 1mm insulation and it didn't break down.

Proud Mary

Thu Apr 24 2008, 12:21PM

Registered Member #543
Joined: Tue Feb 20 2007, 10:26AM
Location: Brighton, UK
Posts: 1636

Plasmaddict wrote ...

Harry wrote ...

PVC has a typical dielectric constant of 3, and a dielectric strength between about 550V - 750V/mm.

Forgive me, I am not used to American notation, and misread the American "mil" as being mm in the table I was using. I have now learnt that 1 American mil = 0.0254 millimeters, which seems a very odd figure to use in metrology. The figures I have misquoted above should be read as American 'mils' and my conclusions about them ignored!

No PMs for any reason thankyou

Bored Chemist

Thu Apr 24 2008, 12:56PM

Registered Member #193
Joined: Fri Feb 17 2006, 01:04AM
Location: sheffield
Posts: 575

Do you folks know one of you was selling one of these spiral thingies?

Andyman

Thu Apr 24 2008, 02:19PM

Registered Member #1083
Joined: Mon Oct 29 2007, 01:16PM
Location: Upland, California
Posts: 210

Yes, one mil is the same thing as 1 thousandth of an inch or .001 inches. Not millimeter

Resistance is futile! (if r<0.7 ohms)
If someone has a gun and is trying to kill you, it would be reasonable to shoot back with your own gun.
My youtube page
I don't have a Ph.D or doctorate of any type so I'm not Dr. Andy, just Andy.

jpsmith123

Thu Apr 24 2008, 07:44PM

Registered Member #1321
Joined: Fri Feb 15 2008, 09:22PM
Location:
Posts: 208

I looked into these spiral vector-inversion generators a few years ago. I may be wrong, but I ultimately concluded that for voltage pulses of several hundred Kv to a Mv or so, with pulse widths of tens to hundreds of ns, delivering energies of tens to a few hundred joules, an air-core pulse transformer might be a better all-around choice. Yes, the spiral generator has the apparent advantages (over a Marx generator) of cheap construction, small size and a single spark gap switch, but a properly designed and constructed air-core pulse transformer would have the same advantages, I think, and may actually be easier to build and more robust.

Proud Mary

Fri Apr 25 2008, 05:25AM

Registered Member #543
Joined: Tue Feb 20 2007, 10:26AM
Location: Brighton, UK
Posts: 1636

jpsmith123 wrote ...
a properly designed and constructed air-core pulse transformer would have the same advantages, I think, and may actually be easier to build and more robust.

But for the home experimenter, amateur scientist and tinkerer like me, the Vector Inversion Generator is a whole new challenge with new concepts and new skills to master. You might as well tell a radio ham he could buy a better transmitter than he could ever build, so why waste his time in the garden shed with a soldering iron. But that is not what hobbies are all about!

No PMs for any reason thankyou

jpsmith123

Fri Apr 25 2008, 06:11AM

Registered Member #1321
Joined: Fri Feb 15 2008, 09:22PM
Location:
Posts: 208

I didn't mean to imply that it's not worth tinkering with, if you're so inclined; rather, I was pointing out that if you're building it because you want an inexpensive source of moderately fast HV pulses for some application, an air-core pulse transformer might be better. (And I don't see how your "buying vs building" analogy applies, since you would build the air-core xfrm and in a sense face "a new challenge" there too).

Tesladownunder

Sat Apr 26 2008, 03:42AM

Joined: Thu Feb 02 2006, 03:45AM
Location: Bunbury, Australia
Posts: 1307

Another upgrade with 40% more windings.
Specs
Output... bright 1inch (2.5cm) sparks (perhaps 30kV)
Input..... less than 12kV peak. (ie 3 times voltage stepup) (NST doubled plus 1MOhm).
Foil........ Equal length copper foils 4.5cm x 367cm plus lead in.
Former... 4.2 x 7cm. Ext diameter 8cm.
Turns..... Calculated 20 turns
Spark gap 2mm
Core...... Ferrrite E cores to fit in 4cm gap.
Cap....... 45 nF (prev one was 0.9nF)
Insul...... Polyethylene groundsheet

Arcing over is a problem with higher input voltages and the higher voltage. Anything further would have to be under oil.
I am still a bit uncertain about the 12kV input with voltage doubled NST across the diode and with a series resistor into the capacitance of the generator. It is probably a fair bit lower as the 2mm spark gap is at the limit of working. I will try with a different supply and metering later.

Has anyone else got any spark photos to show?

Peter

..... "Power tends to corrupt, and absolute power corrupts absolutely." Lord Acton, 1887.
...................... See absolute corruption at

Proud Mary

Sat Apr 26 2008, 04:17AM

Registered Member #543
Joined: Tue Feb 20 2007, 10:26AM
Location: Brighton, UK
Posts: 1636

Well done, TDU.

I've sent off by mail order for some aluminium tapes but I haven't sorted out the dielectric yet, which seems to me to be at the heart of the thing.

No PMs for any reason thankyou

Go to page 1 [2]

Moderators: Chris Russell, Noelle, kalenedrael, Alex, Grant, Tesladownunder, ShawnHV, EastVoltResearch, Dave Marshall, Ben, Bjørn, Matt, Dave Billington, WaveRider, Steve McConner, Anders M., Simon, Tim Koene