Fusor High Voltage Power Supply thread
389 views
Skip to first unread message
Mike Bushroe
May 28, 2014, 3:07:57 AM5/28/14
to heatsy...@googlegroups.com
Just some random notes I pulled off the web about high voltage wires:
I think that racing spark plug wires will work well for us until we get much farther along. However, the typical internal resistance of the current ant-EMI wires of 2k to 4K mean that even at 30mA, there will be 60 to 120 volts lost across each foot of wire. Several feet of wire for the transformer to voltage adder bridge, from there to the fusor cage feed through, and then from the chamber or shield feed through back to the transformer, plus any switching or fusing we might add, and the total length can begin to add up. And if we are starting without the voltage adder bridge with only 4-7KV, dopping 500 to 1000V can make a difference. Still, once we have enough voltage adder stages to get up to 30KV or higher, that will not make much difference. And the lower cost and easy availability of the spark plug wire in bulk rolls and easy of working with it make it a good choice. And we can even buy lots of different size, angle, and shape covered connectors to further reduce risk of arcing or corona discharge. However, the RG-8 with the braided shield grounded has even less risk or corona discharge breaking down the insulation. And it too should be good up to 30KV according to the posts, even though it is technically rated for only 1KV. And there will be much less voltage lost by the center conductor.
Mike
http://home.earthlink.net/~jimlux/hv/hvwire.htm
With bare conductors, air is the insulator, and clearance distances can be calcuated using standard values for the breakdown of air. A common rule of thumb which is very conservative is 1 inch per 10 kV. Since the breakdown field for air is around 71 kV/inch, this provides a 7:1 safety factor.
Popular insulation materials for hookup type wire are polyethylene, PTFE, rubber, and silicone, particularly the latter. Neon signs are a cost sensitive application, so inexpensive wire ($.15/ft) rated at 15 kV with polyethylene insulation is widely available. Rubber is popular for test leads at the 5 kV level, although many rubbers degrade in the presence of ozone, which is often present in HV equipment. High quality high voltage wire has silicone insulation which is quite flexible and high temperature resistant.Typical prices for silicone insulated wire range from $.20/ft for 10kV rated to $2.00/ft for 50 kV rated.
Corona resistant wire is typically constructed with a central copper core surrounded by a semiconducting sheath, which in turn is surrounded by the insulation. The semiconducting sheath effectively increases the diameter of the wire, reducing the tendency for corona discharge. Suppliers of such wire include Belden, Caton, Tally, etc.
Coaxial cable of the RG-8 (RG-213) family is often used as high voltage cabling for several tens of kV. Grounding the outer shield makes the field distribution inside the cable very even, reducing the field concentrations that start corona. RG-8 is rated at 5 kV RMS, however, the polyethylene insulation is (.285-.01??) .120 inches thick which corresponds to 120 kV breakdown. I suspect that the 5kV rating (7 kV pk) allows for a substantial VSWR in transmission line use without breakdown. Certainly, many systems use RG-8 at 25 kV, and I have seen some at 50 kV using RG-8 as a conductor. Also, the field strength at the inner conductor is higher than that at the outer conductor
Equation here.
Having the outer surface of the cable at ground potential also confers some safety advantages. Don't forget though, that in systems with sufficient stored energy, the coax can literally explode in the event of a dielectric failure. If you have several tens of kJoules stored up, the energy has to go somewhere. At least you won't get shocked, just burned.
Coaxial cable using foamed dielectrics (e.g. RG-8X) are not useful, since the nitrogen used to make the foam has a much lower breakdown than the PE. The same goes for RG-59 cable TV remnants, because they are usually foamed insulation (cheaper and lower loss).
http://www.practicalmachinist.com/vb/general/custom-spark-plug-wires-high-voltage-app-139958/
If you are in the 30,000 V range you do not need insulation surrounding the wire other than for your safety.
You just need support insulators with sufficinent over the surface
length. The short time breakdown voltage of most insulators within the
insulating material is in the range of 300 to 1000 V per 0.001"
thickness. As thickness increases breakdown voltage drops
some. So 30 to 50/1000" of Mylar would be a minimum for 30,000 V.
On the outside of the insulator in dry air the breakdown at standard
conditions is about 3,000,000 V/M. Thus, 1/100 M for 30,000 V or
39.37/100" = 0.4". At lower pressure down to a minimum the distance
increases, and as pressure increases the distance is less.
That is why spark plug gaps are so small.
But over the surface of the insulator will be less than the air breakdown value.
Bannana plugs and jacks are fine. If you have sharp corners cover them
with non-conductive epoxy. High frequency radio insulators may be a
useful support. I used one for a 50,000 V scope input voltage divide
that I made many years ago when working on the development
of an automotive ignition system. The dropping resistor was 6" long to
reduce the voltage gradient along the resistor.
Describe more about what you want to accomplish. Also look inside a color CRT and see what is done there.
Also note above about 10,000 V X-rays are generated. That is the reason for lead in the CRT tubes.
Be careful.
http://www.generalcable.com/NR/rdonlyres/F3629AA7-1B9E-40E6-A74B-4D3949CD1B9B/0/AUT_0060_1002.pdf
The demands on ignition wires are increasing
as a result of higher-revving
engines, higher operating temperatures, exposure to potentially damaging fuels and
chemicals, increased utilization of on-board electronics, and increasing number of
miles driven per year. The additional functional requirements of isolating and
delivering the 30,000 – 50,000 volts required to reliably provide the spark needed
for an internal combustion engine,
Good Better Best
Insulating Properties Silicone EVA EPDM
(Dielectric)
http://www.allaboutcircuits.com/vol_1/chpt_12/8.html
Material* Dielectric strength (kV/inch)
===========================================
Vacuum ------------------- 20
Air ---------------------- 20 to 75
Porcelain ---------------- 40 to 200
Paraffin Wax ------------- 200 to 300
Transformer Oil ---------- 400
Bakelite ----------------- 300 to 550
Rubber ------------------- 450 to 700
Shellac ------------------ 900
Paper -------------------- 1250
Teflon ------------------- 1500
Glass -------------------- 2000 to 3000
Mica --------------------- 5000
* = Materials listed are specially prepared for electrical use.
I think that racing spark plug wires will work well for us until we get much farther along. However, the typical internal resistance of the current ant-EMI wires of 2k to 4K mean that even at 30mA, there will be 60 to 120 volts lost across each foot of wire. Several feet of wire for the transformer to voltage adder bridge, from there to the fusor cage feed through, and then from the chamber or shield feed through back to the transformer, plus any switching or fusing we might add, and the total length can begin to add up. And if we are starting without the voltage adder bridge with only 4-7KV, dopping 500 to 1000V can make a difference. Still, once we have enough voltage adder stages to get up to 30KV or higher, that will not make much difference. And the lower cost and easy availability of the spark plug wire in bulk rolls and easy of working with it make it a good choice. And we can even buy lots of different size, angle, and shape covered connectors to further reduce risk of arcing or corona discharge. However, the RG-8 with the braided shield grounded has even less risk or corona discharge breaking down the insulation. And it too should be good up to 30KV according to the posts, even though it is technically rated for only 1KV. And there will be much less voltage lost by the center conductor.
Mike
Eric Hubert
May 28, 2014, 5:28:23 AM5/28/14
to heatsy...@googlegroups.com
I don't think half of us on the group knows exactly what you are saying but darn that was awesome. The fusor is a great project. Would any of us be wrong in calling it a Fusion Reactor (mainly to both scare and interest people)?
Mike Bushroe
May 28, 2014, 11:42:40 AM5/28/14
to heatsy...@googlegroups.com
Eric,
You are not the first to ask that!
Yes the 'Fusor Fanatics' (R) are indeed planning on build one, or more Nuclear Fusion Reactors! Hydrogen nuclei will be smashed into hydrogen nuclei to produce helium. Fusion energy will be released (in very tiny quantity), our Prometheus will steal not the Fire of the Gods but Fire of the Stars themselves! Part of the plan is that each stage of development from very small test model to much larger, longer running fusion reactors will also have a clear section so that people can see the 'Star Fire', a phrase used by other people who have successfully made similar fusion reactors to the ones we want to build. This means that any one can see the tiny bit of nuclear fusion we are managing to create.
Perhaps a short FAQ for what we plan.
Q. Are you taking about real thermonuclear fusion?
A. Yes, but there won't really be any 'thermo', it will never generate enough fusion to warm anything up.
Q. Is this like the 'cold fusion' we heard about a decade or two back?
A. No, that was a hoax. This is real, 'hot' fusion, just like in the core of stars, or the big Tokamaks or Laser fusion reactors.
Q. If it is the same fusion as what the big guys are taking decades and billions of dollars to do, how come you think you can do it with spare parts on the bench?
A. Very good question, and the core of what we are talking about. The 'fusor' is small, fairly easy to make and uses a static electric field to accelerate positive hydrogen (actually deuterium) nuclei to very speed to occasionally bang into each other hard enough to fuse into helium. The simple electrostatic fields are fairly simple to generate (if you have a 45KV power supply in you pocket ;)!) to generate the Inertial Electrostatic Confinement (IEC).
Q. If it is so easy to do, why don't the big guys do it instead Tokomaks and giant lasers?
A. Because the fusor is inherently limited in efficiency. It generates real fusion, but only in very tiny amounts, and always using lots more power to run the fusor than it releases in energy. The fusor design will NEVER reach break even, let alone power our Flux Capacitor :). With a limit on never even reaching scientific break even, let alone produce net power, they don't waste time on it.
Q. If it can't produce any real power, why do it?
A. Because even with this simple method, the total number of living people in the world how have successfully fused hydrogen into helium is very small. There are even lists of who has done it, their age and location, and how long they ran and total estimated atoms fused. We want also be able to brag that we have stolen the fire of the Stars and harnessed the power of nuclear fusion our selves.
Q. If it is Make Magazine, how hard can it be? Why do we need such a fancy setup with special vacuum pumps and super high voltages?
A. Even with the simple basic design and the article in Make magazine/instrcutable, it is still hard to build, tweak, vacuum down far enough, generate the high voltage without it shorting out and a lot of other problems to get even a tiny amount of atoms to fuse. We want to make a machine that doesn't just squeak over the limit. We want to run a strong reaction, with lots of atoms fusing per second, and to be able to clearly prove it by measuring the neutrons emitted by each fusion.
Q. Wait a minute! Yes said "neutrons emitted"? Does this thing generate radiation? Is it dangerous?
A. Well, actually yes, yes, and yes. Even the tiny Make magazine fusor releases a tiny number of neutrons, and high energy gamma rays. The number of actual nuclei fuzed is so small there is little radiation over normal background levels. If we get a bigger, better model going, it will generate potentially dangerous levels of radiation.
Q. Then why don't you DO something to stop all that dangerous radiation?
A. We will wrap the whole thing in 2mm of lead shielding, use radiation vests, and eventually getting a leaded glass radiation proof view port. We will also have geiger counters (probably several!) to monitor safe conditions outside the radiation shielding, and we are looking into dosimeters that people can wear to prove that they did not get exposed to radiation. By using a periscope like viewing port with a mirror inside that reflects the light out in a different direction so that you can look inside and still be outside the line of radiation. Since light waves are easily reflected but neutrons and gamma rays are not, if we have even a single mirror letting us see the fusion reaction taking place (beautiful blue glow!), then we can surround the reactor with lead, and have the light come out at an angle that no radiation can follow.
Q. Only 2 mm of lead shielding? Is that enough?
A. According to our research, only 2 mm of lead are needed for complete safety, reducing the neutrons to 0.4% of the original. But we are planning on using 12 mm (half an inch) because Larry has that much and it will reassure the less technically inclined to feel comfortable with the level of protection. The leaded glass radiation proof view port we intend to add will also provide the equivalent protection of 2 mm of lead (yes, leaded glass has real lead in it!) so that will be completely safe even without a mirror. But, if we put it in the front window or central in the lab and shown running to everyone that comes in, they will have a hard time believing that the perfectly clear glass (and leaded glass is often wonderfully clear and transparent - to visible light!) can possibly protect them from the radiation. So even the leaded glass alone is enough, we would want the display model to still have one or two mirrors to completely reassure people that no radiation can get out.
Q. Are there any other dangers?
A. The high voltage section is actually much more dangerous than the radiation! A single injudicious touch and 30 thousand volts of electricity with more than enough current to stop your heart and burn your skin and it is all over. So we are going to use high voltage wire (see below) and high voltage transformers, diodes, capacitors, and then enclose all the high voltage parts behind safety shields. The high vacuum is no more dangerous than the vacuum made by an AC system repair vacuum pump and we are using nice, strong metal vacuum chambers.
Q. Well then, what are we waiting for?
A. We trying to get it going as soon as we can. Do you want to join us and help?
Mike
You are not the first to ask that!
Yes the 'Fusor Fanatics' (R) are indeed planning on build one, or more Nuclear Fusion Reactors! Hydrogen nuclei will be smashed into hydrogen nuclei to produce helium. Fusion energy will be released (in very tiny quantity), our Prometheus will steal not the Fire of the Gods but Fire of the Stars themselves! Part of the plan is that each stage of development from very small test model to much larger, longer running fusion reactors will also have a clear section so that people can see the 'Star Fire', a phrase used by other people who have successfully made similar fusion reactors to the ones we want to build. This means that any one can see the tiny bit of nuclear fusion we are managing to create.
Perhaps a short FAQ for what we plan.
Q. Are you taking about real thermonuclear fusion?
A. Yes, but there won't really be any 'thermo', it will never generate enough fusion to warm anything up.
Q. Is this like the 'cold fusion' we heard about a decade or two back?
A. No, that was a hoax. This is real, 'hot' fusion, just like in the core of stars, or the big Tokamaks or Laser fusion reactors.
Q. If it is the same fusion as what the big guys are taking decades and billions of dollars to do, how come you think you can do it with spare parts on the bench?
A. Very good question, and the core of what we are talking about. The 'fusor' is small, fairly easy to make and uses a static electric field to accelerate positive hydrogen (actually deuterium) nuclei to very speed to occasionally bang into each other hard enough to fuse into helium. The simple electrostatic fields are fairly simple to generate (if you have a 45KV power supply in you pocket ;)!) to generate the Inertial Electrostatic Confinement (IEC).
Q. If it is so easy to do, why don't the big guys do it instead Tokomaks and giant lasers?
A. Because the fusor is inherently limited in efficiency. It generates real fusion, but only in very tiny amounts, and always using lots more power to run the fusor than it releases in energy. The fusor design will NEVER reach break even, let alone power our Flux Capacitor :). With a limit on never even reaching scientific break even, let alone produce net power, they don't waste time on it.
Q. If it can't produce any real power, why do it?
A. Because even with this simple method, the total number of living people in the world how have successfully fused hydrogen into helium is very small. There are even lists of who has done it, their age and location, and how long they ran and total estimated atoms fused. We want also be able to brag that we have stolen the fire of the Stars and harnessed the power of nuclear fusion our selves.
Q. If it is Make Magazine, how hard can it be? Why do we need such a fancy setup with special vacuum pumps and super high voltages?
A. Even with the simple basic design and the article in Make magazine/instrcutable, it is still hard to build, tweak, vacuum down far enough, generate the high voltage without it shorting out and a lot of other problems to get even a tiny amount of atoms to fuse. We want to make a machine that doesn't just squeak over the limit. We want to run a strong reaction, with lots of atoms fusing per second, and to be able to clearly prove it by measuring the neutrons emitted by each fusion.
Q. Wait a minute! Yes said "neutrons emitted"? Does this thing generate radiation? Is it dangerous?
A. Well, actually yes, yes, and yes. Even the tiny Make magazine fusor releases a tiny number of neutrons, and high energy gamma rays. The number of actual nuclei fuzed is so small there is little radiation over normal background levels. If we get a bigger, better model going, it will generate potentially dangerous levels of radiation.
Q. Then why don't you DO something to stop all that dangerous radiation?
A. We will wrap the whole thing in 2mm of lead shielding, use radiation vests, and eventually getting a leaded glass radiation proof view port. We will also have geiger counters (probably several!) to monitor safe conditions outside the radiation shielding, and we are looking into dosimeters that people can wear to prove that they did not get exposed to radiation. By using a periscope like viewing port with a mirror inside that reflects the light out in a different direction so that you can look inside and still be outside the line of radiation. Since light waves are easily reflected but neutrons and gamma rays are not, if we have even a single mirror letting us see the fusion reaction taking place (beautiful blue glow!), then we can surround the reactor with lead, and have the light come out at an angle that no radiation can follow.
Q. Only 2 mm of lead shielding? Is that enough?
A. According to our research, only 2 mm of lead are needed for complete safety, reducing the neutrons to 0.4% of the original. But we are planning on using 12 mm (half an inch) because Larry has that much and it will reassure the less technically inclined to feel comfortable with the level of protection. The leaded glass radiation proof view port we intend to add will also provide the equivalent protection of 2 mm of lead (yes, leaded glass has real lead in it!) so that will be completely safe even without a mirror. But, if we put it in the front window or central in the lab and shown running to everyone that comes in, they will have a hard time believing that the perfectly clear glass (and leaded glass is often wonderfully clear and transparent - to visible light!) can possibly protect them from the radiation. So even the leaded glass alone is enough, we would want the display model to still have one or two mirrors to completely reassure people that no radiation can get out.
Q. Are there any other dangers?
A. The high voltage section is actually much more dangerous than the radiation! A single injudicious touch and 30 thousand volts of electricity with more than enough current to stop your heart and burn your skin and it is all over. So we are going to use high voltage wire (see below) and high voltage transformers, diodes, capacitors, and then enclose all the high voltage parts behind safety shields. The high vacuum is no more dangerous than the vacuum made by an AC system repair vacuum pump and we are using nice, strong metal vacuum chambers.
Q. Well then, what are we waiting for?
A. We trying to get it going as soon as we can. Do you want to join us and help?
Mike
Michael Mathers
May 28, 2014, 12:43:01 PM5/28/14
to heatsy...@googlegroups.com
That's great research, Mike!
Larry Campbell
May 29, 2014, 3:03:29 AM5/29/14
to heatsy...@googlegroups.com
Stacey brought up some good points and ideas
Duty cycle of automotive spark plug wires....we should consider that....dont want to burn them up with constant current since they were actually designed for less than full duty cycle.
For prototyping simplification of first generation mineral oil recirculation for the cage could be done with gravity feed with reservoirs and a simple pump to reload the upper reservoir , if additional cooling is required merely dropping in some frozen whiskey stones or a teething ring would suffice to keep chamber temps down. (later generations would then know how much additional cooling or heatsyncing would be required by trial and error...(IE incorporate a simple candy thermometer in the coolant)
She also was giving me her ideas on casting shielding plates and possibly casting a solid cylinder to minimize potential for windows in seams as well as ideas for single tube cage design variations that would have minimal blocking potential while still maximizing ion attraction potential....and some ideas on the hydrolysis of deuterium oxide into a trapped deuterium container that could be metered into a chamber..
Plus extra deuterium made from electrolysis could be stored off in spare ampules I have for minimizing waste
I think the next couple brainstorming/planning sessions should be interesting... I picked up some stainless tubing from IMS today to practice variations on large scale cage shapes and Im trying to procure some more parts off ebay too...
The information in this message may be confidential. It is intended solely for
the addressee(s). If you are not the intended recipient, any disclosure,
copying or distribution of the message, or any action or omission taken by you
in reliance on it, is prohibited and may be unlawful. Please immediately
contact the sender if you have received this message in error.
Mike Bushroe
May 29, 2014, 4:04:22 AM5/29/14
to heatsy...@googlegroups.com
Larry,
congrats to Stacey! She is right on the money. Worst case spark plug wire from what I read is 4K ohms per foot. Pushing 30mA through that is 120W/ft which is a bit much. For automotive use, on a four stroke engine each wire only fires once every other revolution, and they probably want the sparks to be closely timed, so perhaps a 1% duty cycle. Assuming the source for the spark is also current limited to around 30mA (I have no idea if that is correct) that would drop it to 1.2w/ft which is fairly easy to dissipate, especially since they are designed to work in a high temp environment. So maybe it will be RG-8 after all. I am going to want to do a full voltage test against a grounded rod before I will be comfortable using it all by itself, but if the shield braid is grounded and the center conductor is at -30KV, then the cable will be safe because the nearest part to us is grounded!
From what I read, you want to avoid the cheaper RG-8 cables with some kind of foamed plastic dielectric. I found one for roughly $0.7 per foot that has solid polyethylene.
More later after I wake up!
Mike
congrats to Stacey! She is right on the money. Worst case spark plug wire from what I read is 4K ohms per foot. Pushing 30mA through that is 120W/ft which is a bit much. For automotive use, on a four stroke engine each wire only fires once every other revolution, and they probably want the sparks to be closely timed, so perhaps a 1% duty cycle. Assuming the source for the spark is also current limited to around 30mA (I have no idea if that is correct) that would drop it to 1.2w/ft which is fairly easy to dissipate, especially since they are designed to work in a high temp environment. So maybe it will be RG-8 after all. I am going to want to do a full voltage test against a grounded rod before I will be comfortable using it all by itself, but if the shield braid is grounded and the center conductor is at -30KV, then the cable will be safe because the nearest part to us is grounded!
From what I read, you want to avoid the cheaper RG-8 cables with some kind of foamed plastic dielectric. I found one for roughly $0.7 per foot that has solid polyethylene.
More later after I wake up!
Mike
Jasper Nance
May 29, 2014, 8:48:03 AM5/29/14
to heatsy...@googlegroups.com
Not sure why y'all think HV wire is expensive...
--
You received this message because you are subscribed to the Google Groups "HeatSync Labs" group.
To unsubscribe from this group and stop receiving emails from it, send an email to heatsynclabs...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
----------------------------------------------
Jasper Nance - KE7PHI
Creative and Scientific Imagery
http://www.nebarnix.com/
Message has been deleted
Eric Hubert
May 29, 2014, 10:09:22 AM5/29/14
to heatsy...@googlegroups.com
I can't resist.......... FREE ePacket delivery from China <------------ china syndrome anyone?
I'd be all for helping with the casting. This may be somewhere I can be less of a watcher and more of a participant.
Larry Campbell
May 29, 2014, 10:11:02 AM5/29/14
to heatsy...@googlegroups.com
You can see why I like having her help on projects...
We could do a stress test of the plug wires, they MIGHT hold up fine however we should not be surprised if they fail. In a failure it would probably be less devastating than a coax failure.
Should we go with the silicone wire 2.50 a foot is certainly not the most expensive wire I have worked with, Ive installed cables that were well over 20 dollars a foot...
Mike Bushroe
May 29, 2014, 10:49:32 AM5/29/14
to heatsy...@googlegroups.com
Well, you know the old saying. I was looking for love in all the wrong places! :)
Considering that I had looked on Bay already for high voltage caps and diodes, I don't know how I slipped up on the wire. I am glad that you are still here, er, over there, to help keep us in line! :)
Mike
Considering that I had looked on Bay already for high voltage caps and diodes, I don't know how I slipped up on the wire. I am glad that you are still here, er, over there, to help keep us in line! :)
Mike
Mike Bushroe
May 29, 2014, 1:06:48 PM5/29/14
to heatsy...@googlegroups.com
I was also thinking about using lower voltage wire and running it through plastic tubing to provide the full voltage protection. Or even if we are within acceptable limits on the wire, just adding the tubing as part of the high voltage safety equipment. Not all of these are flexible enough for tubing.
Science part:
The Cockcroft-Walton voltage multiplier is really a voltage adder. Each capacitor and diode stage stacks another step of the input voltage on top of what all the previous stages have built. So unlike a true multiplier, where the first stage would double the voltage, the second stage would be twice that or f times the input, and a third stage double the second or 8 times the original. If that kind of multiply by two for every stage was available, the parts for the final stage would have to be rated for nearly the full voltage, where as the Cockcroft-Walton requires many more stages to get a large voltage increase, but the components (diode and capacitor) to be much lower voltage rating and so cheaper and easier to find.
The above table of dialectric strength (break down voltage) shows that if we used the RG-8 we would need to make sure that the dialectric needs to be one with a high dialectric strength, and also it was pointed out that it should not be the foamed in plastic as that is more susceptible to point arc through. Not on this list but something I have seen elsewhere is that PVC has a high dialectric strength. But glass is even higher.
Mike
http://file.yizimg.com/305304/2012060809264490.pdf
TABLE 2 Dielectric Strength
A
Data Summary From Four Laboratories
Material Thickness Dielectric Strength (V/mil) Standard Coefficient of
(in. nom.) mean max min Deviation Variation (%)
Polyethylene
Terephthalate 0.001 4606 5330 4100 332 7.2
Polyethylene
Terephthalate 0.01 1558 1888 1169 196 12.6
Fluorinated
Ethylene
Propylene 0.003 3276 3769 2167 333 10.2
Fluorinated
Ethylene
Propylene 0.005 2530 3040 2140 231 9.1
PETP fiber
reinforced
epoxy resin 0.025 956 1071 783 89 9.3
PETP fiber
reinforced
epoxy resin 0.060 583 643 494 46 7.9
Epoxy-Glass
Laminate 0.065 567 635 489 43 7.6
Crosslinked
Polyethylene 0.044 861 948 729 48 5.6
Average 8.7
Mike Bushroe
Honeywell Glendale
.As for the high voltage source itself, as Jasper pointed out the Cockcroft-Walton voltage multiplier is simple to build, keeps the components from see the full, final voltage so that lower voltage, much cheaper parts can be used. With 5KV capacitors fairly inexpensive, and microwave oven capacitors cheap and easy to get but rather low in peak voltage we should be able to find affordable parts to build up the voltage:Science part:
The Cockcroft-Walton voltage multiplier is really a voltage adder. Each capacitor and diode stage stacks another step of the input voltage on top of what all the previous stages have built. So unlike a true multiplier, where the first stage would double the voltage, the second stage would be twice that or f times the input, and a third stage double the second or 8 times the original. If that kind of multiply by two for every stage was available, the parts for the final stage would have to be rated for nearly the full voltage, where as the Cockcroft-Walton requires many more stages to get a large voltage increase, but the components (diode and capacitor) to be much lower voltage rating and so cheaper and easier to find.
The above table of dialectric strength (break down voltage) shows that if we used the RG-8 we would need to make sure that the dialectric needs to be one with a high dialectric strength, and also it was pointed out that it should not be the foamed in plastic as that is more susceptible to point arc through. Not on this list but something I have seen elsewhere is that PVC has a high dialectric strength. But glass is even higher.
Mike
Larry Campbell
May 29, 2014, 5:46:12 PM5/29/14
to heatsy...@googlegroups.com
I spoke to Tom at Chemlab to see about locally procuring some deuterium oxide and some of the other parts we need... he is walking distance from ACME....
Larry Campbell
May 29, 2014, 11:20:36 PM5/29/14
to heatsy...@googlegroups.com
Reading
up on ripple factor ( about now I really would like to pick Fred Sanfords brain since he was THE world renowned expert on ripple....)
It was suggested we could use homemade caps, dont think thats a viable option yet...such as the HV makezine caps I read up on were actually wet leyden jars and were in the nanofarads... good for voltage ratings needed tho....
δV = (I/f *C) {n (n+1)/2} where I = load current in amperes, f = frequency in Hz,
C= Capacitance in farad, n1 = No. of capacitor = 2×No. of stages = 2n (n = No. of stages).
So that, Voltage drop ∆ V = (I/fc) (2/3 n³ + n²/2-n/6)
Regulation of voltage = V/2nEm, Ripple (%) =δ V/2nEm
Calculation of optimum no. of stages for minimum voltage drop / minimum regulation optimum
= √ (Em f C/I).
The size of capacitors used in cockcroft-walton type multiplier circuits is directly proportional to the frequency of input signal.
Caps used in off line, 50-60 Hz application appear to usually in the range of 1.0 to 200 microfarad.....
Those used in high frequency applications, 10+ kHz are typically in the range of 0.02 to 0.06 microfarads
Caps used in off line, 50-60 Hz application appear to usually in the range of 1.0 to 200 microfarad.....
Those used in high frequency applications, 10+ kHz are typically in the range of 0.02 to 0.06 microfarads
It was suggested we could use homemade caps, dont think thats a viable option yet...such as the HV makezine caps I read up on were actually wet leyden jars and were in the nanofarads... good for voltage ratings needed tho....
Mike Bushroe
May 30, 2014, 2:06:05 AM5/30/14
to heatsy...@googlegroups.com
Excellent research there Larry. I ran into that equation today while browsing to find typical capacitor values. But I wasn't willing to dig in to it and try to calculate the size caps we needed. I was still hoping to find somewhere that someone else had done the heavy lifting.
One note, if we really need 30mA into the negative grid on the fusor, at 30KV, that comes to 900 watts of power! The neon sign transformers I am seeing put out 30mA, but at only 7 to 15KV, and that is a half to a quarter the power that we need. We can try a 15KV, 30mA transformer and do a full wave rectifier and voltage doubler to get 30KV DC, but by conservation of energy, that will at best be 15mA, and probably circuit losses will bring that down to 10 to 12mA. That may still be enough for a modest sized fusor, but not the final solution that we want to build up to.
The microwave transformers are good for 700 to 1200 watts delivered, so they would be powerful enough. But even 7KV capacitors in the multiple microfarad range are usually rated in pico farads. It is easy to caps from China that are 0.01uF at 10KV+ for $3 a piece or even less. But the first 10KV 0.1uF capacitor I found is $22 a piece! And that is still one to two orders of magnitude short of what we need for 60Hz input.
So it looks like we will have to chop the power to up the frequency so that we can afford the capacitors. And that probably means AC to DC conversion first, filter capacitors, then 10 to 100Khz chopping and feeding into the voltage multiplier. The big question is how high a voltage do we want the intermediate DC to be? If we could run the chopper directly off a microwave transformer, we would have the power we need, a transformer designed for its input frequency, and the ability to chop the output to make the higher frequency at a voltage we wouldn't need too many sytages to build up to 30KV. However, the best switches for high voltage, high frequency are either MOSFETs, IGBTs, or stacked high voltage JFETs with a standard voltage MOSFET at the bottom, probably to provide rapid cut off. MOSFETs rated to only 4.5KV are $15-$50 and IGBTs of only 4KV are about the same. Where as 500V+ MOSFETs that handle 2A+ are $1 a piece. Unfortunately, this means that a second transformer is needed, that works at both high frequency and high voltage. Iron core transformers, usually the choice for power transformers lose efficiency even at 400Hz, let alone 10Khz. Most high frequency transformers are air-core, which does not have eddie current or hysteresis losses, but tend to have poor magnetic coupling, and are often used for low voltage, audio applications. High frequency, high voltage transformers are sometimes used by ham radio operators, but may be heard to find on short notice. Another possible source is medical or dental X-ray machines. The medical ones tend to have higher voltages and currents and should be able to directly supply the voltage a current we are looking for. The dental ones tend to be lower in voltage and power, but should be good for a modest sized fusor.
Clearly more research is required to find the best combination of transformer, capacitor and switching circuits.
Mike
I begin to think that we should try a microwave oven transformer at 400Hz and see if it provides the same high voltage and doesn't over heat when loaded, then use the 0.1uF capacitors to multiply the 4KV up to 16KV to start. But I am beginning to wonder how high school teams working on a shoe string budget manage to get 30KV at 30mA power supplies!
One note, if we really need 30mA into the negative grid on the fusor, at 30KV, that comes to 900 watts of power! The neon sign transformers I am seeing put out 30mA, but at only 7 to 15KV, and that is a half to a quarter the power that we need. We can try a 15KV, 30mA transformer and do a full wave rectifier and voltage doubler to get 30KV DC, but by conservation of energy, that will at best be 15mA, and probably circuit losses will bring that down to 10 to 12mA. That may still be enough for a modest sized fusor, but not the final solution that we want to build up to.
The microwave transformers are good for 700 to 1200 watts delivered, so they would be powerful enough. But even 7KV capacitors in the multiple microfarad range are usually rated in pico farads. It is easy to caps from China that are 0.01uF at 10KV+ for $3 a piece or even less. But the first 10KV 0.1uF capacitor I found is $22 a piece! And that is still one to two orders of magnitude short of what we need for 60Hz input.
So it looks like we will have to chop the power to up the frequency so that we can afford the capacitors. And that probably means AC to DC conversion first, filter capacitors, then 10 to 100Khz chopping and feeding into the voltage multiplier. The big question is how high a voltage do we want the intermediate DC to be? If we could run the chopper directly off a microwave transformer, we would have the power we need, a transformer designed for its input frequency, and the ability to chop the output to make the higher frequency at a voltage we wouldn't need too many sytages to build up to 30KV. However, the best switches for high voltage, high frequency are either MOSFETs, IGBTs, or stacked high voltage JFETs with a standard voltage MOSFET at the bottom, probably to provide rapid cut off. MOSFETs rated to only 4.5KV are $15-$50 and IGBTs of only 4KV are about the same. Where as 500V+ MOSFETs that handle 2A+ are $1 a piece. Unfortunately, this means that a second transformer is needed, that works at both high frequency and high voltage. Iron core transformers, usually the choice for power transformers lose efficiency even at 400Hz, let alone 10Khz. Most high frequency transformers are air-core, which does not have eddie current or hysteresis losses, but tend to have poor magnetic coupling, and are often used for low voltage, audio applications. High frequency, high voltage transformers are sometimes used by ham radio operators, but may be heard to find on short notice. Another possible source is medical or dental X-ray machines. The medical ones tend to have higher voltages and currents and should be able to directly supply the voltage a current we are looking for. The dental ones tend to be lower in voltage and power, but should be good for a modest sized fusor.
Clearly more research is required to find the best combination of transformer, capacitor and switching circuits.
Mike
I begin to think that we should try a microwave oven transformer at 400Hz and see if it provides the same high voltage and doesn't over heat when loaded, then use the 0.1uF capacitors to multiply the 4KV up to 16KV to start. But I am beginning to wonder how high school teams working on a shoe string budget manage to get 30KV at 30mA power supplies!
Larry Campbell
May 30, 2014, 2:52:34 AM5/30/14
to heatsy...@googlegroups.com
Comments inline:
On Thursday, May 29, 2014 11:06:05 PM UTC-7, Mike Bushroe wrote:
Excellent research there Larry. I ran into that equation today while browsing to find typical capacitor values. But I wasn't willing to dig in to it and try to calculate the size caps we needed. I was still hoping to find somewhere that someone else had done the heavy lifting.
Im much more than a hunk with pretty hair....(::hair toss::) a while of reading pages after pages and I found some guys in india wrote a paper on it...(hence the 50hz reference) I was hoping they were unusually conservative and you would correct it but my look at the math didnt indicate they could be THAT far off for my dreams to come true ....
One note, if we really need 30mA into the negative grid on the fusor, at 30KV, that comes to 900 watts of power! The neon sign transformers I am seeing put out 30mA, but at only 7 to 15KV, and that is a half to a quarter the power that we need. We can try a 15KV, 30mA transformer and do a full wave rectifier and voltage doubler to get 30KV DC, but by conservation of energy, that will at best be 15mA, and probably circuit losses will bring that down to 10 to 12mA. That may still be enough for a modest sized fusor, but not the final solution that we want to build up to.
Ill check mines output... its older so its possible its much higher.than 30mA but I sincerely doubt its THAT high..... Ive had them for over 20 years...
The microwave transformers are good for 700 to 1200 watts delivered, so they would be powerful enough. But even 7KV capacitors in the multiple microfarad range are usually rated in pico farads. It is easy to caps from China that are 0.01uF at 10KV+ for $3 a piece or even less. But the first 10KV 0.1uF capacitor I found is $22 a piece! And that is still one to two orders of magnitude short of what we need for 60Hz input.
thus my willingness to fabricate them...maybe I should just grab some of staceys mason jars and give-er-a-go... the voltage rating would suffice, its just surface area... maybe we should consider plate glass stacks... a few guys in the 50s did that from what I read when I was a kid....(amazing the things I remember reading when I was not even a teenager yet)
So it looks like we will have to chop the power to up the frequency so that we can afford the capacitors. And that probably means AC to DC conversion first, filter capacitors, then 10 to 100Khz chopping and feeding into the voltage multiplier. The big question is how high a voltage do we want the intermediate DC to be? If we could run the chopper directly off a microwave transformer, we would have the power we need, a transformer designed for its input frequency, and the ability to chop the output to make the higher frequency at a voltage we wouldn't need too many sytages to build up to 30KV. However, the best switches for high voltage, high frequency are either MOSFETs, IGBTs, or stacked high voltage JFETs with a standard voltage MOSFET at the bottom, probably to provide rapid cut off. MOSFETs rated to only 4.5KV are $15-$50 and IGBTs of only 4KV are about the same. Where as 500V+ MOSFETs that handle 2A+ are $1 a piece. Unfortunately, this means that a second transformer is needed, that works at both high frequency and high voltage. Iron core transformers, usually the choice for power transformers lose efficiency even at 400Hz, let alone 10Khz. Most high frequency transformers are air-core, which does not have eddie current or hysteresis losses, but tend to have poor magnetic coupling, and are often used for low voltage, audio applications. High frequency, high voltage transformers are sometimes used by ham radio operators, but may be heard to find on short notice. Another possible source is medical or dental X-ray machines. The medical ones tend to have higher voltages and currents and should be able to directly supply the voltage a current we are looking for. The dental ones tend to be lower in voltage and power, but should be good for a modest sized fusor.
I like IGBTs but the price is higher... 500v mosfets have dropped drastically since the UPS manufacturers went IGBT a few years back.......
too bad iron core transformer saturations are not fast enough response to warrant that as an option I have some monster units... as in 1 ton+....
whatever happened to the xray machine we had offered as a donation that I was so hot and bothered about accepting? I KNEW it would come in handy (and people wonder why I hoard tech stuff)
::sigh:: Im thinking we build what we can muster and plug it in and let rip.....should be good for a little while before it goes pop like a rabid weasel....
Jasper Nance
May 30, 2014, 9:43:23 AM5/30/14
to heatsy...@googlegroups.com
The 12kV tesla coil transformer has been modified by removing the magnetic shunts. It can output 180mA instead of the 30mA it was designed for. It does have to be submersed in oil (which it is).
It could be used here if needed.
--
You received this message because you are subscribed to the Google Groups "HeatSync Labs" group.
To unsubscribe from this group and stop receiving emails from it, send an email to heatsynclabs...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Jasper Nance
May 30, 2014, 9:45:29 AM5/30/14
to heatsy...@googlegroups.com
you will not get 10uF out of stacked glass plates that will fit in the building nor layden jars. You will need rolled or stacked mica, with precision cut plates to get this kind of performance. There's a reason caps are expensive.
Are you sure you can't live with the DC ripple?
Larry Campbell
May 30, 2014, 10:41:24 AM5/30/14
to heatsy...@googlegroups.com
I am confident we can tolerate some ripple.. it is something I am not certain about how much we can tolerate before it interferes with the process...
Since we are in the much higher voltages that the minimal required range I would expect it would be comparable to putting a hemi that had a broken spark plug into a small passenger car and wanting to go to the grocery store... it will probably still outrun anything on the block but will sound like hell and make everyone watching very nervous....
I did pick up a few thousand diodes, so we have spares if we fry a few....I wont cry loudly....
On Friday, May 30, 2014 6:45:29 AM UTC-7, Nebarnix wrote:
you will not get 10uF out of stacked glass plates that will fit in the building nor layden jars. You will need rolled or stacked mica, with precision cut plates to get this kind of performance. There's a reason caps are expensive.Are you sure you can't live with the DC ripple?
Jasper Nance
May 30, 2014, 10:49:39 AM5/30/14
to heatsy...@googlegroups.com
What is the current handling capability of the diodes that you ordered?
Larry Campbell
May 30, 2014, 11:07:44 AM5/30/14
to heatsy...@googlegroups.com
one amp....
On Friday, May 30, 2014 7:49:39 AM UTC-7, Nebarnix wrote:
On Friday, May 30, 2014 7:49:39 AM UTC-7, Nebarnix wrote:
What is the current handling capability of the diodes that you ordered?
Mike Bushroe
May 30, 2014, 4:06:54 PM5/30/14
to heatsy...@googlegroups.com
Larry,
How thick is the mylar that you have so many rolls of? I just cam up with an idea to solve the conductive film too close to the edge and arcing over.
I believe that just about any chemical bath that will etch copper PCB boards will also remove the aluminum. At the very least I know from prior experience that caustic lye in water will break down the aluminum (releasing hydrogen from the water), and hydrogen peroxide and muriatic acid mixed 50-50 will attack aluminum faster than it removes copper! We could cut the mylar into rolled up capacitor size, paint a little etchant on the edges, the add the electrical tabs, possibly as the capacitor is being rolled. If the mylar is 2mill (0.0508 millimeters) thick, then it will have better than 10KV dielectric strength. Putting the two sheets with the aluminized surfaces up and stacking one on top of the other and then rolling it up there will always be an non-conductive layer between each conductive layer. Assuming that a spiral rolled capacitor has twice the capacitance of a pair of flat plates since they have opposite voltage plates both above and below, it looks like 12 feet by 1 foot initial sheets would give 1.3uF at 10KV, 96 feet by 1 foot or 48 feet by 2 foot would give somewhere in the range of 10uF at 10KV. With the multi-tap construction you mentioned the inductance should be reduced and it will work at 60hz, probably at 400hz, and maybe at 10Khz.
Of course, if the mylar sheets you have are only 1 mil thick, we will have to find enough non-conductive mylar to put in between each sheet. Still, that might be the easiest way to get fairly large caps at the voltages we want and the prices we think we can afford. And I think the 12 inch high model will be easiest to work with. I am estimating around 2 1/2 inches in diameter if rolled very tight and with no inner core to make winding easier. If we start with a 1/2 PVC pipe, the result will fit within a 4 inch PVC pipe which will provide all the voltage protection we might need, pot the space in between with epoxy or something else that will help carry any waste heat away, attach the internal electrodes to machine screws, glue them to the PVC pipe cover we would have a strong, safe, high voltage, moderately high capacitance capacitor. If the diode strings use a metal tab with whole for their connections, we can just bolt the diodes strings between the caps to make our voltage multipliers and full wave rectifiers. We just need to build about 9 of them for the rectifier and two stage multiplier.
Mike
Larry Campbell
May 30, 2014, 4:23:53 PM5/30/14
to heatsy...@googlegroups.com
Ill have to measure the mylar however I am not opposed to putting another layer inbetween if needed.... and I have a LOT of the mylar...about 160 lb spool!
Michael Mathers
May 31, 2014, 12:27:14 AM5/31/14
to heatsy...@googlegroups.com
Larry, you are good with about 400-500 watts into the grid. 30 mA is too high at fusing conditions but you may need close to that in earlier operating conditions at lower voltage.
35 KV at 15 mA will get you easily detectable fusion. However, early in operation (prior to fusion) you may need > 20 at lower voltage and higher vacuum pressure. But bottom line is 400-500 Watts. You need to hit that in variable current/voltage conditions targeting 30-35 KV @ 10-15 mA.
In my own research I'm coming to the same conclusion with the NSTs. It will be difficult, although not impossible, to achieve this with NSTs. I have a 9 KV, 60 mA NST that I was planning on using. However, upon learning more about how NSTs are configured I'm not sure it's a road I want to continue to go down. NSTs are magnetically shunted. This is the technique used to effectively current limit the transform at the rated current. Fuck me if I really understand what that means but essentially it means the label of 9000 V, 60mA does not mean what we normally read that to mean. NST cannot sustain their 9000V at full load. The study I've seen show approximately 2/3 power at full load. Under the designed load this makes a lot of sense (neon tubes). A high voltage is initially required to ionize the tube but afterwards a much lower voltage and sustaining current is all that is needed. That's the first problem.
35 KV at 15 mA will get you easily detectable fusion. However, early in operation (prior to fusion) you may need > 20 at lower voltage and higher vacuum pressure. But bottom line is 400-500 Watts. You need to hit that in variable current/voltage conditions targeting 30-35 KV @ 10-15 mA.
In my own research I'm coming to the same conclusion with the NSTs. It will be difficult, although not impossible, to achieve this with NSTs. I have a 9 KV, 60 mA NST that I was planning on using. However, upon learning more about how NSTs are configured I'm not sure it's a road I want to continue to go down. NSTs are magnetically shunted. This is the technique used to effectively current limit the transform at the rated current. Fuck me if I really understand what that means but essentially it means the label of 9000 V, 60mA does not mean what we normally read that to mean. NST cannot sustain their 9000V at full load. The study I've seen show approximately 2/3 power at full load. Under the designed load this makes a lot of sense (neon tubes). A high voltage is initially required to ionize the tube but afterwards a much lower voltage and sustaining current is all that is needed. That's the first problem.
Problem #2: Most modern(ish) NSTs are Secondary Ground Fault Protected (ie, local GFI). That needs to either be disabled or the entire transformer needs to be capacitatively isolated (large caps that can support the entire secondary output). Problem #3: Barring that....Any NST 6 KV or over has the center tap grounded to the case chassis. This makes it problematic to get the full output out of the transformer. If you use the transformer end to end the chassis will be floating at (in my case) at 6-7 KV. Not nice (although I may do it anyway). Also, unless the GFI is disabled nor can you just take 120 Hz pulse DC at half the voltage (ground the center and rectify each half) w/o tripping the GFI (it sees unequal load on each of the secondary and trips).
Unless you want to get into switched capacitor voltage multiplication it will be difficult to get to 35 KV at 15 mA.
Unless you want to get into switched capacitor voltage multiplication it will be difficult to get to 35 KV at 15 mA.
I'm all ears if you have any ideas on how to get a hold of Xray transformers but that doesn't seem particularly appealing to me either (was hoping to avoid lots of oil bath).
I know nothing about MOTs so please share if you have any experience with them.
Michael
I know nothing about MOTs so please share if you have any experience with them.
Michael
Jasper Nance
May 31, 2014, 12:31:32 AM5/31/14
to heatsy...@googlegroups.com
Well, first of all no transformer in the world can sustain its open circuit voltage while under short circuit conditions. Second, you can remove the magnetic shunts like I did with the tesla coil NST. Third, you can find old skool NSTs without the GFI fairly easily still. I have three 12kV 60mA models and a 15kV 30mA core.
Is there any reason why we don't use flyback transformers instead? they're much smaller, quite powerful, and can work with much smaller multiplier caps.
--
You received this message because you are subscribed to the Google Groups "HeatSync Labs" group.
To unsubscribe from this group and stop receiving emails from it, send an email to heatsynclabs...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Larry Campbell
May 31, 2014, 1:15:59 AM5/31/14
to heatsy...@googlegroups.com
...my transformers are well over 20 years old... no gfi protection..
Regardless Mike Bushroe came over to ACME again and we went over a number of things including an alternate design for a dc supply theoretically capable of 60kV at 5mA...that should be easily bumped up 50% or to inspire a beefier design to be right between the 400-500 watts required...
Michael,
you have any schematics with reasonably priced components requirements?
all,
anyone know what became of the xray machine that was offered up for a donation that I was all hot and bothered about to accept? my hoarding instinct usually pay off in spades...
On Friday, May 30, 2014 9:31:32 PM UTC-7, Nebarnix wrote:
Well, first of all no transformer in the world can sustain its open circuit voltage while under short circuit conditions. Second, you can remove the magnetic shunts like I did with the tesla coil NST. Third, you can find old skool NSTs without the GFI fairly easily still. I have three 12kV 60mA models and a 15kV 30mA core.Is there any reason why we don't use flyback transformers instead? they're much smaller, quite powerful, and can work with much smaller multiplier caps.
On Fri, May 30, 2014 at 9:27 PM, Michael Mathers <
Larry, you are good with about 400-500 watts into the grid. 30 mA is too high at fusing conditions but you may need close to that in earlier operating conditions at lower voltage.
35 KV at 15 mA will get you easily detectable fusion. However, early in operation (prior to fusion) you may need > 20 at lower voltage and higher vacuum pressure. But bottom line is 400-500 Watts. You need to hit that in variable current/voltage conditions targeting 30-35 KV @ 10-15 mA.
In my own research I'm coming to the same conclusion with the NSTs. It will be difficult, although not impossible, to achieve this with NSTs. I have a 9 KV, 60 mA NST that I was planning on using. However, upon learning more about how NSTs are configured I'm not sure it's a road I want to continue to go down. NSTs are magnetically shunted. This is the technique used to effectively current limit the transform at the rated current. Fuck me if I really understand what that means but essentially it means the label of 9000 V, 60mA does not mean what we normally read that to mean. NST cannot sustain their 9000V at full load. The study I've seen show approximately 2/3 power at full load. Under the designed load this makes a lot of sense (neon tubes). A high voltage is initially required to ionize the tube but afterwards a much lower voltage and sustaining current is all that is needed. That's the first problem.Problem #2: Most modern(ish) NSTs are Secondary Ground Fault Protected (ie, local GFI). That needs to either be disabled or the entire transformer needs to be capacitatively isolated (large caps that can support the entire secondary output). Problem #3: Barring that....Any NST 6 KV or over has the center tap grounded to the case chassis. This makes it problematic to get the full output out of the transformer. If you use the transformer end to end the chassis will be floating at (in my case) at 6-7 KV. Not nice (although I may do it anyway). Also, unless the GFI is disabled nor can you just take 120 Hz pulse DC at half the voltage (ground the center and rectify each half) w/o tripping the GFI (it sees unequal load on each of the secondary and trips).
Unless you want to get into switched capacitor voltage multiplication it will be difficult to get to 35 KV at 15 mA.I'm all ears if you have any ideas on how to get a hold of Xray transformers but that doesn't seem particularly appealing to me either (was hoping to avoid lots of oil bath).
I know nothing about MOTs so please share if you have any experience with them.
Michael
--
You received this message because you are subscribed to the Google Groups "HeatSync Labs" group.
To unsubscribe from this group and stop receiving emails from it, send an email to heatsynclabs+unsubscribe@googlegroups.com.
--
----------------------------------------------
Jasper Nance - KE7PHI
Creative and Scientific Imagery
http://www.nebarnix.com/
Mike Bushroe
May 31, 2014, 2:11:50 AM5/31/14
to heatsy...@googlegroups.com
Jasper,
Smaller, powerful and smaller multiplier caps sounds good to me, too. However, I have never worked with flyback transformers, so I will have to do a little but of research.
If we have to stay with 60hz voltage multipliers, we can use the aluminized plastic film that Servo gave some of us awhile back. However, it is polypropylene, not biaxially-oriented polyethylene terephthalate (mylar). This means that the dielectric constant is lower (not much of a problem) and the dielectric strength is more than 10 times lower (big problem). The 7 mil sheets of PP break down at 3.5KV. If we doubled the sheets by laying pairs face to face to get 14 mil between conductors it raises the breakdown voltage to only 7KV. To get up to a useful voltage 15KV we would still need to add a layer of plastic without conductive coating. And if we do that, we might as well use real mylar and get the high voltage from a thinner film, which will allow higher capacitance per square foot, as well as volumetrically smaller caps for the same voltage/capacitance. If we found 10 mil mylar, we could make caps good to 70KV, and use them on the output of the voltage multipliers to greatly reduce the ripple that small caps and slow frequencies cause.
Mike
Smaller, powerful and smaller multiplier caps sounds good to me, too. However, I have never worked with flyback transformers, so I will have to do a little but of research.
If we have to stay with 60hz voltage multipliers, we can use the aluminized plastic film that Servo gave some of us awhile back. However, it is polypropylene, not biaxially-oriented polyethylene terephthalate (mylar). This means that the dielectric constant is lower (not much of a problem) and the dielectric strength is more than 10 times lower (big problem). The 7 mil sheets of PP break down at 3.5KV. If we doubled the sheets by laying pairs face to face to get 14 mil between conductors it raises the breakdown voltage to only 7KV. To get up to a useful voltage 15KV we would still need to add a layer of plastic without conductive coating. And if we do that, we might as well use real mylar and get the high voltage from a thinner film, which will allow higher capacitance per square foot, as well as volumetrically smaller caps for the same voltage/capacitance. If we found 10 mil mylar, we could make caps good to 70KV, and use them on the output of the voltage multipliers to greatly reduce the ripple that small caps and slow frequencies cause.
Mike
On Friday, May 30, 2014 9:31:32 PM UTC-7, Nebarnix wrote:
Mike Bushroe
Jun 7, 2014, 3:17:37 AM6/7/14
to heatsy...@googlegroups.com
First piece of fusor built!
https://www.flickr.com/photos/34355839@N06/14384430013/in/pool-heatsynclabs
It is just 10 diodes in a line and does even have its protective tubing or heat shrink applied but the construction is already underway! More high voltage diodes strings to come shortly.
Mike
https://www.flickr.com/photos/34355839@N06/14384430013/in/pool-heatsynclabs
It is just 10 diodes in a line and does even have its protective tubing or heat shrink applied but the construction is already underway! More high voltage diodes strings to come shortly.
Mike
Eric Hubert
Jun 17, 2014, 2:34:55 AM6/17/14
to heatsy...@googlegroups.com
Here a local supplier for mylar I found using google. It is 11:30 pm so I can't call them. My sleep schedule has switched to nights so I doubt I will be able to catch them tomorrow either. Is anyone able to follow up on this?
Oh, ironically when one does at google search for "10 mil mylar spool phoenix, az -balloon" this thread comes up third.
Larry Campbell
Jun 17, 2014, 9:07:04 AM6/17/14
to heatsy...@googlegroups.com
Ill see if I can get some time to call them later
Eric Hubert
Jun 17, 2014, 11:53:52 AM6/17/14
to heatsy...@googlegroups.com
I took a quick 7 hour nap at the lab. Since I am up during the day I gave them a call. After speaking to them for a few I felt rather dumb since I didn't know all the specs of what we needed. They do have mylar. They are local. They do not stock a lot of it. It comes in clear or semiclear. They have it in sheets and rolls. It can come in any size or width we want. They have data sheets on their website.
Mike Bushroe
Jun 17, 2014, 3:05:35 PM6/17/14
to heatsy...@googlegroups.com
Eric,
Thanks much for running down that data for us. Aaron gave me the web address awhile back and I took a quick look but didn't call to get more details.
If anyone is checking what is available, we want 2mil and 10 mil with or without aluminum coating on one side, clear/semi-clear don't care.
We need two kinds of capacitors for a 60hz system. 2 mil mylar that can be aluminimized on one side or not, and 10mil mylar that also can be aluminized or not. The completely clear enough to see through or semiclear is not important for this work, so we can take either of those that they have in stock and is less expensive. The only requirement is that it either be brand name mylar or the exact same polymer treated to the exact same process. Mylar seems to be fairly unique in how high it's break down voltage is, and most other plastics have much lower dielectric strengths dropping from mylar's 7,000V per mil to 500 V per mil. With more than an order of magnitude difference, if we put the wrong plastic in it will fail, possibly spectacularly the very first time we apply full voltage to
As for lengths and widths, 4 to 5 feet by 100 is best, 2 mills 3 or more rolls of 4x100, 10 mils one roll may be enough .
The current plan is to build capacitors, in both types we are thinking of 1 foot by 100 foot sheets rolled up to make something 3 to 5 inches in diameter. But remember that it will take two sheets for each to keep the two layers of conductors apart. Also, we will need to etch off about 1 inch of aluminum on the edges of every sheet that comes pre-auminized so that the high voltage won't short across. That means that a 12 inch strip from a 48 inch wide roll will result in 10 inch wide capacitor plates, and that will be slightly below the numbers I was calculating. We can either split only 3 strips off a 48 inch roll and get caps that are 16 inches tall and higher capacity than calculated but only get 1 1/2 caps per roll. Or we can make two whole caps from one sheet, but with smaller capacitance. If we get mylar without the aluminum coating we will also need to buy lots of aluminum foil. The foil will be a greater thickness and probably make better caps with lower internal resistance.
For the TV horizontal flyback transformer will not need any 2 mil caps, but will really need the 10 mil caps to smooth out the saw toothed horizontal sweep which may reach 30KV at its peal, bu will average more like 15K. Also, if I get the microwave oven inverter which works at 20 to 50 Khz switching speed, we also won't need the giant 2 mil caps, the little 10,000pF caps from china will do the job just fine. But we will still need the 10 mill ones to reduce the hopefully much smaller ripple.
Mike
Eric Hubert
Jun 18, 2014, 7:15:09 AM6/18/14
to heatsy...@googlegroups.com
If I understand the making of capacitors correctly the mylar would be rolled along with another material. It would be like taking red Christmas paper and green Christmas paper and rolling them together. So if we find clear or semi clear mylar we'd have to add aluminum foil in a width two inches smaller than the width of the mylar? Would both the mylar and the aluminum have to be two inches smaller than the "other material"? That site had varying widths of 10 mil mylar. As far as 2 mil went I saw a bunch of either ebay or amazon. I believe it was amazon. Mostly it was covers for books but I did see rolls as well. I also glanced at the shipping and in many cases it was included. I am still unsure what exact specification are needed or I'd try to find a direct link. I'd love to see this built and built with as few snags and as most fun as possible.
Mike Bushroe
Jun 18, 2014, 4:23:01 PM6/18/14
to heatsy...@googlegroups.com
Eric,
Short answer: Etch the 2 mil, aluminized, don't bother etching with the 10 mil, just make sure the conductor is cut to be 2" or more narrower than the 10 mil.
Yes, that parallels my research. I have seen short lengths of thick mylar that is used for archive books (not sure what the process is, but that is what the hype it for). I also see LOTS of posts on ebay for 2 mil mylar that is thoroughly coated on one side for 'hydroponics'. One even mentioned how little light leaks through to disturb your sleep (if you sleep while the grow lights are on), and most how efficiently they reflect the grow light energy back into the plants for optimum efficiency.
I have not tested, but believe that the aluminum coating on one side of the 'hydroponic' mylar would be enough to conduct current, if 'taps' are added every few feet to help collect the current and improve the frequency response. And that was a good analogy, yes with the 2 mil, reflective mylar we would take two sheets, like the red and green christmas wrapping paper, wind it tightly around something like a piece of PVC to avoid trying to wrap the exact center, put wires or metal strips in with the winding every few feet with all the wires for the red sticking out on one side and all the wires for the green on the other. Then just solder the wires together, put inside some container like a larger PVC tube, add outside contacts.
Using the pre-coated mylar would require etching away the aluminum that is within an inch or two of the edge of the mylar to prevent arching over. That should be all the protection that we need, but filling the container with mineral oil makes it even safer.
For the thicker mylar, it usually does not come with aluminum on one side, so once again as you said, we would have to add two rolls of aluminum foil to the two rolls of mylar we are rolling up. We could mix the 2 mil with aluminum and the 10 mil without to get 12 mils between layers, but if we have to add conductive plates to the roll, the foil would be much thicker and a better conductor. So I think the answer to your question is that the aluminum foil would be cut to be 2" or more narrower than the 10 mil mylar, but if we use foil instead of aluminized 2 mil mylar, we would not have to cut two things to different widths. If we did use the 2 mil mylar for its conductive surface on one side, we would still cut the 2 mil to be 2" smaller than the 10 mil and not worry about it arcing over because the wider 10 mil provides 70KV protection all by itself. Keeping both sizes of mylar the same width and etching away the aluminum from the edges would give us an extra 14KV of protection, but frankly, on top of 70KV I don't see us ever needing that much. Unless Jasper might want it for the SEM or the Tesla coil.
At this point, with the TV set horizontal high voltage sweep power supply being our starter, and the hoped for high frequency MOV inverter being next, I think we are going to nee the high voltage caps to smooth out the ripple before we need large medium high voltage caps for voltage multipliers working at 60hz.
Mike.
Short answer: Etch the 2 mil, aluminized, don't bother etching with the 10 mil, just make sure the conductor is cut to be 2" or more narrower than the 10 mil.
Yes, that parallels my research. I have seen short lengths of thick mylar that is used for archive books (not sure what the process is, but that is what the hype it for). I also see LOTS of posts on ebay for 2 mil mylar that is thoroughly coated on one side for 'hydroponics'. One even mentioned how little light leaks through to disturb your sleep (if you sleep while the grow lights are on), and most how efficiently they reflect the grow light energy back into the plants for optimum efficiency.
I have not tested, but believe that the aluminum coating on one side of the 'hydroponic' mylar would be enough to conduct current, if 'taps' are added every few feet to help collect the current and improve the frequency response. And that was a good analogy, yes with the 2 mil, reflective mylar we would take two sheets, like the red and green christmas wrapping paper, wind it tightly around something like a piece of PVC to avoid trying to wrap the exact center, put wires or metal strips in with the winding every few feet with all the wires for the red sticking out on one side and all the wires for the green on the other. Then just solder the wires together, put inside some container like a larger PVC tube, add outside contacts.
Using the pre-coated mylar would require etching away the aluminum that is within an inch or two of the edge of the mylar to prevent arching over. That should be all the protection that we need, but filling the container with mineral oil makes it even safer.
For the thicker mylar, it usually does not come with aluminum on one side, so once again as you said, we would have to add two rolls of aluminum foil to the two rolls of mylar we are rolling up. We could mix the 2 mil with aluminum and the 10 mil without to get 12 mils between layers, but if we have to add conductive plates to the roll, the foil would be much thicker and a better conductor. So I think the answer to your question is that the aluminum foil would be cut to be 2" or more narrower than the 10 mil mylar, but if we use foil instead of aluminized 2 mil mylar, we would not have to cut two things to different widths. If we did use the 2 mil mylar for its conductive surface on one side, we would still cut the 2 mil to be 2" smaller than the 10 mil and not worry about it arcing over because the wider 10 mil provides 70KV protection all by itself. Keeping both sizes of mylar the same width and etching away the aluminum from the edges would give us an extra 14KV of protection, but frankly, on top of 70KV I don't see us ever needing that much. Unless Jasper might want it for the SEM or the Tesla coil.
At this point, with the TV set horizontal high voltage sweep power supply being our starter, and the hoped for high frequency MOV inverter being next, I think we are going to nee the high voltage caps to smooth out the ripple before we need large medium high voltage caps for voltage multipliers working at 60hz.
Mike.
Eric Hubert
Jun 20, 2014, 3:45:20 AM6/20/14
to heatsy...@googlegroups.com
When you say etch do you mean like taking the copper off a copper clad board to make a pcb?
Larry Campbell
Jun 20, 2014, 4:54:25 AM6/20/14
to heatsy...@googlegroups.com
Yep, chemically removing the aluminized surface from the edges of the polypropylene.
On Friday, June 20, 2014 12:45:20 AM UTC-7, Eric Hubert wrote:
When you say etch do you mean like taking the copper off a copper clad board to make a pcb?
Mike Bushroe
Jun 20, 2014, 9:41:51 AM6/20/14
to heatsy...@googlegroups.com
Eric,
I remember when being bored waiting for a slow PCB etch to get all the copper off both sides, that I experimented with what metals I could use the agitate the board with when I found that aluminum reacted much more strongly that then steel or brass or stainless. So yes, I am pretty sure that the initial etchant solution that I use, a 50-50 mix of muryatic acid (mild hydrochloric acid for swimming pools) and hydrogen peroxide will attach the thin film of aluminum around the edges. I just have been too busy working on other parts of the high voltage system, making some new and better glass extruder nozzles, the 3D print one will go into, and circuit diagram and board layout that grew much more compex than I had planned, plus all the usual stuff. I hope that I can prove the aluminum etching this weekend. I think that the ferric chloride commonly used probably will work too, but I kind of doubt that the Cupric Chloride etchant would do much.
Mike
I remember when being bored waiting for a slow PCB etch to get all the copper off both sides, that I experimented with what metals I could use the agitate the board with when I found that aluminum reacted much more strongly that then steel or brass or stainless. So yes, I am pretty sure that the initial etchant solution that I use, a 50-50 mix of muryatic acid (mild hydrochloric acid for swimming pools) and hydrogen peroxide will attach the thin film of aluminum around the edges. I just have been too busy working on other parts of the high voltage system, making some new and better glass extruder nozzles, the 3D print one will go into, and circuit diagram and board layout that grew much more compex than I had planned, plus all the usual stuff. I hope that I can prove the aluminum etching this weekend. I think that the ferric chloride commonly used probably will work too, but I kind of doubt that the Cupric Chloride etchant would do much.
Mike
Eric Hubert
Jun 20, 2014, 3:13:00 PM6/20/14
to heatsy...@googlegroups.com
It sounds like a very long process to wait for the chemicals to eat away the aluminum. We also run the risk of some of the aluminum not being eaten away and it being missed during inspection. With my limited knowledge of high voltage it would seem this could make a path of least resistance in a direction we may not want. At least those are my initial thoughts. It would seem that it would be easier to get plain mylar and use aluminum that is two inches narrower than the mylar. We could probably make a jig for aligning the three rolls of materials together and rolling it using Larry's lathe.
Eric Hubert
Jun 20, 2014, 3:25:07 PM6/20/14
to heatsy...@googlegroups.com
I through together a Solidworks model to show the spiraling of two sheets of material around a cylinder. The cylinder is not shown. I did it in case we wanted a visual reference for anything especially for explaining to people who don't do this stuff daily what it is we are doing. Just say the word and I will modify the model to suit.
Larry Campbell
Jun 20, 2014, 4:43:18 PM6/20/14
to heatsy...@googlegroups.com
Well it would be 4 layers, two conductive plates and two insulating layers that would be at least an inch if not two inches wider than the conductive plates to prevent arc over....... I will need to make a rack to hold 4 rolls and Im thinking that even if we use the lathe as a spindle to roll it up on we cannot do that under power because even at 15rpm we would need to stop it too often to put the taps very neatly at regular intervals
attached a quickly butchered rather overly simplified version of how en end of a cap would look like (ideally for a LOT of windings each end would be a different pole on the capacitor because there would be a LOT of taps to bond together...
Mike, how many wraps between taps and what gauge wire for each tap? Im thinking I should set up my lead melting pot for solder and dip each tap to tin the ends so we reduce the chance of having an issue of thermal conductivity compromising the dielectric layers on assembly
Eric Hubert
Jun 20, 2014, 5:02:01 PM6/20/14
to heatsy...@googlegroups.com
That's what I get for posting so soon after I get up. That should be threw instead of through. Would the layers go mylar-insulator-mylar-insulator or insulator-mylar-mylar-insulator?
Larry Campbell
Jun 20, 2014, 5:17:27 PM6/20/14
to heatsy...@googlegroups.com
Since the mylar in THIS instance is an insulator it would be insulator layer on the bottom of a few turns then begin conductive plate A, with the second insulator layer overlapping it by several inches to prevent arc over in the beginning core of the cap, with Conductive plate B beginning as close to the beginning of A , then final layer would have the mylar warp around multiple times to insulate the outer layer of the cal as well
with stops every X number of turns to put a tap wire off conductor A to one end and a tap off of conductor B to the other end, offsetting the taps so they neatly line up in a slow spiral so as to allow for them to be put together neatly
AND making sure the inside end of each tap is smoothed and rounded so it does not make a mini puncture to cause a short some people tape over that end to keep it located and extra protected too....else it go boom
AND making sure the outside end of each tap is pre-tinned while taking extra care to submerse the cap in something to keep it cool when soldering those taps together so as to not compromise the insulative layers they touch....
and....Biaxially-oriented polyethylene terephthalate ...say that a google times fast....
nobody said making caps was easy... its just cheaper than paying a hundred dollars or so per cap....
On Friday, June 20, 2014 2:02:01 PM UTC-7, Eric Hubert wrote:
That's what I get for posting so soon after I get up. That should be threw instead of through. Would the layers go mylar-insulator-mylar-insulator or insulator-mylar-mylar-insulator?
Eric Hubert
Jun 20, 2014, 6:29:52 PM6/20/14
to heatsy...@googlegroups.com
This is something that has physical shape and needs a process to create it. This is something that would be fun for me even if it is not easy. If the task is "take this sheet and that sheet put them together with x wires in a complicated process" I can help with that more than I can with planning the electrical stuff. When I can concentrate enough later I will see if there is a jig of sorts I can design to help with this.
Mike Bushroe
Jun 20, 2014, 7:31:10 PM6/20/14
to heatsy...@googlegroups.com
Eric,
In that case, I have a few 'ask for the Moon' requests.
Is it possible in Solid Works to 'peel back' the visible corner of the outside layer? If so can we split them first into red and green to show that there are two separate sheets. The split each sheet and label one as the sprayed on aluminum conductor and the other as the 2 mill thick insulation layer
The second would be to squeeze in strips of copper tape, two on the left touching the green side's aluminum and two on the right touching the red side's aluminum. That would probably require making a double spiral to show red on one spiral and green on the other. These tabs would then bend outward to become the + and - terminals. Although actually this type of capacitor does not have a polarity requirement, it makes it easier to understand.
Mike
In that case, I have a few 'ask for the Moon' requests.
Is it possible in Solid Works to 'peel back' the visible corner of the outside layer? If so can we split them first into red and green to show that there are two separate sheets. The split each sheet and label one as the sprayed on aluminum conductor and the other as the 2 mill thick insulation layer
The second would be to squeeze in strips of copper tape, two on the left touching the green side's aluminum and two on the right touching the red side's aluminum. That would probably require making a double spiral to show red on one spiral and green on the other. These tabs would then bend outward to become the + and - terminals. Although actually this type of capacitor does not have a polarity requirement, it makes it easier to understand.
Mike
Eric Hubert
Jun 20, 2014, 10:00:30 PM6/20/14
to heatsy...@googlegroups.com
Let me see what I can do. If I don't respond back in a day please remind me. I would have gotten sidetracked with some drama even though I do want to do this.
Michael Mathers
Jun 21, 2014, 2:10:20 AM6/21/14
to heatsy...@googlegroups.com
Honestly, I think stacked with a rolling pin and two 1/8" plates of acrylic with light bolts in the corners is way easier than rollling. The major capacitance variable is a distance between plates. This is tougher to keep this distance minimal in a rolled configuration. Further you easily slice up squares of aluminum (say 6" x 8") simply by cutting a 12" roll in half. A .1uF cap at 40KV would be 1 sheet of 5 mil mylar plus aluminum foil about 6x6" and about 70-80 mils thick. Add two 1/8" acrylic squares and the entire capacitor is < .4". One roll of 5mil x 24" x 10' roll of mylar from Tap Plastics ($31) will make six such capacitors (10 layers). Being square rectangular and flat it would be easy to seal and fill with mineral oil if you wished (not necessary though) and get higher working voltage. 6 flat capacitors of this nature would be 6"x8" (roughly) and 2.5" thick. Not large at al. 100nF from my simulations is enough to keep ripple at 35KV at 15mA or so at about 4 KV. Just fine for this purpose.
Eric Hubert
Jun 21, 2014, 2:28:48 AM6/21/14
to heatsy...@googlegroups.com
It just dawned on me the huge cap we have in the window is rectangular. I was distracted when I first looked at it but I know in the back of my mind I was thinking, "wait, the cap we are designing is round. Why is this one rectangular?" I believe this may be the reason why.
Eric Hubert
Jun 21, 2014, 11:10:54 AM6/21/14
to heatsy...@googlegroups.com
Is this the design we are making? http://makezine.com/projects/make-36-boards/nuclear-fusor/
Larry Campbell
Jun 21, 2014, 3:45:40 PM6/21/14
to heatsy...@googlegroups.com
No, at first Moheeb and I discussed many months earlier making the first one like that however what we are actually talking about making instead is much higher power and far better design.
On Saturday, June 21, 2014 8:10:54 AM UTC-7, Eric Hubert wrote:
Is this the design we are making? http://makezine.com/projects/make-36-boards/nuclear-fusor/
Reply all
Reply to author
Forward
0 new messages