Insulation?

[adalast]

what would be the best way to insulate a flyback in a high voltage generator? i am using this instructable to construct it, and it is working pretty well. so well in fact that on the original one that i built i couldn't get the stupid thing to quit arcing between the ground pin and the other pins/transformer body when i hook it up to my lifter. even with a decent amount of hot glue on the pins, it seemed to arc through the hot glue.

my question that i'm trying to find a good answer to is... what is a good way to insulate the 0 volt pin on the secondary coil of my flyback to keep it from arcing to all hell and killing my CF power supply or arcing on the interior of the flyback and destroying it. that is what happened to my original flyback, i finally got it to stop arcing to the body of the transformer and it decided to arc on the interior and burned out.

any suggestions? i am thinking an oil bath, but i don't have anything but kitchen vegetable oil handy and i'm not sure how well that will work.

 

[Hero999]

The problem is, that because of the high frequency used, there will always be a risk of arcing from the primary to secondary because of the capacitance between the two coils.

 

[adalast]

is there any good way to limit the frequency without killing the voltage? or how could i mitigate the risks? i guess i could live with the risk of blowing the internals of the flybacks since i can get them for cheap... but the CF bulbs can get kinda pricey. also... if someone could suggest another decently cheap way to build one of these, i'd be listening.

 

[Hero999]

You can't limit the frequency because the transformer only works at a high frequency which is generated by the CF circuit.

The interwinding capacitance is part of the transformer and effectively short circuits part of the high voltage secondary to the primary at high frequencies which means the secondary will spark to the primary.

The only way I know of minimising the capacitance would be to wind your own primary coil which is fairly straightforward, just wind 20 turns or wire around the part of the core you can see. This will reduce the capacitance because

Why don't you recycle a broken CF? Most of the time it's the fluorescent tube that breaks leaving the electronic still operational. CFs are cheap where I live but I'd still consider it a waste to use a working one like this when I can get a broken one for free. For me the most expensive part would be the flyback.

 

[adalast]

good idea on the CF bulbs... i hadn't thought of calling one of the recycling centers. I am having trouble keeping the flybacks operational right now. i tried the oil bath and it completely eliminated arcing on the outside of the flyback, but i think it just promoted the interior arcing that kills the flybacks. this makes dead transformer #2 in as many weeks and i'm trying to figure out if it was arcing or the oil that killed it. it seems that whenever i get it insulated to the point that it won't arc across the pins, it kills itself. is all of this arcing being caused by the high frequency coming out of the CF? is the CF putting out too high of a frequency? is there an alternative i could use for the CF that might not give me the arcing problems?

maybe i should look to some other high voltage source. i've heard that ignition coils are a good place to start, but i'm not sure how much they will cost from a junk yard. i think i'm going to get an old TV from goodwill or something and see if i can just rig it to be my generator for the time being. I would do it with the monitors that i'm getting from a local shop for $5 but they keep turning off the flyback because i don't have a computer hooked up to them. not ideal when i'm trying to use it as a HV power supply.

 

[Hero999]

It sounds like the coil is arcing internally which is probably due to the secondary voltage being too high.

Don't forget that the coil is not designed to be operated open circuit, but with a load connected.

The secondary voltage to the flyback can be reduced by reducing the number of turns on the primary. I know that sounds counter-intuitive but the CFL PSU will produce a constant current output so less turns means less voltage on the primary.

You could try winding your own primary as I mentioned above, start with five turns and increase it until the output voltage is high enough.

Another option is to just connect a spark gap to the secondary of the coil to protect it against overvoltage.

 

[adalast]

The spark gap would reduce the effectiveness of the lifter tremendously though wouldn't it? Unless I could make it work by putting the gap in series with the lifter, which is something I haven't tried. Also, I have never wound my own transformers and would be interested to learn how to do it. Could you point me in the direction of a good resource for learning how to do things like that?

 

[Hero999]

Putting the spark gap in series wouldn't help because it wouldn't protect the coil from overvoltage when the load is removed would it?

You don't have to rewind the whole transformer, just add another winding on the opposite side to the other coil.
**broken link removed**

The site I got the picture from has more information:
Solid State Tesla Coil with 555 Timer

You could use the 555 timer based circuit or stick with you CFL circuit which is probably more convenient.

 

[adalast]

so i would have to take apart the flybacks? the ones that i am using are completely contained in a housing. they are out of old CRT monitors.

 

[Hero999]

What does the flyback look like?

If it's the same type as is used on the instructable then adding another winding will be easy.
**broken link removed**

If it's too difficult to wrap it round, you can easily remove the core by undoing the clip shown in the picture below.
**broken link removed**

 

[adalast]

yeah, that's what they look like in general. I think i'm going to have to get a new one, but overall, that is the layout. do i just do the wraps around the plastic bar there? then hook up the power from the CF circuit to the wire i wrap around it? i took off the clip, but it doesn't seem to come apart. Also, could i possibly use one of the ones that i've burned out the interior of for this?

 

[Hero999]

Perhaps the two halves of the core are glued together.

Still, I don't see why you're finding this so hard?

Just wind the coil round the part of the core you have access to, see the attached.

 

[adalast]

no offense meant, but i'm finding it difficult because this is my very first HV project in my life and i have little experience with some of the technology beyond some basic electronics training i received in my computer repair classes. i'm a novice who is looking for a little guidance and i don't appreciate being talked down to with comments like "I don't see why you're finding this so hard?" the image explained part of what i need to do, but i still need to know if i can use the coils that i have destroyed or if i need a new one, or what i hook up to the coil. am i just hooking the CF outputs to the coil and then what is attached to the secondary coil? am i using the output lines i already have attached? also, does the wire need to be insulated or uninsulated. i have quite a few questions that are basically as a result of my having little to no information to start with. my father has been helping me with some of this, but he even only has limited experience with hacking stuff to make it function in a way that was not intended.

 

[Hero999]

Sorry if my comment cause offence: if was a genuine question. Remember I can't see your flyback so I don't know what difficulties you face.

The idea is the winding just replaces the internal primary and should be connected to the output of the CFL supply.

Of course the wire must be insulated to prevent neighbouring turns from short circuiting each other. It doesn't have to be well insulated, enamelled wire will do but don't touch the coil when it's powered as the insulation won't be good enough to protect you from a shock. If you're worried cover your coil with a layer of insulation tape but even so, don't rely on it for safety.

Don't use as many turns as I've illustrated on my schematic, it's just as an example, I found the photograph on Google image search and superimposed, a coil I had drawn before in Inkscape on the picture. As I said before, start with about five turns and increase them if the output voltage isn't high enough.

Take the output from the pre-made secondary coil as you have done before.

The broken transformers aren't good for much, unless you want to wind your own transformer from scratch.

 

[adalast]

thank you for the well elaborated answer, that is what i needed. I will get another monitor and strip the flyback from it when i can make it down to the store. hopefully tomorrow or wednesday. I have some wires from the burned out transformers that will wrap fairly well around the primary coil thing and should work well, if not for them being so hevily insulated. i have some enamaled wire if that won't work. my real concern though is if this will prevent the flyback from dying (which it sounds like it will) and if it will reduce the need for high insulation on the output from the secondary coil's 0v pin.

on the last two i was getting so much arcing that on the open circuit of the lifter i couldn't stop it with hot glue. as i stated previously, the oil bath seemed to work, but i'm just trying to figure out what precautions i need to take going into all of this. i am leaning toward completely submerging the whole thing in vegetable oil to help with cooling it as well since the transistors on the CF tend to get kinda hot and i think that's what usually burns out on them. I would really love it if i could get one of these power supplies to last more than a few days and actually be able to power the Ion drive so that i could get to the experiments i am wanting to do with the technology. they all work on the air gap being larger than the breakdown, so all energy transfer will have to be in ionic form. this seems to be the Achilles heel of the power supply design. do you have any suggestions as to how i can keep them alive in this setup and any advice on the lifters, or should i direct that second question to the experiments forum?

 

[Hero999]

Vegetable oil will improve the insulation but I don't see any need for it. The transformer shouldn't get very hot and therefore shouldn't need cooling. The CFL ballast might get too hot and require cooling though. You might need to improve the heatsinking of the switching transistors and/or add a fan to improve the cooling. I was going to suggest a CPU fan but that would require an additional DC power supply so you'll need to use a small AC powered fan.

I can't help you with the lifter, I've never built one before and don't know exactly what one is. I would discourage you from starting a new thread because those who respond to it wouldn't have seen this one which provides lots of background information. If you do start a new thread, then you must provide a link to this one, otherwise it will irritate some people when they realise there's another thread.

 

[adalast]

ok, i think i can stick to this thread for the time being. basically i need to figure out how to get the power supply to survive dumping power into what amounts to an open circuit. here is a labeled image of the basic lifter design. mine is not far off of this one, only i'm not worried about a payload yet... i still have to get the power supply to be stable so i can get it off the table.
**broken link removed**

basically how it works (if you don't know) is one side of the HV is attached to the Corona wire and the other is to the collector foil. when the high voltage is applied, it ionizes the air around the corona wire and accelerates it towards the foil. this acceleration causes the ions to strike the neutral air around it and generate lift.

the air gap has to be wide enough between the corona wire and the foil to allow for some corona, but no arcing. the corona should be minimized because it represents a loss of energy that is not forming any momentum, but some is ok. as you can see... it is basically an open circuit and i am trying to figure out the best way to make the power supply be able to survive this, since i do know that it is an intensely taxing task for the high voltage supply as the electricity is going to try to find some other way to ground rather than ionizing the air.

i think that wrapping my own primary coil will help with isolating the CF from any feedback arcing that might occur. was all of that arcing i was getting on the flyback pins before caused by the fact that i was using the internal primary and the power connection was too close to the 0v pin? (sorry, doing a little thinking "aloud" and asking as i go through it). maybe this will actually solve the insulation problem i was having with it arcing all over the place and actually stabilize the whole thing. now i'm excited to get another flyback and get it working.

 

[Hero999]

I think I understand what you mean but your picture isn't very clear, is it possible for you to post a higher quality image?

I'm not sure you understood what I was talking about when I mentioned the inter-winding capacitance and high frequency.

A capacitor is formed when two conductive bodies are separated by an insulating material. The thinner the insulating material and the larger the surface area between the two bodies the higher the capacitance. Look at a transformer, it has a primary and secondary coil, separated by an insulating medium, therefore there will be a capacitance between the primary and secondary coils. If the coils are wound on top of each other (how the transformer comes ready made), the capacitance will be much higher than if the coil is wound on the other side of the core (the modification I'm suggesting).

Why does inter-winding capacitance cause arcing?

Capacitors pass AC current and block DC current.

The CF PSU generates a high frequency AC current which is passed to the primary of the flyback. This induces a higher voltage but lower current AC power into the secondary of the transformer. The inter-winding capacitance connects the primary coil to the secondary coil at higher frequencies (remember a capacitor passes AC). If the primary and secondary coils are wound on top of each other, the inner part of the secondary coil will have better capacitive coupling to the primary than the outer part of the secondary - the inter-winding capacitance will not be uniform. This means that that the high secondary voltage can easily arc to parts of the primary circuit.

As mentioned above, winding a separate primary on the other side of the core should reduce the capacitance and therefore the coupling.

 

[adalast]

ok, it took me a few times of reading, but i think i understand what you are saying. basically moving the primary coil to the outside of the secondary coil it is going to reduce the capacitive load on the circuit, thereby reducing the chance that an arc can form between the two. if i'm understanding how the whole transformer works, it shouldn't even lose much of it's voltage generation.

as for a description of how these ionocrafts work, i took this quote from the site that i'm using for much of my research on the topic. i think it explains the function rather well. this is the basic model i am trying to build on, and eventually i will be building more complex models that do not follow the format of the basic lifter, but still follow the general principal of using an ion cloud to impart momentum to the thruster... just hopefully with a greater efficiency and mobility. i just need a good high voltage generator that i can eventually put through a CW multiplier when i need much much higher voltages (that won't be for a long time though). (two sites, same research... the research site and the wiki)

In its basic form, the ionocraft is able to produce forces great enough to lift about a gramme of payload per watt, so its use as a vertical thruster is restricted to a tethered model. Ionocrafts capable of payloads in the order of a few grams usually need to be powered by power sources and high voltage converters weighing a few kilograms, so although its simplistic design makes it an excellent way to experiment with this technology, it is unlikely that a fully autonomous ionocraft for vertical take off will be made with the present battery technology. Further study in electrohydrodynamics, however, show that different classes and construction methods of EHD thrusters and hybrid technology (mixture with lighter than air techniques as those shown on Blaze Labs Research website, can achieve much higher payloads or thrust-to-power ratios than those achieved with the simple lifter design. Practical limits can be worked out using well defined theory and calculations such as those given on the 'Ionocraft mathematical analysis and design solutions' paper (see external links). Thus, a fully autonomous EHD thruster is theoretically possible.

When the ionocraft is powered up, the corona wire generates a very high electric field gradient. The user must be extremely careful not to touch the device, as it can give a nasty shock. At extremely high current, well over the amount usually used for a small model, contact could be fatal. When the corona wire is at approximately 30 kV, it causes the air molecules nearby to become ionised by stripping the electrons away from them. As this happens, the ions are strongly repelled away from the anode but are also strongly attracted towards the collector, causing the majority of the ions to begin accelerating in the direction of the collector. These ions travel at a constant average velocity termed the drift velocity. Such velocity depends on the mean free path between collisions, the external electric field, and on the mass of ions and neutral air molecules.

The fact that the current is carried by a corona discharge (and not a tightly-confined arc) means that the moving particles are diffusely spread out into an expanding ion cloud, and collide frequently with neutral air molecules. It is these collisions that create a net movement. The momentum of the ion cloud is partially imparted onto the neutral air molecules that it collides with, which, being neutral, do not eventually migrate back to the second electrode. Instead they continue to travel in the same direction, creating a neutral wind. As these neutral molecules are ejected from the ionocraft, there are, in agreement with Newton's Third Law of Motion, equal and opposite forces, so the ionocraft moves in the opposite direction with an equal force. There are hundreds of thousands of molecules per second ejected from the device, so the force exerted is comparable to a gentle breeze. Still, this is enough to make a light balsa model lift its own weight. The resulting thrust also depends on other external factors including air pressure and temperature, gas composition, voltage, humidity, and air gap distance. The heavier and denser the gas, the higher the resulting thrust.

 

[tcmtech]

After reading through this I see your basic problem. You are simply over driving your fly back coils.
Most computer monitor fly backs are only a few tens of watts at best.
If you are pounding on one with a 65 watt CFL drive circuit that comes off of a line voltage source you are just over voltaging the insulation to the point of break down.

I have built many fly back HV devices and there is always a limit to how much voltage you can push through one before the insulation breaks down internally.

If you want more output voltage you will need to either get a fly back from a much larger monitor or TV or use an external voltage multiplier circuit on the HV output.
The bigger the CRT the higher the working voltage and power it needed so look for big old TVs for a high powered fly back transformer.