pgtruspace's blog

about things that interest me.

Working on the shell condenser

Getting the form Right

Raw FRP shell

Raw FRP shell condenser shell form

shell form

Condenser shell form

I April 2012

The raw FRP ( Fiberglass Reinforced Polyester) shell form needs to be repaired and trued before the actual condenser can be created on it. Due to working alone there are starved areas in the FRP where there is not enough polyester resin to float the fiberglass. These white streaks are weak areas and need to be ground out and refilled with FRP before the shell can be built up and trued. There are many high areas that must be ground down to a minimum thickness and all other areas built up to obtain a smooth and true curve with little or no ripple under the capacitance foils.

After the minimum thickness is true and strong and stable, the low areas that remain can be filled with resin fillers and further trued. It took most of 6 weeks to get to this point. Now the edge must be trued to the center point so the curvatures are even and smooth.

Electron fields concentrate at points of sharpest curvature and we want the fields to maximize at the disk edge with the least loss. The more true and smooth this surface is the better the quality of the capacitance layers to be created on it.  Any high areas can result in thin dielectric and possible failure. Any low areas will result in loss in capacitance.

marking the shell form

marking the shell form for truing

To obtain a true edge a wire is connected to the top center point and used to draw concentric circles that are tested with the template that is a cut and marked piece of heavy corrugated card board. The use of this shows the areas that need to be ground down or built up to achieve a true shape. After nearly 3 gallons of filler the surface is very true. Now layers of polyester resin are applied and sanded to create a smooth surface, each more smooth and true then the last.  Now I need to clean the work area as everything is covered with a thick layer of polyester dust.  😥

Now to add the coils

The great coil and plasma jet in place

The 130 turn spiral wound high voltage coil has been cut out of the old device and mounted under the disk frame with the plasma jet pointed down.

The black coil around the spiral coil is the 24 turn  primary coil,  and the white coil above it is the bundle of driver coils for the twitch coils.

Inside of the central frame cage is the shell field driver coil that is connected to the groundplain in the rim and to the surface field cap on top of the coil and the outer skin of the shell.


Center coils inside disk

The shell field coil is induction driven 165 turns of 12ga tw covered wire and should create a 100,000 volt field on the shell capacitance due to the 600+ volt per turn field strength.

The shell field coil and twitch coils are loosely coupled to the main field and are jointly connected to the groundplain to make their operation independent of any back EMF from the spiral coil and primary driver that are connected together as an autocoil.

The primary and spiral coils should drive the plasma jet pincushion to about 100,000 volts.  At some time the primary coil can also be connected to the groundplain if needed for testing.

Twitch coils in place in disk edge

The primary coil is a bundle of 24 turns of stranded 12ga THWH inside polyethylene tubing and is powered by the 15,000 volt spark gap set. This should set up a 600+volt per turn field inside of all the coils

The twitch coils are mounted and connected to the groundplain.  2 turns per twitch coil are bundled into the white twitch coil driver necklace above the primary coil to independently provide 1200 volts to each twitch coil LC set. The twitch coils are 90 degrees to the other coils so are nearly independent of the main field.

FRP shell with cap in place

Every other segment is connected to the groundplain and bottom of the shell field coil. A coil/shell cap with carbon fiber whip has been created and covered with aluminum foil to act as the connection from the top of the field coil to the shells’ upper aluminum covering. The upper third needs to be trued by grinding and filling to the finished level as I prepare  to apply the inter condenser copper foil.


Construction of the Actual Shell Condenser

Inter Layer of copper foil


The FRP shell is ready for the addition of the copper foil segments that act as the inter layer of the shell condenser. Each segment connects to one side of the twitch coil behind it

Next, add the shell condenser  copper foil inter layer on the shell form. After several attempts I have found a way to apply a satisfactory copper foil layer to the form.

WOW!  This is starting to look impressive! If nothing else this is a work of Art or at least of PG.   😎    When the outer shell is finished this will be stunning! even if it doesn’t move.

The blue level is tape to hold the foil segments against the wires from the interior circuits. This is removed as I create the outer edge.

Outer edge formed

The outer edge form is done .




Fiberglass gores in place on shell

The outer edge shape is done and I am ready to build up the FRP shell over the inter condenser foil level. The Resin needed is on hand and the fiberglass mat is on its’ way. Hurrah! the end is in sight.

The Fiberglass mat has arrived and is being applied. Gores of mat are cut and taped in place with a slight overlap, then catalyst treated resin is poured on and rolled into the mat until the glass “floats” and air bubbles are eliminated. Fast work is necessary to complete the fill all the way around before the resin sets. To successfully make this, by myself, it is done in 3 layers. Each layer of mat requires the application of 3 and a half gallons of treated resin.

finished shell with aluminum skin

The shell shape is complete and covered with aluminum foil gilding.

A few coats of polyester and sand it smooth.

I just, have to, smoke test this!! Chase a few electrons around and around! light up some old florescent tubes.

It has been nearly 20 years since I energized the great coil

27 February 2013

Golden Shell Condenser

Polyester coated Shell condenser outside the hanger area.

The polyester coated shell is well cured and has been moved outside the hanger area. Now I have a lot more room to work and I have the greenhouse back  😎 for early spring grow out.

NOTE: Be sure to click on the pictures for a better view and more information in comments.  pg

Next see Condenser Disk and  Hurrah!


50 responses to “Working on the shell condenser

  1. Zeke April 20, 2012 at 11:35 am

    It does look like a true form, just eyeballing it (: In fact, it has such a pleasing form it could double as a sculpture!

    I was reading about some of Swedenborg’s inventions and his flying machine reminded me of yours, only he placed some navigational paddles on the top and his was not circular. He was a great inventor and visionary, so I thought you would enjoy the similarity.

  2. p.g.sharrow April 23, 2012 at 9:17 pm

    Swedenborg and De Vinci and many others designed many flying machines. Even Tesla designed and described an Electric Flying Machine. But I don’t know if any human actually created a working one that could be operated by a person inside it. Even the Wright Bros. did not invent the airplane, they just created the first one that could be really flown. Actually a farmer in Orland, California flew the first airplane in his father’s cow pasture. It even looked like an airplane and had a car engine for power. But it had no ailerons so you could not “fly” the wing. The Wright Flyer was a glorified box kite glider that they copied and created their own engine for, but they added “wing warping” so that they could “fly” the wing and actually control the craft in 3 axis. Curtis bypassed the Wright patent for wing warping by inventing ailerons and built aircraft that actually looked and flew like airplanes. So quite a number of people “invented” the airplane. pg

  3. adolfogiurfa April 27, 2012 at 8:50 am

    Dear P.G: Perhaps this will help you:

    Gravity is just a word for a local phenomenon.

  4. p.g.sharrow April 27, 2012 at 1:31 pm

    @Adolfo, Gravity is not a thing that exists. Matter sucks 😎 Kind of like getting old. 😦
    I just finished wiring up the “twitch coils” inside the disk. I hope to order the copper foil this week to start the disk capacitance layers. I will also start building the disk field coil. I am very anxious to do a “smoke test” on the electric circuits. 😮 pg

  5. adolfogiurfa April 27, 2012 at 2:42 pm

    @P.G.: That´s what Langmuir observed: Matter (particles) suck an ions´atmosphere around them. That´s charge, which if of the same polarity, it is rejected.

  6. p.g.sharrow April 27, 2012 at 9:53 pm

    @Adolfo; to my way of thinking, gravity is the result of the charge gradient in the Aether caused by matter and space. as happens in a linear accelerator that uses static charge differences to accelerate particles. pg

    Also I have added pictures into the above post. more to follow. pg

  7. adolfogiurfa April 30, 2012 at 5:28 am

    @P.G.:That´s right, you see it in three dimensions: The two polar forces of electricity, forming a squared (pythagorean) triangle, both being the “legs” of such a triangle, with two perpendiculars to that plane, that of magnetism and that of gravity (four forces total). Changing the ratio between positive and negative charges, changes the size of those two forces, up and down: Magnetism and Gravity.

  8. adolfogiurfa April 30, 2012 at 6:02 am

    I described it wrongly. Both polar forces of electricity (positive and negative) may vary its angle between them (as described in my theory) and, at right angles with them we have magnetism and gravity. The closest the angle between + and -, the closest to neutrality, the bigger gravity.

  9. p.g.sharrow May 1, 2012 at 7:28 am

    @Adolfo, I think your first explanation is closer to the money then the second. Trying to explain vectored forces in motion in 3 dimensions is not easy. You must describe a sphere of space in motion in 3 dimensions from the point of view of the center with forces acting on the outside of the sphere as causes or effects or both. A change in one vector changes all other points.
    You can not truly “grasp” an Orange until you can “see” all of it, inside and out at once. pg

  10. p.g.sharrow May 2, 2012 at 7:25 pm

    completed the top cap for the top of the field coil that is also the top of the shell condenser.
    Yesterday ordered the copper foil for the inter layer of the shell condenser, $274 for 110 square feet of 1oz. copper foil 😥 Sure glad the outer surface will be aluminum foil. pg

  11. adolfogiurfa May 3, 2012 at 8:23 am

    @P.G. You could get zillions of funding just by saying your apparatus will fight against “global warming” by turning the O2 of CO2 into O3, leaving behind the “C” as elemental carbon (soot) which will simply fall down to earth. 🙂

  12. p.g.sharrow May 3, 2012 at 9:07 am

    @Adolfo; I refuse to be part of that scam. However, getting rid of “Roman Candle” space launch would prevent a huge amount of pollution. My fume scrubber does scrub CO2 from the air during it’s operation. When you use “hard” water it makes limestone. 😎 pg

    NOTE: much more has been added to the above post. pg

  13. adolfogiurfa May 16, 2012 at 9:16 am

    @P.G. Can´t help telling you that it looks frail. It would be better if those beams were of titanium.

  14. p.g.sharrow May 16, 2012 at 8:58 pm

    @Adolfo; Man you are right about that! but at present there is no extraneous metal in the device. Wire and foil only! Every thing else is wood, plastic and glass. Everything is glued, no fasteners. I may need to add stiffeners to the lower frames. Although I stand on and work on the shell and I weigh a hundred kilos and it will be stronger when the FRP skin is finished.
    Actually if it works the strength is not important. If mass/inertia is reduced everything is plenty strong for the reduced mass and weight.
    As I am nearly finished adding the copper foil I must photograph and post it. This looks almost too good to bury under FRP and Aluminum skin. 😎 pg

  15. p.g.sharrow May 24, 2012 at 4:46 pm

    @Adolfo: I had the same concern about vortex runaway today as I was working on the shell. When the skin is complete and everything is working as planned, a very strong EMF field should circulate around the outer edge. This should build until the internal resistance eats up all the input power. Maybe I need a circuit kill added to the secondary coils? 😎 pg

  16. adolfogiurfa May 25, 2012 at 8:18 am

    Just a safety breaker.

  17. p.g.sharrow May 29, 2012 at 4:39 am

    @Adolfo; Not sure how to get a circuit breaker that works at many thousands of volts and hundreds of amps. Maybe a spark gap to short the coil. pg

  18. Brian H July 20, 2012 at 4:12 am

    Did it fatally zap you when you turned it on?

  19. p.g.sharrow July 20, 2012 at 2:33 pm

    Hi Brian; No, I have not been ZAPPED! yet. At the energies involved I’d rather not. It mite ruin my day. 😎

  20. Brian H July 20, 2012 at 10:51 pm

    Yeah, that’s why I used the adjective “fatally”.

  21. Brian H July 20, 2012 at 10:52 pm

    Correction: adverb. 8-(

  22. p.g.sharrow July 21, 2012 at 10:09 am

    @Brian; the diagramming of sentences always seemed pointless to me, I had a hard enough time with spelling and writing the alphabet correctly. That was back in the days when a word-processor was a manual typewriter. When computers got spellcheckers, I thought it was the greatest thing ever invented. Now I can write as well as speak! (and correct my mistakes.) pg

  23. Brian H July 24, 2012 at 1:32 am

    Think of it in technical terms. Minimizing ambiguity is the ruling priority.

    Disparaging grammar is like saying colour-coding of wires is boring and optional.

  24. Brian H July 24, 2012 at 1:35 am

    “Awl I right is prefect; my spell-checker tolled me sew.”

  25. p.g.sharrow July 24, 2012 at 6:10 am

    Wires have color codes?? 😯

    I just use whatever wire I can find as long as the size is adequate. 😎

    Actually wire color codes are helpful to find the terminus of wires that are inside the sheath, other then that, better test. Codes are not dependable. pg

  26. P.G. Sharrow August 6, 2012 at 6:08 pm

    I will add an other picture to the lineup soon. Tomorrow I hope to add the last of the aluminum foil to the outside of the disk. pg

  27. BobN August 8, 2012 at 8:45 pm

    @P.G. Sharrow – I haven’t gotten through everything on your site, so pardon me if I’m asking something already explained. What is your power source, I’m guessing your not running an extension chord from the house. With that much capacitance that will look like one big short for a long time. Are you pulsing it to not crowbar. Just curious as your needs don’t exactly seem off the shelf equipment.

  28. P.G. Sharrow August 8, 2012 at 10:25 pm

    @BobN; to start, a 15,000 volt. 600 watt neon transformer powered driver will be used.
    But yes, sort of on an extension cord 🙂 Off the shelf! what is that. At some point I will have to try out a 48 vdc battery powered amp that I built a dozen or so years ago that can drive up to 40,000volts at up to 50Khz. As of yet I don’t know if it will work or blow up. It needs an adjustable frequency generator to provide signal to the power transistors. I have created lots of various parts that can be used if needed. Way too many unknowns at this point to have only one solution to the needs of a working device. pg

  29. BobN August 9, 2012 at 6:08 pm

    This will be most interesting to throw the switch on. The expression “off the shelf” just refers to standard things you pick up at any retail store, Fry’s or Radio shack for example.
    This is all very interesting, I will be following closely what you do.

  30. Simon Derricutt August 12, 2012 at 12:32 pm

    pg – some sort of “circuit breaker” for the coils would be a shorting turn where you can make the contact and thus stop the flux. Since it would only have 600V on it, and would only be an emergency stop, any sort of contact that wouldn’t burn out would be OK, and again it could be pulled with a string so you can remain at a safe distance. I’m guessing that if you hit the resonance you may end up with a lot more power to cut than the 600W you’re putting in for now – nice to have the kill switch if needed. By setting the gap in the Kill Turn you may also set an upper limit on the voltage on the coil before it arcs out – another safety feature.

  31. P.G. Sharrow August 12, 2012 at 3:52 pm

    Good idea Simon, and I know just the place. There is a groundplain just inside of the rim, 1/4 copper tube that serves the twitch coils and is normally an open turn. Shorting it would kill everything. Lets see, I need an adjustable spark gap and deadman switch.
    I already have the gap on hand, just need to add a pinned kill switch.
    At present I’m surface finishing the disk. And controlling my urge to attach power and smoke test! 😎 pg

  32. adolfogiurfa August 13, 2012 at 4:40 am

    Eager to know the result of your smoke test..

  33. P.G. Sharrow August 13, 2012 at 7:39 am

    @Adolfo; As am I.
    One apparent anomalous effect to report. I notice the dust from sanding clings to the topcap, whip and things over the disk at a rate several times greater then the top of the disk and the floor. This must be charge caused, as the material clings all the way around and not just on top of the objects. The design of the disk should cause the charge bodies to accumulate on the outer surface edge and cause the inter surface to be depleted. The sander rips the polyester and charge bodies off the rim and blows them into the air over the disk where they cling to lower charge areas. The disk is now insulated as the surface is being coated with polyester, only the whip and topcap are naked and electrically connected to the interior. pg

  34. adolfogiurfa August 14, 2012 at 6:36 am

    @P.G.: That “static electricity” buildup is what we know as “gravity”, which it is not a “constant” but a variable. Sanding, rubbing on the glassfiber bonding resin causes this phenomenon.

  35. Simon Derricutt August 14, 2012 at 11:53 am

    pg – I was wondering if my email on dielectric losses in the shell reached you. Not the best of news, but it is subject to a lot of uncertainty and it does give you a way to get things better once you’ve got some results. If it didn’t get there I’ll send it again and hope (though “hope is not a strategy”).

  36. P.G. Sharrow August 14, 2012 at 12:54 pm

    @Simon; got it fine and am evaluating the information. I need to get you better information and drawings for a more exact understanding. I believe the dance in the dielectric is a feature as much as a cost in energy. Much more later. pg

  37. Simon Derricutt August 14, 2012 at 2:19 pm

    pg – could be I have the wrong end of the stick, but I would expect that dielectric losses (conversion to heat) are probably just losses. Reducing the losses should give you higher voltages on the foils and in the dielectric, so should confuse the world a bit more convincingly. On getting closer measurements to calculate from, this is not the major source of error. The biggest error will be in my estimate of tan delta (tand) of your resin, and in the thickness between the foils. On thickness, you told me 3/16″, but on “fibreglass on shell” (14th June) you say 125 mil or 1/8″ – found that afterwards. Since energy loss depends on square of field strength, you get a bit more loss with a thinner GRP shell. The reason for the calculations, even though ballpark, is to give you an idea of what to expect (from standard physics, anyway) as you fire it up. I’m glad the Kill-coil idea is useful. I’m also looking forward to reading your report on the smoke test.

  38. BobN August 14, 2012 at 2:38 pm

    PG – you might think about a good discharge mechanism to put things back in save mode when your done and want to touch or work on the device. Have an easy safe way to discharge everything properly. You probably thought of a way, but to think about it after you have turned it on may be problematic.

  39. Simon Derricutt August 14, 2012 at 3:07 pm

    pg – Dust sticking to the shell is probably because you have over-coated the Aluminium with polyester resin. You thus have a good insulator on the surface that will hold a static charge. The sanded dust is triboelectrically charged – always happens when sanding this stuff. I suspect you don’t want to use an antistatic cloth on it, but a wet rag cleans it and discharges it pretty well. One thing that will reduce the problem is an ionised air blower, so the surfaces discharge to air. Using the neon transformer feeding a bed of needles is something you can jury-rig quickly. Just don’t stand in the wind from this.

  40. P.G. Sharrow August 14, 2012 at 5:10 pm

    @Simon; we have several miscommunications. The dielectric thickness is close to 3/16in, 180 mils or 5mm and the active area is about 9500 square inches or about 7 square meters and contains 22lbs of glass, 10kilos? The figures for FR-4 “high resin content” should be close enough. Dust on the shell from sanding is easily cleaned with a wet rag. I was mentioning observed dust collection on the whip and things above the shell that have gathered thick “skins” of dust around them rather than dust “on” them.
    @BobN; Discharge of any stored charge is a very good idea. Once this thing is activated the overall device can create and hold a large charge. pg

  41. Simon Derricutt August 15, 2012 at 3:30 am

    pg – calculations were done using thickness of 4.7mm estimated. Given the dimensions, I worked out around 3.5m² for the active area – this would be the area of the phase-connected Copper foil, not of the total surface area of the foil and not that of the Aluminium foil. About 5400 square inches (so maybe I over-estimated this). Since the phase-connected Copper is coupled to the Aluminium foil and then the Aluminium to an equal area of Copper that is grounded, it’s possible to look at it as two series capacitors with half the voltage on each, which reduces the calculated losses to 1/4 of the worst-case, but it’s still quite a large number.
    It’s going to be almost impossible to work out the waveforms on the foils, but the calculations give you a ball-park figure for peak voltages. Similarly, although you reckon on 20kHz or so, there will be harmonics of this going much higher. I can’t think of a way of reliably measuring these sorts of voltages and waveforms – the meters will certainly tell lies.
    The smoke-test is the only thing that will give you real data. It could be an idea to have an IR thermometer, so you can get some idea of temperatures reached in operation, especially once you move on to the higher-power tests. If there’s corona around, it may give the wrong answers, too, but better than nothing. Otherwise, tape a glass thermometer to the shell and use a telescope to read it.
    It must be getting close to smoke-time.

  42. P.G. Sharrow August 15, 2012 at 6:40 am

    @Simon; My estimation of active condenser area is based on the copper foil area. Our discussion has just now given me an idea of how to pictorially represent the shell condenser segments and twitch coil elements. Lets see if I can “draw” this in a file. I believe you are right that all of this should drag the operating frequency down from the theoretical LC numbers. the numbers say 24Khz. My guess is 18-20Khz. My design work has been ,”better then needed to resist the maximum voltage”. All these handmade parts of questionable characteristics is a big WAG at the outcome. (Wild Ass Guess)
    The coronal field on the disk rim and its’ manipulation is the object of all this design dance.
    Before the real tests can begin, the surface must be finished and one more element should be added to the rim of the disk. this element is not part of the internal circuits so would not preclude a “smoke test” before its’ addition. pg

  43. Simon Derricutt August 15, 2012 at 7:10 am

    pg – in situations like this, there’s no good data to base things on, so the WAG is the only way. Measure the results, make it better next time. Until you see the results of the WAG, you can’t go further. If the first disk doesn’t do all you want, I think you could sell it as a work of art anyway. In case I haven’t made it plain, I’m really impressed by the craftsmanship.

    I’ve also found that having to explain what’s supposed to happen to someone else helps me in getting the ideas straight, and sometimes leads to new ideas simply because the current ideas have been placed nicely in a row.

  44. Zeke September 25, 2012 at 10:19 am

    “The coronal field on the disk rim and its’ manipulation is the object of all this design dance.”

    This is an interesting set up. I have some questions about the effects of a strong field on one or two materials. Perhaps it is possible to place these within 6″ or so of the shell condenser. Do me an e before testing if you think that would be alright. Thanks, Zeke

  45. Pingback: OPEN THREAD « pgtruspace's blog

  46. p.g.sharrow April 15, 2014 at 12:12 pm

    The golden shell above is no more, It is now battleship grey, as it is being trued and smoothed with Bondo and then grey primer. Lots of sanding to get the surface as true as possible before the final coating. Nice to have a real shop area to work in! 😎 pg

    SEE; Condenser Disk

  47. Simon Derricutt April 15, 2014 at 1:32 pm

    This has been a longer-term build than I thought a year ago. Machine-sanding or elbow-grease? Getting the thickness right must be difficult, though. For guitars I’m dealing with relatively flat bits of wood (at least to start with) but here you’ve probably needed to build a large tool to do the measurements. Either that or use templates to get the curves right and hope for the thickness. It will be good to see it fly.

  48. p.g.sharrow April 16, 2014 at 7:47 pm

    @Simon; This has been a longer build then I anticipated as well. It has been 2 years since I posted this, as I started the condenser shell on the former frame.
    When I started there was a lot of belt sander and disk sander work but most of this last work was done with hand sanding. During the original formation I used templates and then tried to make each layer as even as possible.
    I began this project in 1994 at the insistence of a friend when he examined my paperwork thoughts on the subject of mass/inertia and gravity. I started this shell 8 years ago as my second attempt to create a working test device. pg

please say something interesting

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: