As mentioned elsewhere on this site this coil was created from the remains of a previous coil that was wrecked by an irate neighbor. This 'resurrection' was however a hasty affair using a lot of the previous parts, including the already inadequate main base (old TV stand).
So with a view to fitting an even larger and heavier toroid one day, and maybe even a bigger secondary, I decided to completely rebuild the base of the coil.
The new base was to be much heavier to provide stability and also needed to be more robust, so 0.7 inch [18mm] exterior plywood was used, with extra bracing glued and screwed across the middle lengthwise. I had decided to cover the exposed surfaces of the boards with 3mm black Acrylic sheet, while the primary is to be protected by a blue 5mm thick circular top sheet that sits about an inch above the primary coil itself. The MMC is built up of 4 trays of blue 5mm Acrylic that can be pulled in and out for easy access, should a capacitor need replacing at any time.
|
 |
A CAD representation of the finished coil - Phoenix 2.
The top and bottom boards are separated by six 11 inch lengths of 1.4 inch [35mm] diameter clear Acrylic rod. This has been drilled down the centre to take some threaded rod. These rods serve the purpose of clamping the top and bottom boards rigidly together.
In an effort to be able to experiment with the coupling more, I propose to make it easy to adjust by simply pulling out the yellow pins (Shown below) and placing them in a higher or lower hole. This will allow the whole primary assembly, whose four legs project through holes in the top level's board, to be easily raised or lowered. |
 |
Above: The Coupling Adjustment |
 |
Drilling the Acrylic legs for the threaded rod
Despite using a long drill I still needed to drill in from each end. The drilling however has caused the hole which runs down the centre to have a rather crazed appearance when viewed through the Acrylic due to it being magnified, rather like the mercury in a thermometer is.
I next decided to add some colour into the hole to hide the studding that will pass through it, so i filled the rods with paint, and then left them for a couple of days to allow the paint to drain away and dry out inside. |
 |
Sliding MMC Trays
The method employed for the 4 sliding trays of 216 capacitors (101nF 12 strings of 18 in series) was to mill slots into four of the Acrylic rods that support the upper level. The trays themselves are 0.197 inches [5mm] thick and the milled slots that retain the shelves are 0.25 inch both deep and wide. Anymore depth would have weakened the support rod as the central hole down the centre is 0.314 inches [8mm]. |
 |
Above: Milling the 0.25 inch slots for the MMC trays |
 |
New Secondary Mounting
The previous secondary mounting was very unstable as it was originally designed for a 4.7 inch coil with an 18 inch toroid. The much larger toroid I now use (34 x 8 inch) combined with an 8 inch secondary, means that trying to use it in the garden, if there is a light breeze, could be hazardous.
Because of this I made the mounting shown below. |
 |
The Aluminium and brass (see note **) mounting on the base of the secondary (above) has a central threaded boss (circled in blue) that screws down onto the copper threaded rod (shown below, also circled blue), this allows the outer surface of the secondary's mounting plate (circled red) to engage against a matching Aluminium surface on the lower part. As the secondary is rotated it then screws itself down and locks tightly, so no movement is possible.
** The use of brass for the central threaded boss is not ideal with RF, but I wanted a thread that was durable and not easily cross-threaded. The brass boss in turn then screws onto a solid copper rod. |
 |
 |
| Screwing down |
 |
| Fully home and tight |
 |
 |
The SRSG is now in place.
Copper bus bars (0.625 x 0.25 inch) will be used for the feeds rather than cable. The vertical copper bar you will go to the inside of the primary.
This rod is a far larger diameter bar than what is really needed, but it was lying around, so I put it to use. My aim is to have as few as possible leads in the coil, hence the arrangement shown below. This will hopefully mean that the primary frequency stays reasonably fixed rather than varying slightly every time I reassemble the coil. The variation in frequency has in the past been as much as half a turn's adjustment of the primary.
This means that the simple action of altering the primary tap lead and feed it a different route to tap elsewhere on the primary, can in effect make no difference to the resonant frequency. Of course it could also work the other way, as it was possible sometimes to improve the output just by rerouting the tap lead. An easy way of doing it, but far from predictable or repeatable for the next coiling session.
Resistance Table
On the matter of using copper bar, I found this table interesting as it compares the resistance of often used materials compared against copper, which it gives a 100% figure. So for example a Zinc bar would only be 27% as conductive as a copper one would be. (Source)
| Silver |
105% |
| Copper |
100% |
| Gold |
70% |
| Aluminium |
61% |
| Brass |
28% |
| Zinc |
27% |
| Nickel |
22% |
| Iron |
17% |
| Tin |
15% |
| Phosphor Bronze |
15% |
| Lead |
7% |
| Nickel Alum. Bronze |
7% |
| Steel |
3% to 15% |
|
 |
The eventual idea is to have copper bars replacing most of the wire (shown yellow above). |
 |
The basic layout of the SRSG board (Tufnol) takes shape. |
 |
Wiring Circuit
The individual trays of capacitors for the MMC will each terminate onto the two vertical copper bus bars. One of these bars for space reasons has had to be routed underneath the main SRSG motor. |
 |
The Finished (Almost Finished) SRSG 2Hp 3000rpm.
The two upright brass pillars to the bottom left, will hold the horned safety gap. The HT input connections are the two round brass terminals (middle & bottom right). The tall copper rod (rear right) will go through into the top level and will have one end of the primary tap connected to it. While the copper bar running under the motor will go the MMC.
The above photo is called the 'almost' finished because I later modified things - see below. |
 |
The tungsten 0.25 inch electrodes.
The existing electrodes consisted of two short ones inserted into their holders each side. I have now dispensed with this arrangement and used four, one piece electrodes that go through the rotor disc. These have all been ground to be 2.846 inches long, which when combined with the rotor's 2 thou run-out means I can get the gap down to 8 thou.
The use of one piece electrodes means they should align better with the fixed electrodes, and also be safer, with less chance of one coming out. The ends of the electrodes were rounded as is normal practice, to avoid any sharp edges which can cause overheating. Also the corner area, which in effect is really a 'point' electrode, can also cause the electrode to fire early as well. I have however left the very centre of the hemisphere flat.
It has to be remembered in a correctly setup coil, that by the time the electrode have aligned, the capacitor has fired anyway, and the tank oscillations have all but finished.
|
 |
The SRSG on its new base.
The Tufnol base of the SRSG sits on a rubber mat which in turn sits on 3mm black Acrylic. |
 |
The MMC Trays
The blue MMC trays (there will be four in total) will each have just have two short lead connections to the two rectangular copper bus bars you can just see at the tray's rear (middle area of the lower shelf). This will make removing a tray quiet easy. |
|
The whole affair is quiet compact but will dismantle quiet easily if needed. |
 |
Copper Bus Bars
The left hand rectangular copper bar is point B of the MMC (See inset diagram) and runs through a hole to emerge onto the top deck. To the right of this is the earth connection to the secondary in the Acrylic tube. This is entirely electrically isolated from the base construction. Wood can and will conduct, and although it's very unlikely, having the secondary earth in contact could mean it would attract a streamer to the base.
The earth connection is actually underneath the base, and is just a wing nut connection on an insulated mounting block.
The next shorter Acrylic tube feeds the electric to the SRSG motor from the phase controller. Lastly we then have another rectangular copper bar (representing point A). This does not project through into the top deck and is isolated by the Tufnol blocks at its top end.
Point C is a solid round copper bar that runs straight up to the top deck, and can just about be glimpsed behind and between the left hand copper bar (B) and the SRSG's rotor.
The copper bus bars have tapped holes in them to take the two connections to each of the four capacitor trays (only two trays are shown above). |
 |
The partly built four MMC trays
The sheets of blue acrylic had nine shallow channels cut into them with a router. This located the caps nicely, but had the problem of causing the acrylic sheets to bow because of the removed material. This was overcome by gluing reinforcing strips on their underside.
Next a total of 432 holes had to be drilled in all the trays for the cable straps, and although a template was used, this still proved rather tedious to do.
Each tray consists of 54 capacitors comprising of 3 strings of 18 in series. This gives a voltage rating of 36K and a total capacitance of 0.1uF or 100nF for the whole MMC.
Once completed I felt the tray should have been made larger however, as although this arrangement looks tidy, it means the capacitors have fairly short tails to them. Also as the connections involve twisting before soldering, it makes future replacement difficult.
Merely just arranging the capacitors leads side-by-side and soldering without mechanically twisting them, could create an extra resistance, as solder only has around 20% the conductivity of copper. While only a very small resistance point, power lost is I^2*R, and the MMC does pass fairly high current pulses through its 216 solder connections.
Capacitances of this size should really be using dedicated pulse caps, Maxwells for example, to overcome this problem and to reduce the amount of construction, and in the end also the cost. |
9th January 2011
 |
One of the four trays from the MMC with the wiring now in place.
Each tray of caps just slides into place and the two connections (circled yellow) then connect to their respective copper bus bar by bolts. The heavy wiring is in excess of what is needed in places, but much like my use of over-size copper bars in the SRSG below, it will serve no harm.
|
|
Update: 25th April 2011
Above and below:
Additional strips (arrowed in red) were later inserted to stop flashovers that occurred between adjacent strings. This of course makes capacitor removal even more difficult than it was previously! But if proper allowance had been made at the outset, then this method should be quiet robust against flashovers.
|
|
The soldered and twisted 'tails' of each capacitor (arrowed in yellow) have had a dab of hot glue applied and then a short piece of plastic tube slid on to insulate the ends.
A narrow strip of Acrylic (arrowed in red) is then inserted the length of the capacitor string to insulate from adjoining capacitors.
|
|
Finished SRSG - Two Steps Forward, One Step Back
I was never happy with the new electrode's mounting arrangement, as it was possible for the two terminal posts to twist on their mounting (a single bolt from underneath screws up into their base) allowing the stationary and rotary electrodes to crash together. (The original arrangement is shown here)
I prefer to run with a 10 thou gap to cut down losses, so the primary was designed, against modern doctrine, not to have excessive turns.
Modern multi-gap spark gaps need higher primary turns to keep the current down, but a RSG running close tolerances has lower losses.
Therefore to stop any possible twisting I introduced a Tufnol bracing piece between the electrode posts, and secure it to the posts by four small bolts.
This makes the whole affair very rigid, and hopefully I can now run with a small gap.
Originally I had run out of copper bar so had used two pieces of brass, but a recent find amongst my junk allowed me to replace these two brass sections as well. This now means the entire tank circuit is made from copper bar or rod. The size of the rod and bars is far larger than is needed, but it will not in any way be detrimental. |
 |
Top shelf and capacitors removed.
After many design changes the SRSG is finally finished. The two terminals are shown painted white in some of the above photos, but that idea was abandoned, as paint and copper do not adhere very well! The gap is currently at 9 thou and the motor has had a 15 minute run and nothing has clashed, so fingers crossed, and I will slowly decrease the gap when the coil is running and see if there is any noticeable difference (most likely not). |
 |
Cutting a 38.5 inch disc using a router
The Acrylic sheet acts as a guard against streamer strikes, and will also support the primary coil on its underside. To cut the large disc I first cut a 2 inch hole in the centre using a standard hole cutter. A central boss was then made at 2.125 inch diameter, but with a step in the lower 0.25 inch of height that was just under 2 inches.
This boss sits in the disc's hole lightly clamping it down to stop it lifting, but at the same time allowing easy rotation of the sheet around the central boss.
The router was then clamped to the vertical post on the left hand side, while the anchor point on the right fits into the central boss.
The sheet could then be turned by hand and fed into the router. This method allowed me to cut a very clean edge to the disc. |
|
The inner 9 inch circle was cut using a Dremel.
The extension arm was made to fit into the same boss that the router had used to cut the outer edge, thus ensuring two concentric circles.
This time though the sheet stayed still and the Dremel was turned. The hole measures 9 inches diameter as the current secondary is 8 inches diameter. |
 |
The primary coil itself mounted on what will be the underside of the protection cover for the primary. |
 |
Held against the light, you can see the primary mounted on the underside. |
