Stefan's Tesla-Pages

My 2"-System


These are the pictures of the first films I shoot from my 1997'er "small_TC". This is my second TC, but the first one that really works. I've built already a very small one (60W) as a high school project in 1986, this was one of the typical misconcepted designs with a long and skinny secondary, incredible thin wire, no chance to hit resonance frequency with the used parts etc... For a better storage of the current model, the outer dimensions are chosen that I can store the components (primary, filterboard, secondary, glass bottle caps, control cabinet etc.) in standard cardboard boxes e.g. made for xerox paper (that was a very optimistic attempt as my systems grow now more and more :^) ).

Table of contents:

Specs of the 2"-system in April 1997
Photos taken in April 1997
Update Mai 1997
Update June 1997 (1)
Update June 1997 (2)
Photos taken in June 1997
Update October 1997
Update September 1998
Update February 1999-March 2000

Update July 2005

The specs of my 2"-system at this time (April '97) were:

Filter board
Size 10"x10" (25cm x 25cm). There are 6 chokes: two iron powder cores with 250 turns which gives two times 6mH and four ferrit cores with 77 turns which gives four times 34mH so that each feed line has an inductance of 74mH. The dimensions of the choke cores are D=1.4" (36mm), d=0.9" (22mm) and thickness h=0.6" (15mm) (cross section is then 0.6"x0.28" (15mmx7mm)). I also have plastic bypass caps (now blown ;-) & replaced by two nearly indestructible 0.56nF glass bottle caps), a safety gap, 2 resistors 7500Ohms@3W. The resistors blew when I switched to neons with more current (as Jim Fosse said once: "just realize the smoke is the working fluid of all electrical components; if you let the smoke out, it stops working"). They were replaced by two 260Ohm/11W resistors which is enough for short run times at 2600W up to one minute.  
I don't recommend the use of chokes any longer, please have a look down the page where I describe the current status of this project to see how an RCR-network will work instead of the chokes!

Two 50Hz/5kV/18mA xfmrs (from a second hand electronic shop, I don't know for what application they were designed for, the plate says the "totally playtime" should not exceed 5 minutes due to overheating though they are shunted). These OBITs get HOT after 3-5 minutes of runtime! Look at my neon page for more infos on xfmrs.

Series gap
7 gaps made of 8 heavy steel rings (modified set collars) on a wooden rod.

3-4nF glass bottle caps (foil covered, salt water inside). Click here to go to my capacitor page.

Inverted conical section with 11 turns under 21 degrees, AWG12 (4mm dia) halogen litz wire (braided => big RF-resistance!), total inductance of 22µH, mostly tapped at 7-8 turns (around 10µH), spacing between wire surfaces 0.15" (3.7mm), spacing between primary & secondary 0.71" (18mm).

2.16" PP form, winding length 10.3" (1:4.8), 640 turns, magnet wire dia 0.4mm (AWG26/27), inductance around 4.2mH, parasitic capacitance around 4.3pF, natural frequency therefore about 1.2MHz, loaded frequency mostly around 850kHz (varying with the bigger top loads at higher power).

Toroid 1.8" x 7.6" (4.6cm x 19.4cm, 8.4pF, called T5) made of styrofoam, laminated with epoxy, covered with adhesive aluminium tape.

I used my faraday cage as a counterpoise and also connected the whole thing to a cast-iron drain pipe.

max. sparks
3.15"(8cm) with this configuration.

back to top of page

Photos from April 1997:

Sorry that the images are a little grainy, but I only had a very cheap camera which allows only distances over 1m, so the things of interest had to be zoomed during the scanning process. (The photos taken at the end of August'97 of my 4"-system are taken with a better camera).

STK1TC22 Closeup of the whole system with a different toroid (2.4"x6.6" = 6.2cmx17.2cm, 7.4pF, called T4).

STK1TC23 Closeup of the inside (where the glass bottle caps are located).

STK1TC25 Closeup (like #22, but a little more top view).

STK1TC26 Primary (AWG12 braided halogen litz wire) with open strike protection ring.

STK1TC28 Filter board with:
2*250 turns on iron powder core (6mH) + 4*77 turns on ferrit core (34mH)
+ bypass caps (now blown & replaced by two 0.56nF glass bottle caps)
+ safety gap (perhaps I should have set these gaps not so wide with the old bypass caps)
+ 2 resistors 7500Ohms@3W (now blown as I switched to neons with much more current)
+ main spark gap (8 heavy steel rings on a wooden rod)

I don't recommend the use of chokes any longer, please have a look down the page where I describe the current status of this project to see how an RCR-network will work instead of the chokes!

STK1TC29 5A control board (analog meters for U and I, main switch, key switch, pushbutton, fuse, various indicator lamps, sockets for ballast, HV-xfmr and two auto xfmrs with different ratios or one variac).

sorry, to bad
to publish


3.15"(8cm) streamer in air and "power" strike to wire (great event, but awful photos).

STK1TC35 Plasmaglobe: BIG (2000W) light bulb lying on top of toroid with damaged fluorescent lamp in front of it (both ends at ground potential). Photo taken at 180W input power.

back to top of page

Update Mai'97:
Right after making the first photos, I found additional identical 5kV-xfmrs at the same second hand electronic shop (surprisingly, because the two old ones are 10 years old and I never saw one of this type again!) and bought 6 more (about 6$ one). But though I used additional caps (up to 7 nF) and various bigger toroids, I was not able to generate sparks longer than 4.7" (12cm) (I tried it not hard enough...). I decided to try it with 9nF or 10nF again sometimes.

back to top of page

Update June'97 (part 1):
As I want to make things always better as they are today, I decided to make a flat version of the RQ-gap for better performance of my TC. I used 10 sections of 0.9" (22mm) dia copper, 4" (10cm) long, which gives me 9 gaps. The spacing was around 0.028" (0.7mm) as was suggested by Richard Quick in the old archive files at funet. By replacing my old gap against this one (June 15th, 1997), I achieved 8.7" (22cm) instead of 4.7" (12cm) arcs. The only thing I wonder is that my 10kV-xfmr is only capable to use 4 (max 5) of the gaps (RQ: 1.4kV/gap => 7gaps ???). After widening the old gap and trying it again, I found that the sparks were more spindly and shorter than with the RQ-gap, independent of the number of gaps (until no spark occurs due to a too wide setting). I think this is due to a better quenching of the RQ-gap.

back to top of page

Update June'97 (part 2):
As the experience grows, also the toroids did and now I'm the owner of MANY handmade toroids with different sizes. As I said, the space in my basement was very limited this time. My old basement was extremely small and allowed only one foot of sparks with an 8"-toroid (total available space between the shelves was 31" :-( ). To overcome these restrictions and have some fun in my own basement, I made a conical section out of an old sheet of plastic laying around for years in my basement. The height is 11.8"(30cm), bottom dia 7.9"(20cm), top dia 4"(10cm). I taped it with aluminium tape as I did it with the toroids. The thing was placed between two toroids on top of my coil. This allowed me to arrange my top capacitance more vertical to get longer arcs without going to a toroid with a larger diameter. Current record set on June 18th. 1997 with 540W input power is 14.4"(36.5cm) arcs (point to point). On June 20th. 1997, I performed a test with one of my 63mA@8kV-neons and achieved 17.3" (44cm) with 635W input power. The primary capacitance was about 10nF in this case. I think the spark length can be increased furthermore by the use of a bigger primary capacitor of up to 18nF (and of course a better one than the salt water caps of now).

Some of the photos taken in June'97:


Beautiful display of an 20cm plasmaglobe (about 300VA input power).


Arc from the discharge electrode to an 12cm plasmaglobe with grounded filament (about 540W input power).


Arc between an 6cm and an 12cm plasmaglobe, filament connected to the toroids resp. grounded (about 540W input power).


Streamer from the discharge electrode.

back to top of page

Update October '97:
The braided halogen litz wire of the primary is now replaced by a bunch of 7 twisted solid copper wires, each 1.8mm (AWG13). The main system spark gap is replaced by a small modified RQ-gap, consisting of 7 pieces of 22mm dia copper tube (2" long) lying next to each other in a distance of 0.5mm. The HV-xfmr is now located inside of the housing at the bottom. Above it (yes, that will be a bad situation if one bottle will rupture...), there are the salt water caps (9 small mineral water bottles). The capacitance of them is much to small (only 5.7nF instead of 18nF), so they will be replaced in the future by plastic caps (click here to go to my capacitor page). With the new primary I think I have around 14µH and a frequency of around 550kHz with the conical section electrode on top of the toroid. The current in the primary coil seems to be around 200A assuming a breakdown voltage of 10kV in my gap.

back to top of page

Update September '98:
After a long search, I found some ceramic disk type capacitors 1nF@10kV. Assuming this as the DC rating, I put 6 of them in series and replaced the old bulky saltwater bypasscaps of my 2"-system (currently driven by 9.5kV@540W) with them. For safety reasons, I decided to add bleeding resistors over each cap. I took four 1MOhm@600V-resistors per cap for a total of 24MOhm@9.5kV. The discharge time constant is 4ms, bleeding current 0.4mA and reactive current in the ceramic bypass caps about 0.5mA at line frequency. So the losses in the filter board are about 2% of the input power. I don't recommend the use of chokes any longer. Instead, I'll use an RCR-network and perhaps some MOVs to filter the R coming backwards form the coil.

back to top of page

Update February '99 - October '00, reworked in March 2002, Update October2004:
The ERO caps (MKP = selfhealing aluminized PP, 62nF, 1kVDC, 330VAC, 1500V/us in series/parallel combination) that I use in Vitamini still perform very well. I aquired some more of these caps and plan to use them here in the 2"-system to reduce its size (height) and improve the performance. I expect at least 50cm sparks (2 times the winding length) with 540W input power and I'm even dreaming of getting up to 75cm (3 times the winding length). John Freau's formula
   spark length = 5.8 * sqrt (power input) / 4th root ( BPS)
predicts 108cm, but this is for more common systems, I'm sure I'll get not as much because my secondary diameter is far to small and I use not enough inductance in the primary circuit.

Due to the good lifetime of these (ERO KP1836, 1000Vdc, 62nF) and other caps, I decided to go for only 15 caps per string. I will use about 120% (160%) of resocap size, so I will make 2 layers of up to 4x 15 caps each. Resocap size is approx. 18nF, so the optimum 120%-LTR value will be around 22nF.  
Terry measured the ERO1845 and the result was an improvement from 40C/W (based on lead spacing) down to 27.5C/W for them. Hopefully, this improvement can be assumed for the ERO1836 with their 37.5mm lead spacing, too. So the 33C/W might be reduced down to 22C/W. Terry' MMC-Calc says 14 caps per string (540VA at 9.5kV) with a static static gap. So I will try the solution with 15 caps per string.

  • In Vitamini I have 20 caps in one string for 9450VAC. So I overstress the caps there by 145%AC, 67%DC and 197%dU/dt.
  • Toni, a fellow coiler, stressed a string out of 20 caps with 15kVAC (212%AC and guessed 99%DC and 62%dU/dt).
  • Reinhard (RWB) is also using these caps. His configuration is 67nF made out of 13 strings with 12 caps each used together with a 7.5kVAC supply on his 8"-coil. He stressed them for 3 hours until I lost track on this (189%AC, 88%DC and 47%dU/dt) and reported no cap failure using these ERO-caps.
  • With the planned 15 caps in series for my 2"-system, I'll overstress them 193%AC, 90%DC and 150%dU/dt.  

Compared to the old sw-caps, I will be able to reduce the total hight of the 2"-system by about 16cm when using the new plastic caps (and additional 4.5cm by rearranging the filter board and static gap).

Update July 2005:

Wiring scheme:

A gap (~2mm wide) between the individual rows of caps has the advantage that you can inspect the sidewalls of the caps in case some of them blew. Height of the caps is approx. 2.7cm, I arranged them in two layers. I use an 1.8MegOhm bleeding resistor across each cap. For increased cap lifetime, I added about 14Ohms in series to the safety gap (8x 1.8 Ohm) for reducing the voltage risetime stress (=current) on the caps in case of saftey gap firing down to a value near standard TC operation.

First test with the new MMC was done on 08.07.2005:
I did not use a couterpoise but connected the bottom of the secondary only to the earth ground in my cellar (concrete floor not connected!), so this was really a bad RF-gnd. The layers were not arranged on top of each other in the housing but layed out just next to each other (=> longer wires). I tested with 4x 15 caps (16nF) and the gap fired pretty fast at approx. 100V. Using 2 additional packets of 2x 24 caps from the old setup in parallel gave me the 120% LTR value of 22nF. I tapped 4.25 primary windings (approx. 4uH) and used an additional small toroid T3 approx. 1.5cm above the top end of the windings to suppress corona. I topped the whole thing with my toroid T12 (6cm above T3). I calculated a frequency of 550kHz for this setup. To suppress racing arcs (perhaps a sign of bad tuning - my old tap-clamp was not easy to move around), I raised the secondary 2.5cm up for loosing the coupling. Result was 30cm arc length to a grounded wire. Still got some corona and racing arcs. I hope to improve this by proper tuning after making a new clamp and by using a counterpoise as an improved RF-gnd. Surprisingly, the arcs were longer when I shorted out one or even two sections of my main spark gap. Measurement (only filterboard, no caps) showed that the main gap will need 190V for 4 gaps, 220V for 5 gaps and 250V for 6 gaps. Nominal voltage here in germany (without using a variac) is 230V. I've set the xfmr safety gap to fire at 250V input voltage which is the maximum I get when cranking the variac up as much as possible.

Energy per bang will be approx. 1.92J (compare this with the 0.52J of the latest sw-cap arrangement with 9x 0.63nF or the 1.92J with which I achieved 1.5m arcs using my 4"-TC with sw-caps in 1997!), peak current in the primary around 670A (I have to recalculate this for the new toroid I'll use). 

New secondary:
Since I used only 3 turns of the 9.4-turn primary, I decided to wind a new secondary with approx. 4 times the inductance than the old one. That will enable me to tap the primary at turn 8, leaving a tuning range of 38% of inductnce using the full primary.
Since I had only a 0.22mm wire at hand (wanted to use 0.19mm originally), I searched the DIY-store for a suitable tubing. I found it in the rain piping section, where 53mm tubing is common instead of the usual 50mm tubing in the plumbing section. The end caps which can be found in the plumbing section fit perfectly into the 53mm tube of the rain pipe section, so another problem was solved automatically! Furthermore, I found an adapter piece for 50mm tubing to a metal pipe with approx. 75mm diameter, which fits perfectly into the 53mm tubing an allows a much easier (more stable) placement of the toroids on top of the secondary. To increase the diameter, I decided to slit an equal length of pipe and slide it onto the other, filling the gap with epoxy glue (sanding it smooth afterwards of course) 
The design parameters for the new secondary  are: 53mm-tubing with "sleeve" for an outer diameter of approx. 58mm, winding length 11" (28mm), approx. 1240 turns (at 97% fill factor) of 0.22mm diameter enameled copper wire.
I used two endcaps and removed the outer ring so that they can slide into the tube. This way I can seal the tube and have still some bare tube at the end where the mounting for the toroid can be placed inside as well as another endcap screwed to the baseplate where the primary resides. That way, I can mount all the pieces together very quick when setting up at a new location (remember, this will be my transportable and reday-to-fire TC) and in a sturdy way (don't want do damage another toroid...)  

Results so far (20.07.05):
I got heavy racing arcs with the new secondary and had to reduce the coupling compared to the old secondary. Despite this, I get now 40cm with the new secondary (35cm with the old secondary in this setup) at 22nF and approx. 540VA.

back to top of page