DC tesla coils (theory) / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

DC tesla coils (theory)

1 2 next

Move Thread

LAN_403

northern_lightning

Mon Mar 03 2008, 04:19AM

Registered Member #1348 Joined: Sun Feb 24 2008, 01:48AM
Location: Michigan's Upper Peninsula
Posts: 22

i'm looking to build a DC driven spark gap coil. i would have stuck to an AC one, but after doing the math i realized there is no way to do it with my low voltage high voltage transformer while maintaining good coupling. i need to increase the voltage!

anyways, what formula do i use in order to calculate how long (in seconds) it will take a particular capacitor to completely charge given voltage and current in a full wave DC supply? basically i want to know how many times a second my single gap static gap will fire with a particular capacitor.

i have all of Richie's formulas for a resonant DC charging topology. but i don't want anything fancy with inductors and 2X input voltage. just a diode bridge and my cap bank.

which brings up one last question--sorry i have lots of questions. does the bridge need protection against kickback from the primary coil? or is using diodes of overrated voltage enough?

thanks

p.s. my emoticon menu won't work!

MOT_man

Mon Mar 03 2008, 06:50AM

Registered Member #1127 Joined: Mon Nov 19 2007, 12:08AM
Location:
Posts: 139

Rather an easy task...

If you go MOT - a single MOT can be ramped up to 10 kVDC using a 2 stage multiplier. I believe Scopeboy -

Has a rather interesting page on it. Its full of good stuff!!
I have tried the same modification and yes - I got rather awesome results. 2 MOT's in series with a similar set up produced 100" discharges point to point -
So yes, it is very easy to get good results.

Try it out - have fun!

Dr. Slack

Mon Mar 03 2008, 08:26AM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

how long (in seconds) it will take a particular capacitor to completely charge

infinitely long, however, if you replace "completely" with "sufficiently" you then get some calculable answers

i don't want anything fancy with inductors and 2X input voltage

a charging indcutor is not really "fancy", it's practically essential. You *can* charge a capacitor with a diode bridge and a resistor, but
a) you generate a lot of heat in the resistor
b) not only don't you get the "free" 2x multiplication of an inductor, you don't even get the full input voltage, unless you wait for an infinite time
c) you charge slowly
however it is cheap and cheerful
and you do need the resistor

The performance of resistor versus an inductor + diode is chalk and cheese.

The diffiuclty is sourcing the inductor. Scopeboy bought some 2.5H inductors that were IIRC intended for plate bias in valve amps. I am using several flourescent chokes in series. Many use old transformers and create an airgap in the core (don't do it without an airgap). The diode is easy, 1N5408s are dirt cheap, string enough in series.

Electroholic

Mon Mar 03 2008, 08:52AM

Registered Member #191 Joined: Fri Feb 17 2006, 02:01AM
Location: Esbjerg Denmark
Posts: 720

Conner used a two stage half wave CW made with strings 1N5408 and 450V lytics (with resistive and capacitive balancing), and I'm pretty sure he used 4x 12.5H in series, which gave 30H total.

As for capacitor charging, just divide the energy capacity of the tank cap by your power input.

BTW, two stage is enough to get 10kV.

Sulaiman

Mon Mar 03 2008, 12:43PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

Assuming that you're using a low-current high-voltage source
(such as several camera flash inverters in series)
you can use Charge (Q in coulombs) = I.T=C.dV
so T=C.dV/I

e.g. C=10 nF, dV(change in capacitor voltage)=10 kV and I=1 mA
T= 10E-9 x 10E3 / 1E-3 = 0.1 second ... about 10 sparks-per-second.

Usually the inductance of the transformer secondary is enough to protest the circuit.
(for such low power inverters, NOT for high power inverters)

northern_lightning

Mon Mar 03 2008, 06:17PM

Registered Member #1348 Joined: Sun Feb 24 2008, 01:48AM
Location: Michigan's Upper Peninsula
Posts: 22

it looks like there're too many benefits not to use DC res charging. so DC resonant charging it is!

i think i'll use a bridge, followed by a smoothing cap just large enough to achieve reasonable ripple, and then all the DC resonant charging stuff ; ) that should give me a fairly smooth 5.65kv charging voltage for the tank : )

Sulaiman

Mon Mar 03 2008, 08:46PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

OOPS !! .. I misunderstood the question;

IF you have a good high voltage source that doesn't droop etc.
and use a charging inductor, then it's a completely different scenario;

Immediately after a discharge there will be some voltage of either polarity remaining on the capacitor
due to primary resonance and the spark gap 'extinguishing' with some primary energy left;
for a simple analysis I shall ignore this small offset.
So the primary cap starts with zero voltage/current/energy, as do the inductors.
Until the spark gap 'fires' the primary inductor can be ignored.

The circuit 'looks' like a series inductor (the charging choke) and capacitor (the primary capacitor)
The voltage across the charging inductor is equal to the dc supply voltage.
The resulting voltage across the primary capacitor will look exactly like a -cosine wave,
starting at zero passing Vdc and going to 2 x Vdc.
If there is no diode in series with the inductor then the capacitor voltage will continue on the cosine path
and pass back through Vdc down to 0V and continue oscillating at F = 1 / 2.PI.SQRT(Ldc.Cpri)
with the oscillations dying down due to losses, eventually the voltage on the primary capacitor will be Vdc.

With a series diode (often called a 'de-Qing diode') the capacitor voltage will remain at 2 x Vdc.

The spark gap should be set to 'fire' at a voltage slightly less than 2 x Vdc.

The time to charge the capacitor will be 1/2 of a cycle at F = 1 / 2.PI.SQRT(Ldc.Cpri)

You may have spotted the problem with this arrangement - how does the sequence start
if the spark gap is set to fire at almost 2 x Vdc ?

Paul Benham

Tue Mar 04 2008, 03:51PM

Registered Member #570 Joined: Wed Mar 07 2007, 03:41PM
Location: Winchester Uk
Posts: 42

You will need to use a rotary spark gap to allow charging to twice the DC voltage. In practice it never gets quite that high.

I use 14H for my charging inductor and it consists of four MOT secondaries in series. The I's have been removed from each MOT using a hacksaw and sawing through the welds. I have 10kV DC approx and it has not flashed over or failed yet.

Cheers,

Paul.

teslacoolguy

Tue Mar 04 2008, 03:55PM

Registered Member #1107 Joined: Thu Nov 08 2007, 10:09PM
Location:
Posts: 792

if it is not that nig of a coil you dont have to use a rsg and a let's say 10 stage spaced 1mm apart multigap will work just fine but a rsg always works better

northern_lightning

Wed Mar 05 2008, 01:22AM

Registered Member #1348 Joined: Sun Feb 24 2008, 01:48AM
Location: Michigan's Upper Peninsula
Posts: 22

thanks for all the info.

Sulaiman wrote...
Usually the inductance of the transformer secondary is enough to protest the circuit.
(for such low power inverters, NOT for high power inverters)

is there a simple explanation as to why this happens?
________________________________________________

and back to resonant charging...

You will need to use a rotary spark gap to allow charging to twice the DC voltage. In practice it never gets quite that high.

if it is not that nig of a coil you dont have to use a rsg and a let's say 10 stage spaced 1mm apart multigap will work just fine but a rsg always works better

what about a single gap with the electrodes spaced far enough apart to allow ~2X the DC voltage and a strong fan/ vaccum to quickly extinguish the spark between firings?

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob