Calculation of core area required for spark

[mercurial]

Hello,

I'm trying to play with small sparks and was tinkering around with few circuits.
I was trying to achieve sparks with a single pulse similar to a ignition coil. I was wondering what determines the size(core area) of a ignition coil for single pulse sparks.

I started with a small ferrite toroid 10mm in diameter and wound few turns in primary and 200 odd turns in the secondary, and discharged a capacitor in the primary, the capacitor was charged to 30 volts, since the turns ratio was 60:1 i expected around 1.5 to 2Kv in the output, but I couldn't see any spark.
The primary ohms was around a few milliohms and the secondary was around 20ohms, the gap that I kept was around 0.4 mm keeping in mind that air has a 3KV /mm breakdown.

I'm wondering why this didn't work, is the transformer not producing the desired voltage or is there any definitive calculation for core area or any rough way to estimate the  minimum core area required, mine was a small toroid cross section area of 8mm2.

[Jwillis]

Core area determines the power of the transformer. Represented as VA or Volt Amperes. For example a 100VA transformer can handle 100 volts at one ampere (amp) of current. VA can be directly converted to Watts. Small cores have less power than larger cores at a given frequency.

The sparks are produces when the field discharges. If  DC (Direct Current) is applied to a transformer it will charge but won't discharge until the voltage is removed. So you use AC (Alternating Current) to achieve sparks because AC will charge and discharge every cycle.

You also need enough voltage to break the dielectric strength of air to create an arc (spark). Dielectric strength increases as the distance between the secondary terminals increases. 10mm of air has more dielectric strength than 1mm of air.   
The dielectric strength of air is 30 kv. per cm . So to get a spark across 1cm of air you need 30000 Volts. 1mm of air would require 1/10 of that value or 3000V.   
A 60:1 transformer with 30V at the primary would give you 1800V. Turn ratio is equal to Volts ratio. Minus losses of course. 

[mercurial]

I couldn't understand the part regarding DC vs AC,
Currently I'm pulsing the transformer with DC voltage a short pulse using a mosfet.
Does AC have more advantages over DC for spark creation.

[Terry Bites]

The input is DC, the output is AC.
The indction coil is not acting as a conventional transformer. Other factors determine the input to output voltages.
Closing the switch charges the primary winding which is acting as an inductor. The energy stored in the primary is given by 1/2 (Li^2)
When the circuit is interupted, the magnetic field decays releasing the stored energy.
This creates a high voltage spike of back emf. [infinite in an ideal indcutor]
This waveform will couple into the secondary but the energy transfer is poor.

Adding a capacitor across the switch completes an LC tank circuit. When the switch the opens , a high voltage under damped oscilation occurs. This burst is very many times higher in amplitude than the DC source. This high voltage ringing waveform couples into the secondary where it multiplied by the turns ratio. A a car coil will have a turns ratio of about 100:1. It is a poor ac transformer with a lot of "leakage inductance" ie loose coupling. Its designed for a fairly low Q.

Its an unregulated free runing flyback converter in essence. You need a high voltage switch with a prtotective clamp, say an MOV. To het the HV, you can stack identical MOVs if you needs be. The MOV has to soak up a lot of energy so it needs to have a substantial power rating.  A stack will help here too. I suggest using a brutal HV IGBT rather than a whinging MOSFET in this kind of project.

[themadhippy]

Currently I'm pulsing the transformer with DC voltage a short pulse using a mosfet.

Cor posh,before i hit my teens   i  was using a  DC (mechano) motor in series with the door bell transformer to generate sparks and loose the trust of my mates,hold these wires its only a battery

[Jwillis]

I couldn't understand the part regarding DC vs AC,
Currently I'm pulsing the transformer with DC voltage a short pulse using a mosfet.
Does AC have more advantages over DC for spark creation.

My Apologies for not explaining  further . Not to insult your intelligence but many people I have encountered wonder why a steady DC voltage won't pass a transformer.
As Terrry Bites explains if you pulse DC, the primary coil charges when the pulse is high then discharges when pulse goes low.
So essentially, in laymens terms, it is still an alternating current, Alternating between 0 volts and some other voltage.

[mercurial]

The input is DC, the output is AC.
The indction coil is not acting as a conventional transformer. Other factors determine the input to output voltages.
Closing the switch charges the primary winding which is acting as an inductor. The energy stored in the primary is given by 1/2 (Li^2)
When the circuit is interupted, the magnetic field decays releasing the stored energy.
This creates a high voltage spike of back emf. [infinite in an ideal indcutor]

This waveform will couple into the secondary but the energy transfer is poor.
Can you please explain, Why is the energy poor?

Adding a capacitor across the switch completes an LC tank circuit. When the switch the opens , a high voltage under damped oscilation occurs. This burst is very many times higher in amplitude than the DC source. This high voltage ringing waveform couples into the secondary where it multiplied by the turns ratio. A a car coil will have a turns ratio of about 100:1. It is a poor ac transformer with a lot of "leakage inductance" ie loose coupling. Its designed for a fairly low Q.
I never knew they actually resonate the pulse with a cap, how many cycles would the oscillations last?, What causes it to be a poor ac transformer, is it by design? Is the core make with lossy material??

Its an unregulated free runing flyback converter in essence. You need a high voltage switch with a prtotective clamp, say an MOV. To het the HV, you can stack identical MOVs if you needs be. The MOV has to soak up a lot of energy so it needs to have a substantial power rating.  A stack will help here too. I suggest using a brutal HV IGBT rather than a whinging MOSFET in this kind of project.
I tried a few low voltage mosfets around 100V and they get damaged, then I scoped the drain and saw huge kickback around 250V dc when the current actually switches off, on further research I found that this could be reduced with a snubber across the switch or a zener clamp, i was using a ferrite toroid, is ferrite a good core for this application or do we need gapped cores or powdered iron cores?