replacing spark gap with semis

what type of semiconductors are well suited to replace capacitor current dumping as would otherwise a spark gap perform in its place?

the capacitor currents can exceed well over 200amps peak but it seems that even the most inexpensive thyristors can pulse handle well over 150amps.

The issue is the current on the spark gap side is high frequency ac which can cause the whole thing to quit working unless maybe its a bidirectional thyristor.

Hi,

a spark gap usually is a (cheap and unreliable) solution to prevent from high voltage.
But you talk about high current.

Please clarify.

Klaus

KlausST said:

Hi,

a spark gap usually is a (cheap and unreliable) solution to prevent from high voltage.
But you talk about high current.

Please clarify.

Klaus

Click to expand...

I was referring to those found in spark gap tesla coil arrangements.

The capacitors charge until they reach a voltage defined by spark gap length and discharge loads of current making the spark gap a high current carrying device.

The issue is which semi is ideal to switch the high peak AC currents.

A clash of wording!

Zak28, when we speak of a spark gap we usually mean a device or construction designed to deliberately arc when excessive voltage is applied across it. What I think you are referring to is a switch to discharge a capacitor into a transformer (Tesla coil in your case).

If the primary current is high frequency AC you probably don't want to discharge a capacitor directly but drive a continuous signal into it. Thyristors and triacs are not really suitable in that application because they will tend to self trigger on the rapid and sharp voltage changes. I have used power MOSFETs in a design that had to switch 200V at 1KA but they need a gate driving circuit that can supply several Amps to work efficiently.

Brian.

betwixt said:

A clash of wording!

Zak28, when we speak of a spark gap we usually mean a device or construction designed to deliberately arc when excessive voltage is applied across it. What I think you are referring to is a switch to discharge a capacitor into a transformer (Tesla coil in your case).

If the primary current is high frequency AC you probably don't want to discharge a capacitor directly but drive a continuous signal into it. Thyristors and triacs are not really suitable in that application because they will tend to self trigger on the rapid and sharp voltage changes. I have used power MOSFETs in a design that had to switch 200V at 1KA but they need a gate driving circuit that can supply several Amps to work efficiently.

Brian.

Click to expand...

In either case a spark gap is fundamentally a switch.

Those FETs sound as if they were very expensive TO-264 parts or SOT-227 modules unless you ganged a load of inexpensive FETs in which case would demand a top dog gate driver to pump loads of current.

Can you share the diagram? I prefer to use 2 high voltage capacitors in series to the primary coil such as this.



Its an uncommon arrangement but it has its peculiar effects.

I think the question is too general. Need to talk about actual discharge circuit parameters (C, L, Q, breakdown voltage) to see if there's a possible semiconductor switch replacement for the spark gap.

I guess you don't even know the exact parameters?

- - - Updated - - -

Taking the post #5 circuit parameters as granted for the time being, what's the initial voltage? The charging voltage source needs to be disconnected or supplemented with a series resistor of course.

Hi,

spark gap:
This came into my mind: https://electronic-products-design....nics/pcb-design/general-pcb-design/spark-gaps

But now I see you look for something different.

Klaus

FvM said:

I think the question is too general. Need to talk about actual discharge circuit parameters (C, L, Q, breakdown voltage) to see if there's a possible semiconductor switch replacement for the spark gap.

I guess you don't even know the exact parameters?

- - - Updated - - -

Taking the post #5 circuit parameters as granted for the time being, what's the initial voltage? The charging voltage source needs to be disconnected or supplemented with a series resistor of course.

Click to expand...

Lets make this 580v initial, caps should be 35uF instead of the 9.8nF depicted in image and closing the solid state "spark gap" switch at ~200hz or more.

The spark gap is not "fundamentally a switch". The negative
resistance characteristic is what makes a Tesla coil self-
oscillate from a DC supply.

If you want to replace it with a semiconductor (or other)
"switch", you will need the stimulus to get the oscillation
you allude to. This could be a feedback network tuned to
the "right answer", but you will also need to feed the gate
drive (or base drive) power from some auxiliary supply.

The spark gap doesn't care about any of that.

But a spark gap across an ideal voltage source of higher
than breakover value, goes by another name. That is,
"arc welder". Or "self eating plasma cutter". I think the
LTSpice cartoon is not right, about that.

A triac or SCR with suitable trigger circuit, e.g. a zener diode as most simple variant will also self oscillate. We've used similar circuits as strobe light or plasma burner igniter.

Limitations of SCR circuits are however maximum dI/dt capability and trigger speed. I'm not sure about the imagined application circuit. 580 V and 2 µF e.g. makes 290 A/µs, a bit above the rating of smaller SCRs and triacs. Exceeding di/dt will quickly damage the device because the current is only carried by part of the junction and respectively causes local overheating. Paralleling a number of SCRs and applying a sufficient high trigger current to reduce the trigger delay skew can work.

The issue is the current on the spark gap side is high frequency ac which can cause the whole thing to quit working unless maybe its a bidirectional thyristor.

Click to expand...

No problem actually. Due to the DC bias, an antiparallel diode will work.

I've never constructed a tesla coil but as far as I remember, SIDACs and IGBTs have been used in actual designs to replace the spark gap switch. Not sure about thyristors+antiparallel diodes or TRIACs. You may still want a lower primary voltage vs. basic spark gap designs. I'm not familiar with the details but there's plenty of material on the internet.