Making low volt spark gaps

[Capernicus]

How small has anyone hand made a spark gap before?
It seems pretty tricky on the surface, Anything below 100volts would be excellent.
Anyone with any hints?

[Gyro]

Welcome to the forum.

At roughly 3000V/mm breakdown voltage in air, I think you'd have great difficulty in getting something narrow enough gap to break down at less than 100V. You'd need micrometer adjustment at the very least.

A couple of ideas that might help. Neon bulbs break down at around 90V (although once they have struck, the glow (or heavens forgive, Arc) voltage is considerably lower. You can also get dedicated Spark Gap tubes that break down as low as 90V, they're used for telecom protection (although 230V ones are more common).

Some idea of the application would be helpful.

[Capernicus]

Thanks for the welcome,  your welcome.

I've got lots of ideas in front of it, but I'm only starting small with a basic oscillation to run some blinky lights,  I'm using a dc technique to generate the voltage, its similar to a voltage doubler, its a mini marx generator that is only going to increase the voltage a little, because I dont want to make it so large, its too many parts.    If I do a 5 capacitor, it gets to about 150volts from my 30 volt beginners power supply.

Im using a version that has diodes instead of the spark gaps per capacitor,  but I need 1 spark gap to time the oscillation.

I could make a larger one,  but its just too big and clumsy, if I could do it lower voltage it would be so much easier.   Maybe I should just buy it, I bet it doesnt cost much, but I thought maybe theres some crafty way of making one that ppl already know.

[CaptDon]

A spark gap for timing? Worst idea in the world. A low voltage spark gap, also a bad idea. The smaller the gap the bigger by percentage is the point at which the arc will occur. Dust, lint and surface moisture begin to look like a car driving through your gap. Practical commercially made gaps start at around 500 volts and their tolerance is wide. Years ago it was common in CRT televisions and monitors to see a ceramic capacitor of maybe .001uf to perhaps .01uf with a built-in gap on the end away from the leads. I think your project needs a total redux!!! B.T.W., it was common to use a very small transformer which was used as an audio output matching transformer for very small transistor radios and a unijunction transistor set up to pulse the low impedance side of the transformer and rectify the high voltage pulse coming from the high impedance side to get around 100vdc and feed that to some NE51 neon lamps with small capacitors across the lamps to make neon blinky-bulb razzle-dazzle time wasters to stare at. I built one with 3 bulbs that would flash in sequence!!!

[Capernicus]

Thats really cool,  thats what im looking for, except for more boring and just using ordinary LED's.

What do you mean by arc?

Can't you seal spark gaps pretty easily with some shrink tubing or something to stop contamination?  (But thanks for warning me.)

[perieanuo]

Thats really cool,  thats what im looking for, except for more boring and just using ordinary LED's.

What do you mean by arc?

Can't you seal spark gaps pretty easily with some shrink tubing or something to stop contamination?  (But thanks for warning me.)

https://en.wikipedia.org/wiki/Electric_arc

[Siwastaja]

I always forget the name and exact number, but there is a threshold voltage for arcing happening in air at all, so even micrometer accuracy adjustments are not going to cut it; you simply can't make 100V arc. So while 3000V might typically jump 1 mm, and 600V would jump 0.2mm, 100V won't jump 0.033mm, it won't jump even 0.001mm.

The threshold is around 200V IIRC, Google it.

The reason why lower voltages than this apparently seem to spark is due to inductances in circuits. Shorting by actual contact causes inductance (even parasitics are enough for a tiny spark) to charge up, then release higher (hundreds of V) voltage peak, however short, enough to start the spark which can be sustained by lower voltage for a while.

[Capernicus]

Isn't the smaller the distance the less of an arc,  thats what confused me,  I was thinking you (CaptDon) were meaning the machine blowing up usually happens at the smallest spark gap or something.   

A little mini 100 volt gap the spark is almost non-existant,  its what I want, because I want to lose less energy in the gap,  But I'm not in the know about that yet, the percentage loss, I have to try it out,  going from a capacitor discharge, into another capacitor and test the voltage loss. (do a discharge with and without the gap, and get the difference should be a good result.)

Just a tip,  capacitors share voltage like how resistors share the current, except in series instead of parallel.  its an interesting property I'm looking at right now.

Ive seen 60volt spark gaps for sale on the net.

[mikerj]

I always forget the name and exact number, but there is a threshold voltage for arcing happening in air at all, so even micrometer accuracy adjustments are not going to cut it; you simply can't make 100V arc. So while 3000V might typically jump 1 mm, and 600V would jump 0.2mm, 100V won't jump 0.033mm, it won't jump even 0.001mm.

The threshold is around 200V IIRC, Google it.

Paschen's law.  Looks like 327v with a 7.5um gap is as low as it gets at one atmosphere.

[Capernicus]

I see!
If you could be so kind what would a 35 micrometer gap come to?
That graph, I cant read it any good...

[Terry Bites]

Small gaps are not good at self extinguishing. Ionised air will hang about making your spark gap characteristics very (very) unpredidcable. Stop while you can still see the shore.

[Capernicus]

Paschen's law is wierd,  it means once you get close enough with two tracks,  it only jumps once you bring the wires apart from each other.

Electricity does not seem to make sense at times.   Inverse spark gap. 

Also it means in cpu's, theres no chance for the electricity to jump wires at all - because its all in micrometres.
or... actually it would, but it would jump alot of tracks at once to get to the right further distance i guess...

[TimFox]

Paschen's Law  (see  https://en.wikipedia.org/wiki/Paschen%27s_law  ) gives the breakdown voltage as a function of the product of gap distance times absolute pressure.  This is often called a "glow discharge", and is the type of discharge in neon bulbs.  The curves show a distinct minimum, where the voltage increases rapidly at low (pressure x distance, i.e. a good vacuum) values and slowly at higher values.
In air, this is at p x d  = 7.5×10⁻⁶ m·atm. ("atm" means atmospheric pressure, roughly 100 kPa)  This is 327 V in air at standard atmospheric pressure at a distance of 7.5 μm.
Starting from a given condition (such as the above), one can increase the breakdown voltage by either decreasing or increasing the product p x d, by decreasing or increasing one or both of the two variables.

[mikerj]

I see!
If you could be so kind what would a 35 micrometer gap come to?
That graph, I cant read it any good...

It's not a very useful chart for fine details, but plotting the values in e.g. Excel gives something more useful.  A 35 micrometer gap @ 1 atmosphere has a breakdown of 450 volts.

[Gyro]

If you are going to go down this (rather strange) way of flashing your LEDs, then your best method would certainly be to use a spark gap tube - or more commonly GDT (Gas Discharge Tube) as I mentioned previously.

Eg.  https://uk.farnell.com/epcos/b88069x0720s102/surge-arrester-90v-axial/dp/2529288?st=gdt

These operate with larger gaps with a low pressure mixture of specific gasses and often a small radioactive source to promote ionisation and triggering. This avoids the atmospheric pressure problem.

Incidentally there was once a fairly popular gas lighter in the UK (in fact it was offered as project and kit in Practical Electronics magazine) which used a simple inverter to step up 1.5V to around 300V. This was used to charge a capacitor to the point where a GDT in series with the primary of a high ratio step-up pulse transformer would trigger, discharging the capacitor and generating the final spark from the secondary. This cycle would repeat about 4 times per second. Despite the high discharge currents and rate, the GDT would survive for thousands of cycles before failure.


EDIT: I even found the magazine issue - July 1975      https://worldradiohistory.com/UK/Practical-Electronics/70s/Practical-Electronics-1975-07.pdf

[Capernicus]

Yeh thats what I need.   They are cheap as chips too!

The magazine was a nice read too.   That lighter looks well made.

Micro spark gaps wouldnt be able to ignite gas I dont think,   way too small.

[amyk]

Interesting further reading: https://en.wikipedia.org/wiki/Spark_micrometer

[penfold]

I always forget the name and exact number, but there is a threshold voltage for arcing happening in air at all, so even micrometer accuracy adjustments are not going to cut it; you simply can't make 100V arc. So while 3000V might typically jump 1 mm, and 600V would jump 0.2mm, 100V won't jump 0.033mm, it won't jump even 0.001mm.

The threshold is around 200V IIRC, Google it.

Paschen's law.  Looks like 327v with a 7.5um gap is as low as it gets at one atmosphere.

Not that its terribly relevant, but worth correcting a bit of a misconception...
 
The problem with Paschen's law is that it does not scale very well for atmospheric pressures and particularly small gaps or non-uniform field distributions. When you push below a certain separation with "plane" fields the tendency is for discharges to initiate from electrons tunnelling out of the surface (with real electrodes surface imperfections offer plenty of local field enhancement that the necessary fields strengths come about at the relatively low potentials) which is a mechanism that doesn't show up in the ranges of pressure and potentials for which Paschen's law is typically relevant.

The ~300V numbers that crop up as a lower limit of arcing are more of a limitation of Paschen's law than a fundamental limitation, in the context of the OP's question, whilst it would be possible, any atmospheric air spark gap would have have such a considerable leakage below the threshold that it would be impractical.