Where's the spark ??

Cutting magnets was an early production use for EDM. Before digital watches, Timex cut the little kidney-shaped alnico magnets for their electronic watches on an Eltee-Pulsitron ram-type EDM. The electrodes were pieces of copper tubing squashed into the kidney shape and plunged straight into the alnico. They ganged something like 30 or 40 electrodes into a tool block.

Ed Huntress

--
No dimensional changes because of mechanical stress or heat of
machining.

Kewl !!

thanks Ed

Um... You don't know? And you just IGNORE the breakdown equation without comment? OK, I'd like to hear your reasoning for why that "Paschen rule" isn't important. We can send your discovery to Jim Lux and he can add it to the High Voltage Handbook.

:)

Stick welding has nothing to do with the electrostatic breakdown of normal air which the OP was asking about. To start a welding arc you have to TOUCH the stick to the workpiece. Once the arc has formed the physics (and the voltages involved) are entirely different, and you can pull the electrodes apart to a fairly large distance. The physics is then that of e-fields within conductive plasma which keep the plasma "alight." Seen a "jacob's ladder?" The spark can only leap across an inch or so, yet once it has appeared it can grow to many inches in length. The original question was about that spark-leaping phenomenon, not about the length such a spark can grow after it has started.

The rules and handbooks you're talking about are things with which I'm not familiar, but it sounds like this is being made more complicated than it is.

The dielectric strength of air is around 3 x 10^6 V/m. That means that you can create a spark with something like 90V at a gap of 0.0012 inches.

That's within the gap range of a typical EDM. Gaps run from perhaps 0.0002 in. for fine-finishing to maybe 0.005 in. or a little more for roughing.

It doesn't take a very high voltage to initiate a spark if you have a sensitive servo mechanism to maintain a close gap.

Ed Huntress

I'm afraid it's 'way down on a very long list, Paul. I know EDM, and at one time I probably was writing more about it than anyone else in North America. I've worked for two EDM companies (Sodick and Mitsubishi) and I've researched EDM power supply designs, including their history.

My interest is practical, not scientific. In this case, it seems to have left me in a less confused state.

Be assured, EDMs, particularly modern ones, run at open-circuit voltages well below those suggested by some here. The one I'm most familiar with had a maximum open-circuit voltage of 350V. That was 20 years ago. But there have been EDMs that run well with a maximum open-circuit voltage of 90V.

Ed Huntress

Bill Beaty is correct. The minimum breakdown voltage for air is at STP is around 300 volts. Dielectric breakdown of air is a complex process. Free electrons in the gap must be sufficiently accelerated by the electrical field so that they can create additional electrons when they collide with neutral air molecules. Furthermore, the rate that new free electrons are created must be greater than the rate that existing free electrons are being lost through recombination - the average lifetime of a free electron is only about 11 nanoseconds in air at STP. If the rate of free electron creation exceeds the rate of loss, then "avalanche breakdown" of the air occurs and a spark jumps the gap.

If you are to the right of the minimum point on the Paschen Curve, decreasing the gap size will continue to decrease the breakdown voltage in a fairly linear manner since there is still sufficient distance for electric field to accelerate electrons in the gap sufficiently to trigger avalanche breakdown. However, once you reach the Paschen minimum (about 320 volts for a 1 mil air gap for air at STP), any further decreases in gap distance will require a HIGHER electric field to break down the air. Because the gap is so short, a higher E-field is needed so that electrons can reach the velocity needed to trigger avalanche breakdown before disappearing into the positive electrode. The commonly accepted Paschen minimum sparkover voltage for air is 320 volts at a distance of about 1 mil. It can be more or less for other gases.

However, much of the above is moot since most EDM'ing is done under a dielectric fluid using actual mechanical contact and high pulse currents to "blast" away the small areas that make contact as the working tool/wire is slowly advanced into the work piece.

-- Bert --

--
--------------------------------------------------------------------
We specialize in UNIQUE items! Coins shrunk by Ultrastrong Fields,
Lichtenberg Figures (electrical discharges in acrylic), & Scarce OOP
Technical Books. Stoneridge Engineering -- http://www.teslamania.com
--------------------------------------------------------------------

No problem Ed

I only know Paschen's Law because it is in an obscure book at work on insulating oil by Frank Clark.

Otherwise, (I guess now with net it is easy to find, here it is

but otherwise many EEs have not heard of it.

Your experience would count for a lot.

all the best Paul

Ed Huntress wrote:

Most interesting, tending to make Paschen's Law somehow not practical here or over-ridden !!

Happy New Year

NO physical contact, Bert. Ever. Physical contact and a short circuit, or even an arc that lasts more than a few milliseconds, usually will destroy the workpiece and/or the electrode. As I said, servomechanisms are employed to keep the gap somewhere in the range of 0.0002 in. - 0.005 in., more or less, depending upon the other operating conditions. Fine finishes are obtained at gaps of less than a thousandth of an inch.

EDM as we know it today is indeed run in a liquid dielectric, but the precursors to EDMs, called "tap busters," used air. These machines are still made. I don't know who's in the business today but two of them that were around at least through the '80s were Electro-Arc and Camman. Maybe there is some info about them on the web that would tell you their operating voltages.

In my shop we had a crude air-dielectric tap buster sold by Do-All. It had a

240V transformer, but it's not a good example because it had no automatic servo. It often *did* achieve contact, but it didn't matter, because we weren't trying to machine accurately, as one does with an EDM machine today.

I just looked at the parameters published for some currently-marketed ram-type EDMs, and I see that open-circuit voltages fall into the range of

50V - 300V, which is about where they have been since the beginning. These machines use a light mineral oil for dielectric.

Also, wirecut EDMs, which use deionized water for dielectric, seem to run around 90V open-circuit.

Ed Huntress

had

I suspect it's just some additional phenomena that we aren't accounting for here. If you're interested, here's a technical paper about EDM pre-spark phenomena that may clear it up. I believe it concerns liquid dielectrics, and they used a 600V one-shot power supply to create the observed events, but it may suggest to you what's happening with air dielectric:

To you too, Paul.

--
Ed Huntress
(remove "3" from email address for email reply)

Paschen's Law only holds for uniform fields. In other words it is valid when the electrodes have a large radius with respect to the gap distance. But if one or both of the electrodes is sharply pointed, or other shape which produces a nonuniform field, ie is hollow, has complex features, etc, the field strength is much higher at points in the gap than the simple potential divided by distance between the electrodes would indicate.

That allows the arc to establish, then cascade, ie one electron emitted from the cathode can knock loose more than one electron in the gas, which then go on to knock loose more, etc. Secondary electrons are also emitted when the stripped ions impact the cathode. These actions cause the arc impedance to fall, so that large currents can be carried between the electrodes at very modest voltages.

Pressure, density, temperature, and gas species also play important roles. For example, Argon has a 3 times lower breakdown potential than ordinary air. That's why it is the gas of choice for GTAW welding, ie arc starting and stability are much better with Argon than with other shield gases. Breakdown can occur as low as 137 V/cm in Argon at .5 torr.

As the arc begins to establish, temperature skyrockets. This grossly lowers the needed ionization potential. That's why the arc can sustain at much lower voltages than are required to strike it.

But this is all pretty much moot since EDM isn't done with arcs through a gas. It is done with arcs through a liquid. Paschen's Law doesn't apply at all in that case.

Gary

Ratios here go up to about 4:1 for actual E over uniform for a point source- like a needle perpendular to a flat surface but a bit away from it.

^ | ________|_______

A protruding somewhat rounded source is about 3.

So as Gary says you get a corresponding reduction in V.

Lets try my 4

V =~ 1350/4= 338 volts (not bad eh!)

but as Gary points out Paschen's Lay only true when gas present.

Paul

Ratios here go up to about 4:1 for actual E over uniform for a point source- like a needle perpendular to a flat surface but a bit away from it.

^ | ________|_______

A protruding somewhat rounded source is about 3.

So as Gary says you get a corresponding reduction in V.

Lets try my 4

V =~ 1350/4= 338 volts (not bad eh!)

but as Gary points out Paschen's Lay only true when gas present.

Paul

Ratios here go up to about 4:1 for actual E over uniform for a point source- like a needle perpendular to a flat surface but a bit away from it.

^ | ________|_______

A protruding somewhat rounded source is about 3.

So as Gary says you get a corresponding reduction in V.

Lets try my 4

V =~ 1350/4= 338 volts (not bad eh!)

but as Gary points out Paschen's Lay only true when gas present.

Paul

Ratios here go up to about 4:1 for actual E over uniform for a point source- like a needle perpendular to a flat surface but a bit away from it.

^ | ________|_______

A protruding somewhat rounded source is about 3.

So as Gary says you get a corresponding reduction in V.

Lets try my 4

V =~ 1350/4= 338 volts (not bad eh!)

but as Gary points out Paschen's Law only true when gas present.

Paul

BIG SNIP OF GOOD STUFF

little snip

Hey Gary,

I think this last part of your reply is not really correct. Most EDM I see is set up and started in free air. After location and some other things are adjusted to suit the operator, then the work is flooded, either by flooding the tank to immersion, or on smallish parts or small burn area just with direct flow of die-electric externally or through internal created flow ports. After the die-electirc is applied, the rate of burn is then further adjusted to the need, which seems to me to almost ALWAYS be MAX!!

Ed Huntress says that the arc starts without actual physical contact, and I can't argue, but if so it sure isn't obvious. I always figured that on die-sinkers the electrode (always carbon in my cases) does touch and then retract ever so slightly, and now-a-days maintains the arc through sophisticated controls on the newer CNC EDM's.

Happy New Year.

Brian Lawson, Bothwell, Ontario.

The only reason they ever touch is to establish dimensional zeros, Brian. And that's done with the EDM power supply turned off.

BTW, I've never seen an EDM started in air. The electrode has to be submerged before you start.

Ed Huntress