HT 'low current' resistance fundamentals in Tungsten Welder - event horizon

[marcophys]

Does anybody understand the fundamentals of what is happening in this circuit?
Tungsten welding info is all about high current... but this is a low current scenario.
AC input, CDI Coil generated 20Kv (estimated) suffers voltage drop over 100mm of tungsten rod.

The larger the diameter... the less voltage drop... Yet...
I believe a pointed electrode will produce a better spark... but by creating a point, the resistance is increasing, as the diameter gets smaller, to effectively 0.0mm diameter.

Is there an event horizon, where the spark is going to jump, regardless of diameter?
What is exactly happening, as the spark jumps?

But more than that... how can we get a grip of the fundamentals at play here?

For example: I don't even know what the coil output voltage is, as a multimeter cannot test HT output.
Searching on the web... people throw figures around, but nobody seems to actually 'know' what is happening.
Official figures don't seem to be quoted, presumably because the public cannot test HT output.

Further... if a 5k resistor is used in the plug cap... this dramatically alters the equation results of Ohms law

I'm clearly floundering.

I don't know the HT voltage
I don't know the current
I don't know by what degree a shallow taper impedes voltage transmission.
(welders go for a very shallow taper... but they use high currents... maybe that changes things)

Does anybody understand the fundamentals of what is happening in this circuit?
And, or, how to measure the HT output?

 

[zoki85]

The larger the diameter... the less voltage drop... Yet...
I believe a pointed electrode will produce a better spark... but by creating a point, the resistance is increasing, as the diameter gets smaller, to effectively 0.0mm diameter.

Is there an event horizon, where the spark is going to jump, regardless of diameter?
What is exactly happening, as the spark jumps?

The "event horizon", as you call it, could be a breakdown el. field of normal air (E=2-4 kV/mm).
After the spark jumps, arc voltage is low and current magnitude depends on the source's impedance.
Minimum current required for arc sustentation differs from case to case.

 

[marcophys]

The "event horizon", as you call it, could be a breakdown el. field of normal air (E=2-4 kV/mm).
After the spark jumps, arc voltage is low and current magnitude depends on the source's impedance.
Minimum current required for arc sustentation differs from case to case.

Can you expand on this?
How do we use this, in understanding that an electrode ground flat, has less resistance to a pointed electrode.

Do we lose any voltage along the taper?
Or is it so close that the spark has already been structured, and is just being channelled to produce a larger concentration of energy (at the point)?

And does the 'point' work, in a low current scenario.
(Ie. Why did almost all spark plugs have just a slightly domed electrode... rather than a point? )

 

[Windadct]

The breakdown Voltage of the air is typically defined between two parallel surfaces, the sharp point (and high frequency) increase the voltage gradient and stress on the air(like a lightning rod), and the spark jumps (plasma) - typically this is only used for the start ( HV Start - used in TIG and MIG (MIGAW)-VS Scratch start) - then the welder switches to current control ( well the better welders do IMO- as opposed to V control). Until there is current flow there is no voltage drop along the 100mm Rod - all of the voltage is across the air-gap.
now the plasma gives off heat according to P = V * I, and this heat is used for to melt the metal. In a welder the challenge is arc stability, so the weld has a uniform heat input and proper temperature - the higher the current usually the more stable the arc, so in this mode high current / lower V is usually used.

 

[marcophys]

The breakdown Voltage of the air is typically defined between two parallel surfaces.
The sharp point (and high frequency) increase the voltage gradient and stress on the air(like a lightning rod), and the spark jumps (plasma).

Typically this is only used for the start ( HV Start - used in TIG and MIG (MIGAW)-VS Scratch start),
then the welder switches to current control ( well the better welders do IMO- as opposed to V control).
​

Until there is current flow there is no voltage drop along the 100mm Rod - all of the voltage is across the air-gap.
Now the plasma gives off heat according to P = V * I, and this heat is used for to melt the metal.

In a welder the challenge is arc stability, so the weld has a uniform heat input and proper temperature - the higher the current usually the more stable the arc, so in this mode high current / lower V is usually used.​

​

Thanks for that... the thread seems to be homing in on the crux.

The breakdown Voltage of the air is typically defined between two parallel surfaces.
So there are laws relating to this 'breakdown voltage of the air'.

The sharp point (and high frequency) increase the voltage gradient and stress on the air(like a lightning rod), and the spark jumps (plasma).
So, am I right in my understanding... that by creating a point on the electrode: the voltage increases as it advances up the taper, to the point?

Now you say:
Now the plasma gives off heat according to P = V * I, and this heat is used for to melt the metal.

Okay... so with a point, the voltage has increased, and we've created more heat... the plasma is hotter... but presumably there is less of it (plasma).
EG. with parallel electrodes (say 3mm diameter)... the same energy is created, but the spark/plasma is larger, so cooler (compared to point to point)

If my understanding of what you said is correct... a point creates the maximum heat, in the smallest possible area.
Grinding the taper to leave a flat 'circle' allows one to lower the heat, whilst increasing the what?... the spread of energy (fatter spark)?

So can we say:
If we achieve the required temperature, then a fatter spark is the way to go.
(like placing a steak on a thin pan at correct temp...it goes cold immediately...
... better a thick pan which holds more energy.

Do we know this... or are we guessing?
Cos I can grind the electrode to a point, or leave a circle, or grind it flat.
(I actually have to make that choice!)

I guess the question is: whether a 'small high temp spark' is better than a 'fat lower temp spark'.
This presuming that there is no voltage drop on the taper... but rather... the voltage increases.

However...the concern remains:
That most spark plugs were designed 'domed' rather than tapered.
The above, doesn't explain that real life result... or does it, and it's just me that's missing it?

 

[zoki85]

I guess the question is: whether a 'small high temp spark' is better than a 'fat lower temp spark'.
This presuming that there is no voltage drop on the taper... but rather... the voltage increases.

It is a "small low temp spark" vs a "fat high temp spark".
But do you know how V-I characteristics of plasma arc looks like?
If you don't, google it.

 

[Windadct]

As for "Laws" regarding breakdown voltage - more like standard ways to measure and describe, this really applies to most materials. For air (and many fluid insulators they measure between 2 flat circular electrodes - or 2 semi-rounded electrodes to try to get a uniform gradient ( voltage change over a distance in the material) in the gap between the electrodes. The Sharp point does not increase the voltage - you could say it focuses the stress caused by the voltage on a smaller area of the material ( in this case air). Similar to poking a balloon with your finger vs with a pin - the same force is applied, but the pin applies all of that force in a very small area and the balloon "fails".
For welding - everything I am finding ( and my memory - I worked in robotic welding systems) - the HV is only for ignition of the spark, once the arc is established and you are welding you are operating in high current mode. Basically the plasma becomes the conductor - and much lower resistance than air.

Okay... so with a point, the voltage has increased, and we've created more heat... the plasma is hotter... but presumably there is less of it (plasma).
EG. with parallel electrodes (say 3mm diameter)... the same energy is created, but the spark/plasma is larger, so cooler (compared to point to point)

If my understanding of what you said is correct... a point creates the maximum heat, in the smallest possible area.
Grinding the taper to leave a flat 'circle' allows one to lower the heat, whilst increasing the what?... the spread of energy (fatter spark)?

So the sharp point makes it easier to initiate the arc (voltage has not increased) and not necessarily created more heat. You may be creating a smaller hotter plasma - but total heat (energy) is about the same. As far as shaping the electrode - we used TIG for more detail and unusual metals ( Inconel, thin stainless) for most we used MIG - even for SS and ALU - so the wire was the electrode, in general the larger the wire the more current - all in all you are balancing a number of parameters ( Current, feed-wire speed, electrode distance, stitch, travel speed, gas mix, work piece temperature etc... - all to control total heat, HAZ, penetration etc... - it's an art really)

So the shape of electrode, amount of current and amount of added metal - all are determined by the weld needed - but I did not work with shaped TIG electrodes too much. Small detail in SS needs a small plasma , lower current - SS does not conduct heat well and the material can be seriously (negatively) affected by the heat. For a larger structural ALU work piece you may want a lot more current (=heat) and a lot of added metal - and you may have to move & stitch pretty fast. For the robotic welds we set up - a person could not do it ( consistently - all day...).

As for the TEMP I would also say probably abut the same ( I am thinking plasma temp is PROBABLY one of the keys to stability - but you can not measure or control the actual temp) - the small arc is less heat, big arc more heat.

As for a Spark Plug - the dome is to provide a UNIFORM gap over a long period of time. A sharp point would degrade pretty quickly - so you leave the mechanic all tuned up and 2000 mi later your MG is running like -- well an MG.

 

[marcophys]

It is a "small low temp spark" vs a "fat high temp spark".
But do you know how V-I characteristics of plasma arc looks like?
If you don't, google it.

I found this
https://mysite.du.edu/~jcalvert/phys/dischg.htm

... it is very thorough and detailed in its explanation... it will take some studying
It states:
If the two points are separated by a vacuum, there can be no discharge. The transfer of matter between the two points is necessary, since only matter can carry electric charge.
It seems to be stating that it is the air carrying the charge... though surely there is a transfer of material from the anode to the cathode... or not?

I note you are suggesting that the pointed electrode produces a fat high temp spark.
If this is the case, then I should be proceeding with a point... perhaps leaving just a small circle, to aid gapping measurement.

The Sharp point does not increase the voltage - you could say it focuses the stress caused by the voltage on a smaller area of the material ( in this case air). Similar to poking a balloon with your finger vs with a pin - the same force is applied, but the pin applies all of that force in a very small area and the balloon "fails".

So the sharp point makes it easier to initiate the arc (voltage has not increased) and not necessarily created more heat. You may be creating a smaller hotter plasma - but total heat (energy) is about the same.

As for the TEMP I would also say probably abut the same ( I am thinking plasma temp is PROBABLY one of the keys to stability - but you can not measure or control the actual temp) - the small arc is less heat, big arc more heat.

As for a Spark Plug - the dome is to provide a UNIFORM gap over a long period of time. A sharp point would degrade pretty quickly - so you leave the mechanic all tuned up and 2000 mi later your MG is running like -- well an MG.

In this use, the spark is generated by capacitor discharge.
The cap is filled by way of a rotating component, at each pass filling the cap a little more, until the timed release of the charge.
See here an excellent explanation with a live graphic


We want the biggest spark possible, hot enough to ignite a fuel air mix under compression.
From the 1st quoted doc... I tried to discern what is happening to the spark under compression, but failed.

Anyway, from common knowledge... a weak spark fails under compression.

Re the spark plug domed anode
This must be because the anode is not hard like tungsten or iridium.
Iridium anodes are tiny 0.6mm in some cases.

Why they chose iridium over tungsten... I can only presume that it creates a better spark, and possibly lasts longer than tungsten.
Though this is not like welding.
I would presume a slightly flattened point will last many thousands of miles in an engine.

Nobody has commented on the paradox:
... of the reducing diameter of the taper, increasing resistance no?

If I used a 1mm D electrode, the spark is weaker than if I used a 3mm D electrode.
I see it like a small tube and a large tube carrying water under pressure... at any point in time the large tube is carrying more water.
Is that fair?

So if the 3mm anode is ground to a taper... as it goes to a point... is it gradually restricting the potential difference?
(because the spark is weaker with a small diameter electrode, for the same voltage)

The current is low (but how low, I don't know).
It's high voltage, low current.
I've had a few belts from spark plugs @ >10kv but with low current it doesn't kill us.​

 

[zoki85]

I note you are suggesting that the pointed electrode produces a fat high temp spark.

No. I suggested that higher currents produced a fat, high temp arcs.

 

[marcophys]

No. I suggested that higher currents produced a fat, high temp arcs.

Ah... Okay...
The problem/question revolves around a 'low current' circuit.

We know this works... but I don't understand the part that 'current' plays, in creating this spark.

We can't change the current/voltage... as that is being delivered by the CDI.
We just need to decide on a taper or no taper, or angle of taper and size of the 'flat spot'.

Understanding what is happening at the taper, might help us better specify the taper.

 

[zoki85]

If the two points are separated by a vacuum, there can be no discharge. The transfer of matter between the two points is necessary, since only matter can carry electric charge.

Hey, first sentence isn't entirely correct. There are and can be electrical discharge between two electrodes in a complete vacuum. Such discharges occur in HV vacuum circuit breakers. Much higher electrical fields than in normal air are needed.

 

[marcophys]

Hey, first sentence isn't entirely correct. There are and can be electrical discharge between two electrodes in a complete vacuum. Such discharges occur in HV vacuum circuit breakers. Much higher electrical fields than in normal air are needed.

Oh dear... a disagreement at such a fundamental level.
It doesn't impact on our circuit, but still...

In this respect... the question is whether material from the anode or cathode makes the jump.
Perhaps the required matter (to carry the energy in a vacuum) is material from the electrodes?

Remember contact breaker points on old cars.
One contact would build, while the other would crater.

Perhaps very little material leaves the tungsten because it is so hard?

 

[zoki85]

Oh dear... a disagreement at such a fundamental level.
It doesn't impact on our circuit, but still...
In this respect... the question is whether material from the anode or cathode makes the jump.
Perhaps the required matter (to carry the energy in a vacuum) is material from the electrodes?

Under very high E-field, anode electrons make first jump. Then, if the bombardement of cathode is sufficient, a cathode "spot" forms. Then, metal material starts to evaporate and flow through interelectrode space.

 

[marcophys]

Under very high E-field, anode electrons make first jump. Then, if the bombardement of cathode is sufficient, a cathode "spot" forms. Then, metal material starts to evaporate and flow through interelectrode space.

Our circuit will function between 10Kv - 30Kv depending on the coil being used.

Does the metal material only fly across, due to evaporation... in which case, if the tungsten doesn't melt... it won't degrade?
I believe tungsten has the highest melting point of all metals (3,422 °C)

Tungsten seems to be a good choice of material.
The question then, is 'it's design'?

It has been stated that the pointed anode focusses the energy to increase the stress on the air (gas).
This improves the chance of creating a spark... that's a big positive.
I wonder if at its very point, it would melt?

But anyway, I think a perfect point is not practical... you would just scratch your gauge every time you inserted it :)
Leaving a small flat spot on the end of the taper, may not overly impact on the energy focus.
It would still be pointed.

The question over the voltage drop, as the conductor narrows, hasn't been explained.
You would think that the taper would create a bottleneck to the flow.
Could the charge 'back up' behind the point... creating a longer sustained spark?

Apparently, grinding the taper lengthways (rather than a rotational grind), prepares the grains of tungsten correctly for transmitting the energy.​

What's going on there?
The direction of the cut across the grain, substantially effecting the flow of energy.

 

[zoki85]

Does the metal material only fly across, due to evaporation... in which case, if the tungsten doesn't melt... it won't degrade?

99,9% of plasma material in a typical current arc are electrons and gas ions .
Only in very strong arc currents, where electrodes get melted intesively, metal ions become important ingredients of arc plasmas.
Tungsten has very high melting point, but it is a very heavy material as well. And even when it melts, melted material of other electrode is more abudant and mobile in plasma. Vacuum arcs, and gas arcs can't be directly compared. Their mechanism and dynamics differ in many things.

Regards

 

[Windadct]

Oh boy -- this is getting a little out of hand...Marco - you are primarily asking about welding correct? Like a TIG torch? -- In the plasma generated, you feed a gas "shield", and this gas is actually turned to the plasma, in a plasma the electrons are stripped from the gas molecules making them Ions - the science of the gas shielding in welding - is very complex.
If you are using a stick electrode - the surface of the electrode is the flux - the heat of the plasma turns the flux into the shielding gas. ( hoever I have not heard of CDI ( high voltage start) for stick welders)

If you are talking about welding - don't bring the vacuum cases in. Each one is unique - and IMO does not help the discussion on welding. ( there is Vacuum plasma welding - again a very special case)

The "science" of the shape of the tungsten - is specific to welding. Determined more by experimentation than electrical theory. You ask about the tip melting - I believe that can occur, but then there is too much heat...by the time the tungsten melts you are overheating the work piece. I would hazard to bet that a good welder does much less tip maintenance ( sharpening) because he controls the process better.

As for

The question over the voltage drop, as the conductor narrows, hasn't been explained.

- are you asking about the increased resistance in the tip? Then this is not significant if it was wee would see the heat in the tip - and that is not where we want the heat - the resistance is "in the plasma" as the current flows though the ionized gasses. - Remember as the arc is on you are continually ionizing the gas - this takes a lot of "work" and gives off heat.

Found this

 

[marcophys]

First thing... No this is not a welding application.
I think welding knowledge has been important, because it is well understood in terms of sub-surface theory, and the application of that knowledge.

This is a 'capacitor discharge circuit' (CDI)
Description in this post above:
https://www.physicsforums.com/threa...ten-welder-event-horizon.779932/#post-4903658

We don't know the HT voltage between 10Kv - 30Kv depending on the coil being used.
We don't know the current... though considered negligible.
The gas is a fuel/air mix

The voltage rises with the increased revolutions of the magnetic flywheel around the stator.
The starting voltage available to the CDI system is 45v AC

We want the best spark possible, from the HT voltage that is delivered to the electrode tip.

I was concerned about the potential bottleneck to energy flow, if the conductor is tapered to a point, because we cannot turn up the power to compensate...Our HT voltage is beyond our control.

Windadct says:

the resistance is "in the plasma" as the current flows though the ionized gasses.

So the air becomes the conductor between the cathode and the anode.
It still doesn't explain the increasing bottleneck, as the conductor gets ever smaller.

If it doesn't matter in real terms... the question is academical.

I appreciated the video of the electrode formation @ 2:20 in the above video.

I think I'm good to go with the taper - ground to a point and then flatted, like for aluminium welding.

 

[Windadct]

Sorry TMI in the posts - I saw the ignition case but did not realize that was the actual application.

Based on this - actually a Flat tip will be harder to get the spark to jump - but for an ignition circuit a sparkplug is SOOOO engineered - it seems you are trying to re-invent one? It is all about getting the spark to jump - not the resistance of the circuit once the spark is made.

Try this - seems like some good info : https://www.fairchildsemi.com/application-notes/AN/AN-8208.pdf

 

[marcophys]

That doc was a nice find.
I will give it a thorough read.
I note that the CDI spark is 'short duration'.

It is all about getting the spark to jump - not the resistance of the circuit once the spark is made.

No... I was not concerned about the resistance 'after' the spark is made.
I was concerned about the resistance 'before' the spark is made.

In effect a 1mm dia. electrode will provide a weaker spark than a 3mm dia. electrode.
It is so weak, the gap has to be very close... too close for good operation.

Replacing most of the electrode with copper (shortening the tungsten rod) helps... but I believe a bigger diameter (of both copper and tungsten) overall is best.

But imagine 3mm dia copper... brazed to 1mm dia tungsten electrode.

The energy flow would reduce to that allowed by the tungsten no?
This is before the spark is made, but as the circuit is energised.

Re: flatted point making spark formation harder.
Yes... but I have never seen a pointed spark plug.

And from previous testing... it will strike a spark... only that I'm just looking at establishing what is best, from a theoretical perspective.
Actual testing without a specific lab setup is impossible.

However... I will be able to hear if there are serious ignition problems.
Minuscule differences will perhaps go unnoticed.

 

[Windadct]

This point is probably lost above - if there is no spark (current in the circuit) ALL of the voltage is across the GAP - the shape of the electrode has no effect on the resistance - the sharper the point the FARTHER away you can be and still get the spark to jump. However - since the "flatter" electrode needs MORE voltage to make the jump you will probably get a stronger spark - ( so harder to ignite - but stronger) This has to do with the higher voltage - not the resistance in the electrode.
IMO - on the HV side replacing the electrode with copper will not help in the way you are thinking..

 

[marcophys]

if there is no spark (current in the circuit) ALL of the voltage is across the GAP - the shape of the electrode has no effect on the resistance - the sharper the point the FARTHER away you can be and still get the spark to jump. However - since the "flatter" electrode needs MORE voltage to make the jump you will probably get a stronger spark - ( so harder to ignite - but stronger) This has to do with the higher voltage - not the resistance in the electrode.

"the shape of the electrode has no effect on the resistance"

Hmmm... let's say we have a 3mm dia. conductor tapering to a 0.05mm dia conductor.

Examined here in a real world scenario:

Point 1.
We cannot control the voltage, (IE. provide MORE voltage)
We can only provide a larger diameter conductor and larger diameter electrode (and then shape the electrode).

We know that a small diameter conductor will limit energy flow from the coil...
... as will the use of almost any other conductor instead of copper.

I have seen diagrams somewhere (can't find the link) showing the increase required in diameter of different metals to match a given copper diameter.
Tungsten is a poor conductor 34% compared to copper 100%.

Therefore, with tungsten in the circuit, there is more resistivity.
Therefore, shortening the tungsten electrode, and increasing the copper, does help.
Hence the copper cored spark plugs.

In my tests, the spark was better when I replaced a big section of the tungsten rod with copper.

Why is this... as yet it is unexplained, but it is fundamental to the circuit?

In common parlance: a thin conductor can't deliver the punch, whereas a thick conductor can.

What is happening when we transition the conductor from say 3mm dia to 1mm dia.?
Or similarly transition the conductor from copper to tungsten?

... and therefore the taper... going from 3mm dia to almost zero.

Is it voltage drop?
Or is it current drop?

All I know is that the spark is weaker.

I started with 100mm of small diameter tungsten... not good.
I replaced 90mm with copper... better!

Next test is to up the diameter of both copper and tungsten.
This should produce a bigger spark.

Why?​

If we have the answer to that, then we might have the answer to the 'reducing diameter of the conductor in a taper - and why it does not behave like 'reducing the conductor diameter'.

Note Zoki said:

I suggested that higher currents produced a fat, high temp arcs.

So maybe, the already low current, is lowered even further, by any transition in diameter, or to a different metal?

Actually Zoki... The gas plasma being now understood, I thought you would have also have a crack at answering this tapering/transitioning question?​

Point 2.
You indicate that the flatter electrode may produce a stronger spark, but would require more voltage.

I did raise this point earlier, however instead, we examined how the point 'stressed the air better'.
This explains the creation of the spark beautifully... but it didn't explain how the poorly conducting tungsten can continue to conduct, as it gets smaller.

I think that another viewpoint would be welcomed, because you did say 'probably get a stronger spark'.

However, at all times we must remember... 'the voltage is as it comes - 10Kv - 30Kv ... but I don't know the current... it's low... maybe 50mA (I read somewhere, but not confirmed)'.
So when comparing performance of a 'point' to a 'flat'... it must be with the same voltage/current arriving at the electrode.

Wouldn't it be nice, if we could simply attach a multimeter to the circuit, and discover everything :(