Magnetic copper from HV experiment Inventor asks for help

[Chet K]

http://www.youtube.com/watch?v=-kO8aMm9uW8

Thanks for looking
Chet
PS Many thanks to "Wesley" for translating this video

[Centraflow]

My speaker cable is magnetic and apears to be the same as his


Mike

[BEP]

It is not uncommon for lower grade speaker wires to be copper plated soft iron core. However, it was shown that only a small portion of the cable would react to a magnet.

I have a solution. Folks may take it as they please.....

ECM (electrochemical migration). This effect is a very common problem in my industry.

Proof would be to remove the magnetic portion of the wire and repeat the experiments with one catch..
Keep the humidity of the room below 80%. ECM requires the following to happen:

1. Humidity above 80%
2. Certain atmosphere contaminants (ppm can be extremely small but the contaminants aid in ionization)
3. Voltage difference (poor connections  )
4. Heat (poor connections)
5. A metal to migrate (silvered, tinned soft steel like found in most passive components and terminals)

This is why iron seems to form at a carbon spark gap.

Insulated wires are extremely susceptible to this problem. The migration only travels a short distance from where the migrating metal is.

None of it can happen unless the humidity is fairly high. It can happen at lower humidity levels but the other factors need to increase to make it happen.

Like the heat, voltage, difference between the metals, etc.

An example is ribbon control cables. In the circumstances we deal with, ribbon cables must be replaced yearly because iron from the components on the circuit board leaches up the cables and causes the insulation to crack/open. All this with only spilt-rail 15V supplies.
Another example is multi-layer circuit boards shorting between traces in the middle layers because of ECM dendrite growth.
Another is high power DC choke failure in motor drives. Iron from the copper coated and silver plated terminals migrates along the windings of the choke causing a gradual increase in choke inductance.

In the old days it was called 'whiskers' when solder grew whiskers on a junction.

All unusual to see in the public except where folks twist wires together on high voltage connections and don't expect problems  

[Chet K]

WW
Thank you for that interesting possibility,Would the Migrating metal be visible on the wire with a magnifying glass or Microscope?
thx
Chet

[BEP]

The migrating metal is generally visible but may be in an oxide form of that migrating metal.

Ex. Migrating copper can appear as green, black, copper(color) or a pink/orange. (The green and white corrosion on the positive terminal of your bad auto battery connection is ECM of both copper(green) and lead(white/black) unable to bond on either side due to the very small potential difference across the connection)

Iron may appear as black, silver(color) or rust(orange in color).
Silver may appear as Silver(color), white or brown.

With the amount of energy required to make iron migrate (not easy at low energy) the effect should look more like electroplating of iron onto copper.  

Even though copper is diamagnetic very little iron is required to make the copper stick to a magnet.

Truth be told, our check for iron ECM on the ribbon cable example is to see if the cable is attracted to a magnet  

[ion]

While I agree with WW's explanation and analysis of possible metallic ion migration in dissimilar metals, and how common it is in DC circuits, In my experience, I have never seen such migration in AC circuits. There is no known mechanism for it to occur in an AC circuit, unless the AC frequency is very low (sub Hertz).

Electroplating requires a difference of DC potential, or in effect a current flow for the migration to occur. We see the migration as WW explained in common battery operated appliances and DC electronic circuits where a difference of metals is joined in high humidity conditions, or in AC circuits that have some DC bias.

Note that the connecting wire segment was used on the secondary of a transformer (AC circuit) that fed his coil. In that short 2 foot segment of wire, there should have been some apparent magnetic gradient. Another important point is the wires were joined to similar wires, not to wires of dissimilar metals.

While I fully respect WW's expertise and life experience, however,  I suggest the "migration" explanation in this particular case is at a low probability, e.g. on the order of probability of the proposed atomic transmutation.

Hopefully the presenter will retrace his steps with a new batch of wire in controlled conditions and attempt to duplicate the "effect" such that useful data is secured.

[ion]

Did a little search and it appears copper clad steel speaker wire is quite common. See Here:

http://www.alibaba.com/product-gs/503226351/Transparent_Red_Edge_Speaker_Cable_with.html

Specifications

Speaker Cable
Good Quality & Lower Price
Conductor :copper clad steel
Specification:45*0.12mm*2*50m

[BEP]

While I agree with WW's explanation and analysis of possible metallic ion migration in dissimilar metals, and how common it is in DC circuits, In my experience, I have never seen such migration in AC circuits. There is no known mechanism for it to occur in an AC circuit, unless the AC frequency is very low (sub Hertz).

Electroplating requires a difference of DC potential, or in effect a current flow for the migration to occur. We see the migration as WW explained in common battery operated appliances and DC electronic circuits where a difference of metals is joined in high humidity conditions, or in AC circuits that have some DC bias.

Note that the connecting wire segment was used on the secondary of a transformer (AC circuit) that fed his coil. In that short 2 foot segment of wire, there should have been some apparent magnetic gradient. Another important point is the wires were joined to similar wires, not to wires of dissimilar metals.

While I fully respect WW's expertise and life experience, however,  I suggest the "migration" explanation in this particular case is at a low probability, e.g. on the order of probability of the proposed atomic transmutation.

Hopefully the presenter will retrace his steps with a new batch of wire in controlled conditions and attempt to duplicate the "effect" such that useful data is secured.

ION,

Of course ECM is most prominent in DC or AC circuits with even a minor DC bias (generally > 5V) but I would think your experience with SMPS's would have shown that dissimilar metals at junctions is not required.

A dendrite between two copper unplated traces will almost always be black to dark brown unless certain contaminants exist or the device's operation was near a sea air environment. In the latter case the dendrites may be orange to pink in color. One interesting point is that you can tell which side provided most of the metal. That copper trace will have turned brittle and crystalline. Sometimes to the point where no act will make solder stick to it again.

While most ECM follows the rules of electroplating this isn't always the case. There need only be a formation of a dielectric between charged surfaces. Once there is an ionized path between charged surfaces ECM may happen. ECM does not require current flow. All it needs is an electric field. That field may even just be from static electricity.

Metal migration does generally require that the average current direction be more one direction than the other. Pulsed DC at high frequency, as in an SMPS, is one of the first places to look for this problem.

One thing is for sure.... Out of all the research I've read on the subject none agree completely on all points. In many cases you can't predict which direction the migrating metal will propagate. Indeed, in some cases the metal migration goes one direction regardless of the current polarity. One paper I read suggested that the migration direction could only be determined by the overall electric charge of the source metal and the magnetic field direction where that charge resides. Lorentz?
Chances are good that both of his speaker wire conductors exhibit attraction to a magnet. That would just make me think that he had swapped the connections many times over a period. Once the migration happens it is unlikely to reverse direction simply because the original source of iron was still a greater source than the earlier migrated iron.

[ion]

WW

Most of the SMPS design I was involved with were low power,  ten to twenty watts off line, isolated types with lots of space between high potential traces. I must admit I have never seen the effect you speak of in these, so I will defer to your greater experience with a larger variety of high power SMPS designs.

I have a lot of experience designing thermocouple probes, but have never seen the migration occur to such an extent that the measuring junction was moved away from the tip where the wires were joined e.g. iron to Constantan (Cu Ni).

Anyway thanks for your experience on this subject, as I always learn from you.

[Allcanadian]

@WW

Of course ECM is most prominent in DC or AC circuits with even a minor DC bias (generally > 5V) but I would think your experience with SMPS's would have shown that dissimilar metals at junctions is not required.

It may be that it is not only the conductors but also the insulators that determine the effect. A cat whisker diode in crystal radio's can simply be an oxide layer between two conductors. This could bias the boundary condition between the conductors at a given threshold or act as a bidirectional capacitance. In other cases it may act as a negative resistance and produce oscillations depending on where any capacitance is in the circuit.

Imagine that, a FE device may work in one instance but not another simply because a bad connection was made at the right place, lol,

One good example is the infamous sulphated battery which may produce a non-linear capacitance. It acts like a leaky capacitor with a variable theshold and the breakdown voltage is also biased by the state of charge. It's no wonder people have trouble figuring out what's going on in certain circuits when our battery can randomly turn into an oscillator.

AC

[EMdevices]

May i suggest that oxygen can play a big part in this observed phenomena?

[BEP]

ION,

And I from you 

T-couple junctions have never come up as suffering from ECM. I imagine it is due to the very low potential difference and the junctions are generally well sealed.

The amount of power has little to nothing to do with it unless that power contributes to the circuit heat. I'm sure you are familiar with Acopian PS's and other open-frame little bricks. We see these problems in all but the hermetically sealed types, regardless of power or voltage. Paper mills and tire manufacturers have the greatest problems. Next to them, almost any installation having sea (salt) air has copper of all sizes (fine PCB traces to 110 grade silver plated bus bars) which literally rots (crystalizes) from the inside-out. This happens on aluminum structures and sheeting even when U.S. Navy paint specs are met.

When we see 'Pink-Death' (Copper Oxide(II)) on traces we generally just trash the device. I've seen five layer boards explode from the inside where the source was 3.3V traces. There are many instances where the dendrites 'un-grow' during shipping for repair. Very hard to detect. It is always fun

The solution that seems to work for PCBoards is to follow the latest recommendations for manufacture where de-ionizing each layer before bonding is required and the layers are pressed together under high pressure to the point where you can no longer find a layer with a fine knife.

Fun for all!!

Did you know that some forms of carbon are ferro or para - magnetic? These details make it difficult for us who love to experiment and then find weird results.

[BEP]

For AC & EM,

Oxygen is key and so are the insulators.

I've seen zero-gain op-amps circuits that turn into pulse generators even with no input due to ECM. The silver leached between DIP leads and formed a negative resistance oscillator junction. Kewl