metals!

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Clarification on thermocouple effects

"I am told that the junction between two metals creates a thermocouple. This has an effect on signals. This is why "High Quality" cables use a welding connection method so the metals "mix" to reduce this effect."


Just want to (hopefully) add some scientific clarification on thermocouple effects... Thermocouples (ie. the ones used for measirement, gas pilot lights etc) are actually manufactured by welding.

For a thermocouple (Seebeck) effect to work there always has to be a 'hot' and a 'cold' junction with a temperature difference between them to produce a voltage. Any intervening metal makes no difference to the thermocouple effect as long as the two junctions (ie, first metal - intervening metal, and intervening metal - second metal) are held at the same temperature. Soldering, for example, silver to copper, produces no more thermocouple effect than welding.

If you have a interconnect lead using, say, silver plugs and copper wire, either soldered or welded, then there will be no net thermocouple effect unless the plugs are at different temperatures... for the lead itself - there mating sockets at each end might have different contact metals, in which case there will be other thermocouple effects created.

Audio equipment is full of temperature gradients and dissimilar metals. Thermocouple effects are only a practical issue in Metrology and high precision DC instruments where great care is taken to minimise temperature gradients across dissimilar metals.
 
Talk of thermocouples at junctions is a red herring. This is for at least three reasons:
1. The thermal voltage depends on the outer metals, not whatever is between them as long as it is conductive. Solder, welding, crimping are all the same.
2. The thermal voltage is not developed at the junction! Not a lot of people seem to realise this. The thermal voltage is developed in the metal wires as they go through a temperature gradient away from the junction. The temperature gradient causes a voltage gradient.
3. Thermal voltages are DC (unless the temperature is fluctuating rapidly). Music is AC. Any decent equipment will reject DC at its input, or simply present it as a DC offset at the output (to be rejected by whatever comes next, hopefully).

Talk of thermal noise above 0K is a red herring. Thermal noise is created by resistance. If your cables are sufficiently low in resistance to transfer the signal without significant attenuation then they are sufficiently low in resistance that thermal noise from the cable will be negligible.

Talk of skin effect at audio frequencies is almost a red herring. I can't be bothered to do the maths right now (or look it up), but skin effect only really begins to be relevant a bit above the audio band. That means that exotic platings are likely to be irrelevant, except for their effect on surface corrosion. I suppose that it is just possible that surface effects could create 1/f noise (oops - I seem to have fed the trolls!) but plating might then increase the problem by adding new interfaces.

Oxygen-free etc. etc.? I haven't done the maths, but my guess is that the mean free path of an electron before it gets scattered is sufficiently short even in the purest copper that its progress is randomized in all except the most tiny conductors. Fortunately all these electrons influence each other (they are highly correlated) so we don't get noticeable amounts of shot noise from our wires.

My personal view is that all these explanations are put forward because people want to offer a 'scientific' explanation for something which is actually caused either by RF pickup (which could be sensitive to details of cable construction) or the placebo effect.
 
I wasn't going to touch this thread, but... Soldering "is" a type of welding, albeit a low temperature version.



No, soldering is not a form of "welding" it would be closer suited to low temperature brazing, where a dissimilar metal is used to join pieces of similar/dissimilar metal, where as welding comes down to joining 2 metals with a similar metal. In the case of ferious metals, you have tons of options! from steel, to copper and semi-exotic welding practices, such as carbon arc welding.
 
The base metals may not have been melted in soldering or brazing, but an alloy was created for at least a few molecules deep in the surface. As to the thermocouple, I've built several by soldering - at first the lab tech was aghast that I would "corrupt" the thermocouple, but when I walked him through it and showed him that the starting and ending metal wires were the same and it didn't matter what happened in-between, he finally understood. And I think he eventually realized that the mechanical solidity and freedom from oxidation was an improvement.

Personally I like teflon coated silver wire, but mainly for the added durability of the teflon coating.
 
..
1. Solder containing silver as or opposed to lead tin Etc

Which is the preferred %?

2. The cable question, silver plated PTFE or Teflon insulated as opposed to copper linear crystal / oxygen free copper.

Silver solder almost always is 2% silver. Electrically I doubt there is any difference because the soldered connections are first mechanically joined. For example wires are twisted together before soldering or Resistors leads are placed through holes in the PCB. The solder job should really be mechanical. But certainly some current does flow through the solder.

Silver plated wire has a slightly lower resistance at higher radio frequencies than unplated wire of the same size, at RF skin effect begins to matter. But mostly hookup wire has a diameter selected for mechanical properties and is way over-sized electrically. You do see silver plate used in some radio transmitter parts

There is a trade off too. Silver is not as reliable if the equipment is in a damp location. Dissimilar metals will corrode. so if the plated wire a nicked and the copper exposed the wire can fail.

For tube powered equipment the best hookup wire is likely solid #22 copper with the 600V PVC insulation. With tube gear lead dress and routing matters a lot, the exact location of a wire matters more than the type of wire and the stiff #22 solid wire allows you to move it and have it stay in place for decades.

The term "oxygen free copper" is right up there with "fat free watermelon". The standard for "standard" pure copper is in effect oxygen free. More details here.
Oxygen-free copper - Wikipedia, the free encyclopedia
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.