Thermal emfs: alloys

[TimFox]

Does anyone have a link to a table of thermal emfs against copper that is more complete than that provided with Fluke equipment?
Specifically, I am looking for emfs for copper alloys (different brasses, bronze, Be-Cu, and Te-Cu).

[chuckb]

Vishay has a nice list in the attached Technote.

[TimFox]

Thanks!  That is exactly what I was looking for.  I've always wondered why there is not more difficulty encountered with the huge emf of silicon against copper.

[NWerner]

You should note however, that thermal EMF is temperature and strain dependent.
This vishay note neither mentions measuring conditions nor cites its sources.
I don't trust it

[TimFox]

They claimed (without details) that these were measured (presumably by their staff?).
As can be seen in the standard tables of thermocouple emfs, such as https://srdata.nist.gov/its90/main/ , the thermal emf for "normal" thermocouples is not strictly linear.  However, for metrology we are normally interested in the thermal emf values near room temperature.

[NWerner]

Oops .. while crossreading I skipped over this paragraph  . Sorry for the confusion.

[WattsThat]

I worked for the resistive systems group from 1976 to ‘81. Yes, some of that data was generated in house but it was not for publication at the time, they were trying to figure out how to use commercially available packaging methods (ceramic dips, TO-3 cans, etc) and keep the thermal emf as low as possible. AFAICR, most of the info came from existing commercial sources. What I find ironic is a common material they struggled with, Kovar, isn’t on the list.

They were keenly aware of the issues since a recent change (at that time) to the primary product, the S102 was improved by eliminating the flexible interconnects between the tinned copper leads and the foil based element. That design change remains in place with the S102C today where the interconnect leads are directly welded to the foil, eliminating one dis-similar metal junction.

As already mentioned, best to measure it yourself under the exact conditions of use.

PS. I assume this the Fluke document you’re referring to?

http://download.flukecal.com/pub/literature/p18-21.pdf

[TimFox]

Yes.  I believe that Fluke printed that in some of their user manuals as well.  Note that it does contain Copper-to-Kovar.
When I was working with precision analog electronics, I tried to get useful application information about reducing thermal emf problems with precision op amps and surface-mount PCBs, since most of the op amps had Kovar or similar leads, but the sales engineers for the semiconductor vendors did not have any useful information.  Since microvolt/kelvin offset drift is readily obtainable with IC op amps, I didn't want to mess it up with the connections.

[Kleinstein]

It sometimes only take minute impurities to change the thermal EMF. This is especially the with the semiconductors and also the near pure copper (high conductivity) alloys. So not all Cu-Be is the same.
The resistor alloys are usually less critical, as they have lot of electron scattering anyway.

With silicon the seebecke constant can go from some +600 µV/K to some -600 µV/K depending on the doping. So the question about thermal EMF from chips is very real. One point that helps here is that the thermal conductivity in silicon is high and the temperature gradients are thus often small.

[TimFox]

I noticed that Vishay quoted values for two of the standard Be-Cu alloys, with different Be fractions.   Their emf data are significantly different for the two alloys.
CA172 is (1.8 to 2.0)% Be, balance Cu.
CA175 is (0.5 to 0.7)% Be, (2.4 to 2.7)% Co, balance Cu.
There is a noticeable spread of the "minority" component in both alloys.

[razvan784]

Kovar is discussed in the Vishay appnote as giving high TEMF, even if it is not listed in the table.
Semiconductor manufacturers also discuss thermal EMF in their datasheets and appnotes for precision parts.
Take for example the LTC1050. https://www.analog.com/media/en/technical-documentation/data-sheets/1050fb.pdf There is ample discussion on page 6.
Now that's an old part (circa 1990); they've been copy-pasting that section into datasheets for newer parts, that don't come in hermetic packages with kovar leads, so the effect is reduced; the principles are the same though and resistors may still generate TEMF.
EDIT: the LTC2057 datasheet for example https://www.analog.com/media/en/technical-documentation/data-sheets/2057f.pdf gives an updated discussion starting at page 19.

[MegaVolt]

A little about solders.

[TimFox]

Thanks for the solder information.  Surprised that they omitted my favorite:  Sn62, Ag2, balance Pb
I was a little confused by the high values (at high temperatures) until I noted the discontinuity at the melting points for each of the different alloys.

[eplpwr]

I don't see CuTe in any table. It would be interesting since it's used in machined parts like low-EMF binding posts.

[Nauris]

Kovar is discussed in the Vishay appnote as giving high TEMF, even if it is not listed in the table.

It is actually in the table, just a little bit hiding under "Iron Alloys" Fe53-Ni29-Co17-Mn0.2 -35 µV/K.

[TimFox]

Thanks.  "Kovar" is a trademark.