Thermocouple signal amplification

Of course! I hadn't made the connection between the type B's shallow voltage reversal around room temperature, and its reference characteristics, but now that you state it it's obvious. Indeed, both Type B extension wires are just high-purity OFHC copper! Within rather wide limits, the reference junction's room temperature voltage is indistinguishable from zero. (I'm using B's right now :-).

...Yep, the defined range is 0 to -3uV to +3uV from 0C to 53C. Above about 260C the rate is 3uV or more per degree. In the 1200C+ range it's

I'm pretty sure none of the other standard TC's has a B-style inflection around room temp; do you know whether any of the nonstandard alloy pairs do?

jp

I don't know of any, but there are an infinite number of possibilities. If one existed, it might be very useful commercially.

Best regards, Spehro Pefhany

Got it. (The bookseller in Kent kindly sent it by airmail at the price of surface).

No info on type B, in fact they poo-poo rhodium content higher than

Type S was the original, however it was found that British type S T/Cs were significantly off in calibration compared to similar American sensors. This was found to be due to a slight iron contamination of the rhodium.

Type R alloy (more rhodium in the + leg) was created to allow instruments already in use to be used without re-calibration (though the curve is a bit different shape).

Best regards, Spehro Pefhany

Hello Speff,

Much the same historical information on types R and S is in NBS Monograph 125, published by Omega Press, Library of Congress number 73-600214. It is in the preambles to each of the thermocouple tables.

The NBS Monograph recommends the use of type B for better stability at higher temperatures (up to 1800C).

This post of yours is the only post seen in this thread, so I'm not sure what it is about. However, I have done interfaces for these precious-metal thermocouples, and they are right pigs....... damn-all output that needs a chopper amp, (but usually no cold-junction compensation), and a curve that would nicely fit a dog's hind leg. AFAIR though, the working range could be fitted quite conveniently by a single raising_to_a_power function.

Do you know if it has any historical info on type B?

Not important, I was just continuing a side branch of the original thread.

Best regards, Spehro Pefhany

In article , Tony Williams wrote: [snip]

Umm, no. Checking an old notebook shows that it was the Land Pyrometer that needed a raising_to_a_power function.

Type R and S thermocouples were reasonably linear in their intended working ranges, so could be handled with a simple offset and gain, (Ax+B), function.

Thank you very much, Tony. I'll make more inquiries from those pointers. Sounds like it was developed to increase stability at very high temperaturess.

Degussa and Heraeus in (postwar) Germany seem to pop up a lot.

A nasty (slightly) non-monotonic curve it is too. It's pretty much impractical to fit with a single polynomial over the whole range.

Best regards, Spehro Pefhany

I have NBS monograph 125 (with all the tables and polynomials) and STP470B, ASTM's thermocouple manual, if anybody needs data.

John

Useful looking quotes from page 47......

.....The 30-6 thermocouple was first introduced in Europe by Degussa of Hanau, Germany. Reference curves and table were published for their thermocouple by Obrowski and Prinz [1962]......

.....Studies by Ehringer [1954], by Walker, Ewing, and Miller [1962 and 1965], and by Glawe [1970] have demonstrated that thermocouples in which both legs are platinum-rhodium alloys are suitable for reliable temperature measurement at high temperatures.....

.... (1) improved stability, (2) increased mechanical strength, (3) higher possible operating temperature...

.... At temperatures below 450C the Seebeck coefficient of Type B becomes quite small and is almost negligible in the room temperature range.....

Thanks again! Of course all of this stuff is too old to be on the web.

There's a lot of good info in "Noble Metal Thermocouples", especially empirical data on contamination and hydrogen embrittlement, which are serious issues.

Just one polynomial or two? If just one, it's probably inaccurate as all get-out below 600 or 700 C. IIRC, around 250°C would be the peak error.

Best regards, Spehro Pefhany

Found some refs earlier on in the book Speff.

Obrowski.W and Prinz.W. Neu Bestimmte Grundwerte fur die Thermopaarkkombination Pt 30% Rh - Pt 6% Rh. Arch. Eisenhuttenwesen 33, 1-4 [1962].

Walker.B.E, Ewing.C.T, and Miller.R.R. Study of the Instability of Noble Metal Thermocouples in Vaccuum. Rev, Sci. Instr. 36, 601-666 [1965].

Glaw,C.E. Thermal Electromotive Force Change for Some Noble and Refractory-Metal Thermocouples at 1600K in Vaccuum, Air, and Argon. NASA Tech Note TN-D7027. [1970].

The same paper was also presented in Vol4, page 1645, (Instrument Society of America, Pittsburg, Penn.) [1972].

It looks nearly a straight line (plus offset) from 1000C to 1800C. The NBS Monograph 125 also has a power series expansion from 0 to 1820C... 8 terms, from T^0 to T^8.

NIST keeps up-to-date info, including polynomials and tables, at

John Perry