The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
From: Tony Williams
Subject: Re: Reference Junction for Thermocouple
Date: Tue, 29 Oct 2002 15:12:52 +0000 (GMT)
NNTP-Posting-Date: Tue, 29 Oct 2002 15:13:56 +0000 (UTC)
User-Agent: Pluto/1.14i (RISC-OS/3.60)
In article ,
> I would like to build a solid state reference junction for my
> thermocouples. I know it's been done many times, so I don't want to
> "reinvent the wheel." Anyonel have a schematic handy?
I don't know if this is useful. It's a copy of
a post made here some months ago.
A copper resistance is the cheapest/easiest.
especially when used with the 7106/7107 DVM
chip (because the differential inputs nicely
interface directly to a resistance bridge).
The 7106 (LCD) is preferable because of the
much reduced self-heating (no LED driving).
T/C+ +----------------------->INPUT HI
| | 7106
T/C- +------+ +------>INPUT LO
Rcu is clamped across the point where the
T/C wires change to copper/copper wiring.
A Type K thermocouple (say) is about 40uV/C
at 20C. So it's output reduces by 40uV for
every 1C increase in ambient temperature.
The T/C is stood on a pedestal formed by Radj+R1
and Rcu. Rcu has a temperature coefficient of
about +3930 ppm/C. If Rcu has a value of (say)
about 10 ohms, then 1.017mA through it will
result in a dV/dt of +40uV/C, as req'd to
compensate for the T/C dV/dT of -40uV/C.
Because Rcu measures absolute temperature it
has an offset equivalent to 273.15 at 0C. So
R2 and R3 provide the balancing offset, where
R3/(R2+R3) = Rcu/(Rcu+R1+Radj) when Rcu = 0C.
Rcu is easy to obtain.... it can be formed
as a pattern on a pcb, or it can be wound with
fine wire, etc, but the cheapest source is
small rf chokes (the size of a 1/4W resistor.
You can see them on sale in most broadline
distributor catalogues, and they cost peanuts.
Choose one of about 10 ohm to 20 ohm nominal
resistance. The mfr'ing tolerance is going
to be about 20%, but it doesn't matter if
you follow the design steps below.
1. From the given nominal value of Rcu, at
whatever temperature they quoted it at
(usually 20C), use the +3930 ppm/c to calc
dR/dT and consequently the current needed
to exactly compensate the T/C.
2. That current (and Vref) give a value for
(Radj+R1). Proportion Radj to get a 20%
adjustment range on that current.
3. Using the +3930 ppm/C again, calculate the
the value of Rcu at 0C, and the voltage
across it at 0C.
4. Calculate R2/R3 so that the voltage across
R3 is equal to the voltage across Rcu at 0C.
Actually, R3= Rcu(0) and R2= (Radj+R1), as
calc'd in (3) above, or as near as practical.
On the final thermometer, stick the business end
of the T/C into an ice-bath and adjust Radj for
a 0C reading.
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup