Thermal Divider

RST Engineering (jw) wrote:
I have a thermometer that reads from 0 to 200°F. I'm trying to measurea
cylinder head on an aircooled engine where my range of interest goes from,
say, 200 to 400°F. The classic way of measuring this temperature is with a
thermocouple mounted on a copper washer underneath the spark plug of the
cylinder that you determine (by trial and error) to be the hottest.

I can think of several ways of measuring a cooler spot on the cylinder that
will probably be in rough proportion to the actual temperature at the plug
seat, but most of them are dependent on the airflow over the cylinder(s)
remaining constant from day to day. With the baffling on the engine being
rather thin and wobbly, I can't count on this airflow being truly constant.

The sensor on my thermometer is a plain old silicon diode that won't
directly take the heat that I'm trying to measure. Anybody got a clever way
of making a thermal divider that won't be subject to the day to day shuffle
of the airflow over the cylinders?

Why don't you get a better sensor? Either a thermocouple or a platinum
resistance sensor will go up to 400F (204C). Farnell do a thin-film
Pt100 sensor that is only 5 mm by 2mm by 1.1mm.

Omega do similar looking sensor with leads attached (good o 500F)

http://www.omega.com/ppt/pptsc.asp?r...D-2&Nav=temc06

--
Bill Sloman, Nijmegen

scrobe on the papyrus:


RST Engineering (jw) wrote:
I have a thermometer that reads from 0 to 200°F. I'm trying to
measure a cylinder head on an aircooled engine where my range of
interest goes from, say, 200 to 400°F. The classic way of
measuring this temperature is with a thermocouple mounted on a
copper washer underneath the spark plug of the cylinder that you
determine (by trial and error) to be the hottest.

I can think of several ways of measuring a cooler spot on the
cylinder that will probably be in rough proportion to the actual
temperature at the plug seat, but most of them are dependent on the
airflow over the cylinder(s) remaining constant from day to day.
With the baffling on the engine being rather thin and wobbly, I
can't count on this airflow being truly constant.

The sensor on my thermometer is a plain old silicon diode that won't
directly take the heat that I'm trying to measure. Anybody got a
clever way of making a thermal divider that won't be subject to the
day to day shuffle of the airflow over the cylinders?

Why don't you get a better sensor? Either a thermocouple or a platinum
resistance sensor will go up to 400F (204C). Farnell do a thin-film
Pt100 sensor that is only 5 mm by 2mm by 1.1mm.

Omega do similar looking sensor with leads attached (good o 500F)

http://www.omega.com/ppt/pptsc.asp?r...D-2&Nav=temc06

A type K thermocouple will go much higher than that. Look at the
MAX6675, it will resolve to 0.25C from 0C (32F) to 1024C (1875F) with a
type K and it has an SPI bus.

--
John B

John B wrote:
scrobe on the papyrus:


RST Engineering (jw) wrote:
I have a thermometer that reads from 0 to 200°F. I'm trying to
measure a cylinder head on an aircooled engine where my range of
interest goes from, say, 200 to 400°F. The classic way of
measuring this temperature is with a thermocouple mounted on a
copper washer underneath the spark plug of the cylinder that you
determine (by trial and error) to be the hottest.

I can think of several ways of measuring a cooler spot on the
cylinder that will probably be in rough proportion to the actual
temperature at the plug seat, but most of them are dependent on the
airflow over the cylinder(s) remaining constant from day to day.
With the baffling on the engine being rather thin and wobbly, I
can't count on this airflow being truly constant.

The sensor on my thermometer is a plain old silicon diode that won't
directly take the heat that I'm trying to measure. Anybody got a
clever way of making a thermal divider that won't be subject to the
day to day shuffle of the airflow over the cylinders?

Why don't you get a better sensor? Either a thermocouple or a platinum
resistance sensor will go up to 400F (204C). Farnell do a thin-film
Pt100 sensor that is only 5 mm by 2mm by 1.1mm.

Omega do similar looking sensor with leads attached (good o 500F)

http://www.omega.com/ppt/pptsc.asp?r...D-2&Nav=temc06

A type K thermocouple will go much higher than that. Look at the
MAX6675, it will resolve to 0.25C from 0C (32F) to 1024C (1875F) with a
type K and it has an SPI bus.

But the OP only wants to go up to 400F, and thermocouples are pretty
nasty temperature sensors at the best of times - the output voltage is
low and you have to provide cold junction compensation. The Maxim part
may promise all kinds of nice performance, but it would still be
digitising the output voltage of thermocouple at the other end of at
least a foot of so of wire, if you could buy the thing in the first
place - Farnell doesn't stock the MAX6675, which means that it isn't
exactly a commodity part.

--
Bill Sloman, Nijmegen