Am 22.10.2011 19:49, schrieb Jed Rothwell:
Peter Heckert <[email protected] <mailto:[email protected]>>
wrote:
This discussion about "close contact to the metal" and
"chemogalvanic or electroosmotic voltages" is blather. I am sorry
to be harsh, but it is irrelevant, evasive, nitpicking blather.
It is not.
Put 2 identical copper electrodes in water. Heat one, and the
other not. You get a voltage and a current between the electrodes.
You misunderstand. It is blather because it is not important. Even if
you right, it does not affect the conclusion, and it does reduce
confidence in the results. As I said, if you had cut the thermocouple
wires with scissors, or Mats Lewan had accidentally dropped his
temperature log papers off a bridge -- if we did not have a single
numeric measurement of temperature -- we would still know with
absolute certainty that massive anomalous heat was produced, far
beyond the limits of chemistry.
We know there was energy and we know there was a heater.
Nobody denies there was energy.
You cannot make a conclusion from heat to anamolus energy. This is junk
size.
You are discussing thermocouples instead of confronting this fact. You
are nitpicking.
Now let me nitpick a little. I suspect you are wrong Omega sells these
bare wire K-Type themocouples, and they do not say in their catalog or
on-line guides that things do not work when you tape them to a pipe.
People do that all the time. I have done that. I have not seen
unexplained fluctuations.
Having not seen something means nothing. Quite often it means that you
closed your eyes.
It is absolutely correct to connect a thermocouple directly to a metal
pipe, but only if the measuring amplifier is not electrically connected
to the pipe otherwise. Normally this is the case.
If you have a multichannel thermoamplifier and if the inputs are not
isolated, then you cannot connect two thermoelements to one and the same
pipe, because the thermovoltage will be partially shorted.
This is VERY obvious.
And if the pipe is unequally heated it can itself generate a
thermovoltage and this can add to the measurement. (Yes, even when you
have a temperature gradient in homogenous metal, you get voltages)
Of course there might be high end devices that are internally isolated
but these must be much more expensive.
Refusing to think about this very reasonable consideration and close the
eyes is junk science.
The TCs remained in reasonable agreement with dial thermometers and
other non-electronic devices.
Possibly they had no firm contact. To understand this, you must think in
a microohms and microvolts.
If no close contact is there then the resistance is in range of infinite
to milliohms and because these thermocouples have low impedance this
might not influence the measurement. But if you have bad electrical
contact you also have bad thermal contact and you measure the
airtemperature via heat conduction over the wires as an additional error
factor
Of course it is simpler to close the eyes and the brain and to go on as
usual.
But this is junk science. If this works sometimes or works mostly for
you, this does not mean it will work always under all conditions.
It is common in science and technics, some people repeat the same error
over and over because they refuse to think and instead judge from
experience and belief. They think if it worked 3 times for then it will
work 1000000 times for others.
But this is junk science,
Peter
