Temperature controller for non-standard value RTD

Steve wrote:


Please check if a "calibration value" solve Your problem.

--
Pozdrowienia!!StaCH
Sieci,systemy,serwis PC,alarmy,kamery,domofony,sterowniki PLC.
http://www.uniprojekt.com   mobile:wap.uniprojekt.com

A temperature controller with the possibility of an "input shift" will solve
your problem.
Look at the omron controllers type E5CK or E5CN. They have that possibility.
I used the "input shift" to correct the temperature reading of a Pt100 in an
explosive surrounding that was connected to the controller by zener
barriers.

Ronny

One of my non-standard (non 100 ohm)  value RTD devices has a
resistance of 250 ohms at room temperature, another is 750 ohms.  I
tried using an Omega CN9220A controller which has offset compensation,
but it would only compensate -128 degrees which was not enough.  It
actually gave a sensor error because of the too high 750 ohm value.
Using a potentiometer in place of a sensor, it gave a sensor error at
300 ohm.  I will look into the other suggested controller.  Another
option for me is to use the controller's voltage input instead, after
converting the rtd current to a voltage since my rtd resistance vs
temperature is linear.

Steve

proclaimed to the world:


Almost all of the microprocessor based temperature controllers I have
worked with have the ability to "zero" adjust. Some controllers may
have trouble making this large of an adjustment. Someone already
suggested Omeron. Honeywell also has this adjustment. If you have a
decade resistance box, it will make the job easier.

Another possibility is to add a precision temperature stable resistor
across the RTD input to bring total resistance down to 100 ohms. This
will affect the slope of the resistance/temp relationship slightly but
the calibration adjustments will be able to correct for this. Again, a
decade resistance box makes the job of determining the "shunt"
resistor you add easier. I have even used a potentiometer in place of
a fixed shunt resistor.

If you need more detailed explanation on doing this, just ask.

Paul M wrote:

It is not a matter of zero offset, but scale. A platinum RTD is usually
excited by a fixed current and the voltage across it is measured in a
three- or four-lead configuration. Changing the nominal resistance from
100 to 160 requires either a reduction of meter gain to 100/160 (.625)
of the original, or a reduction of excitation current by the same ratio.

Jerry
--
        "The rights of the best of men are secured only as the
        rights of the vilest and most abhorrent are protected."
            - Chief Justice Charles Evans Hughes, 1927
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

proclaimed to the world:


Jerry, have you ever tried doing any of this? I don't believe that I
am wrong here. Perhaps I am not describing this well. The RTD should
be 100 ohms at 0 deg C. (in most cases). I understood him to say his
was reading 160 at 0 deg C. He needs to lower the resistance to match
a standard 100 platinum RTD or change the zero of the measuring device
to 160 ohms. The meter does indeed supply a constant current source
and then measure the voltage across the RTD. In a three or four wire
system you are just measuring the drop at the RTD, subtracting the
drops across the leads. Lowering the current source value will correct
the offset. This is how the meter makes zero offset adjustments (in
most cases).

Keep in mind that I have been involved in building and designing these
meters from the days of using mechanical analog meters. It's not my
understanding of how these things work that I question, but perhaps
how I am explaining it.

I can't see how scale is the issue here but then my definition of
scale may differ. I define scale as difference between the minimum and
maximum reading of the meter. Perhaps I need to look up the
definition. I learned it in the lab, scaling meters for customers.
Sometimes the meanings of words used in the industry deviate from the
definition given in school, and this is why I asked if you had ever
done this. What do you base your statement on?

You did remind me of something I did not specify. The shunt resistor I
suggested should be connected across the two sense leads at the meter
and not the "comp" or excitation lead or leads.

Paul M wrote:

   ...


To answer your first question, I have done it, but only with equipment I
designed myself. I now believe that we have essentially the same things
in mind, but coming from different backgrounds, we use use different
terms. To me "offset" is the addition of a constant to all readings;
"scale" or "gain" it the slope of the readout vs. the change in measured
quantity. (Think of your old trusty VOM. Offset is adjusted by moving
the pointer spring, and the range knob changes scale.)


As I wrote above, we use different terms to mean the same thing. That
wouldn't be a problem if those words didn't have other meanings in out
individual lexicons. When you say "reset", you probably mean what I call
"integration". When I say it, I mean clear to zero or to an initial
condition.


Doesn't the temperature coefficient of the shunt resistor need to be
taken into account? How would a shunt work in a three-wire setup?

Jerry
--
        "The rights of the best of men are secured only as the
        rights of the vilest and most abhorrent are protected."
            - Chief Justice Charles Evans Hughes, 1927
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

proclaimed to the world:


Yes, part of this is terminology, but mainly it is because I was wrong
in what I wrote. Looking back, I realize that memory is failing me.
Putting a fixed resistor in parallel will not only change the zero
offset but also the slope or scale. That is not what I had done. I
believe it was this: Put a fixed resistor or resistors in series with
the lead compensation lead(s). This tricks the meter into thinking it
has high lead resistance. In effect, the meter will increase the
voltage applied to the RTD, shifting the zero point.

It really is moot. I don't think the meter circuit will be able to
handle that much of a change. He had 160 ohms and wanted to shift that
down to 100. That is a big change.

I've started taking some medicine for pain that is really messing with
my head right now, so I will use this as an excuse and ask everyone to
humor the doped up idiot until I am clear headed.

Terminology and the way we learned our craft does present a problem
sometimes. I started in the military and was taught PID as being PRR
or proportional, rate, reset. This was in the context of pneumatic
control systems. Have you ever messed with a fully pneumatic PID
control system? Hagen and Baily were the two systems used. Both used
volume chambers, fulcrum bars and needle valves. The system controlled
the main boilers on an aircraft carrier and had to be tuned for
quickly changing loads. The steam catapults could draw so heavily on
the system that you would go from 10% to 90% in about 15 seconds. The
limit was moisture carryover, where the steam rushed out so quickly
that water got pulled with it, essentially the mother of all water
hammer. Anyway, the controls had to be constantly tuned and kept
clean. It was a nightmare. I am aware of the terminology problems and
try to use more standard words. Still, the way we approach problems is
affected by our experience. I see this in Tim a lot. I find it
interesting more than irritating and a good topic of discussion once
in a while.

I hope everything is going well for you Jerry.