[SS2S-Main] Re: temperature measurement

  • From: Richard Nakka <richard.rocketry@xxxxxxxxx>
  • To: sugarshot@xxxxxxxxxxxxx
  • Date: Mon, 24 Mar 2014 11:18:00 -0500

Thanks for the valuable information. We are looking at using the
DI-149, for budgetary reasons.


On Mon, Mar 24, 2014 at 11:13 AM, Grant Saviers <grants2@xxxxxxxxxxx> wrote:
> One additional comment - the output filter shown for the AD8495 for 60Hz
> rejection requires a unity gain op amp buffer to drive the input of the
> DATAQ, assuming its input impedance is something less than many megohms.
> Which DATAQ is being used?
> Grant
> On 3/24/2014 8:19 AM, Grant Saviers wrote:
>> A lot depends on which DATAQ unit as there are specific models for
>> thermocouple measurements.
>> However, my guess is you have a one with a single ended high (0 to 5v?)
>> inputs.  So the few millivolt output of the thermocouple will have
>> significant errors.  Generally, an IN-amp (instrumentation amplifier) with a
>> gain of 100 or more is used to amplify the signal and eliminate other error
>> sources such as high common mode signals. The general properties of in-amps
>> are high gain, low bandwidth, high common mode rejection, low input offset
>> voltages, and very low offset drift.  A dozen components plus the IC can
>> make a decent preamp for each channel.  The preamp design needs to fit the
>> input specs of the DATAQ - single ended or differential inputs.
>> A basic tutorial and simple preamp optimized for K thermocouples you can
>> build easily is on page 3 of
>> http://www.analog.com/library/analogDialogue/archives/44-10/thermocouple.pdf
>> I use Expresspcb.com for proto boards and the 3 boards for $51 option will
>> easily get you 8 channels of preamp.  The AD8495 is in stock for $4.81 at
>> digikey.com
>> A fairly deep dive on this topic is at
>> http://www.analog.com/library/analogdialogue/archives/43-09/EDCh%202%20other%20linear%20circuits.pdf
>> Having a reference junction is nice for high accuracy, but probably not
>> needed given what you will measure.  The tables and formulas for linearity
>> correction are dependent on the thermocouple type and for a K type in this
>> application not needed.
>> Grant Saviers
>> On 3/24/2014 7:32 AM, Richard Nakka wrote:
>>> For an upcoming motor test, we would like to use thermocouples
>>> (k-type?) to measure motor casing and nozzle temperatures.
>>> We would like to use an 8-channel DATAQ unit for the data acquisition.
>>> We understand there are challenges to this due to small voltage
>>> signal, non-linear temperature-voltage relationship and the need (?)
>>> for a reference junction compensation.
>>> Question to list members: Can anyone give advice on how to do this?
>>> What kind of amp would we need? Could anyone be of assistance in
>>> designing/building such?
>>> thanks in advance
>>> Richard

Other related posts: