High Accuracy Temperature measurement


Thread Starter

Les Bond

I need to measure a temperature in an industrial process environment with an accuracy of +- 0.1 degrees C over a range of 80 to 120 degrees. I want to bring this signal into a Delta V system either on HART or 4-20 mA. Rosemount offer a transmitter 3144 that will give +- 0.24, as do Yokogawa and Honeywell offer something similar. Any ideas control.com?

Matthew Hyatt

Are you using TC or RTD? It should not be a problem to calibrate the sensor and transmitter over this range to get a +/- 0.1 Deg C accuracy.

Do a little searching, various manufactures out there provide calibration equipment nad all transmitters can be scaled and calibrated to meet your needs.

What you are trying to do is difficult to achieve but not impossible. You are either going to have to consult an expert in the field or do a fair bit of homework.... probably both.

A PRTD is probably the right element but there is much more to it than that. RTD's need to be protected from stress and shocks so even the element design, construction and quality are important.

Application design is going to be critical in order to get that accuracy. The element needs to be very closely coupled to the process temperature to ensure that the difference between your process temperature and the element temperature is much less than your specified accuracy requirements. The errors here can be very large.

You will need to consider how you will calibrate your system. Unless there is a suitable calibration facility available to you then you will probably need a fluid bath and certified reference thermometer. You may need and certainly should consider some means of checking your certified thermometer. I can calibrate to an uncertaintly of 0.017deg C but the cost of the equipment is considerable. I have a certified thermometer, fluid bath and accessories, a water triple point cell and a Gallium melt cell supplied by Isotech. Hart Scientific also make similar equipment. At the cell temperatures (0.01 and 29.7646) I can check to an uncertainty of 0.0005 deg C)

When you are comparing precision/accuracy of transmitters you need to add up all the errors. The affect of ambient temperature on the transmitter is usually the greatest source of transmitter error. Try to locate the transmitter in a temperature controlled environment. You may need to use a measurement system of the standard used as reference
thermometers. Some names... Cropico, Hart Scientific, Isotech. Alternatively you might consider a good quality multimeter such as Agilant or Keithly.

Staying with industrial transmitters you might like to look at the Accutech range. In particular look at the AI-1500 which is the best they do with HART communications. Don't even consider analog signals with your specification and range. I have just calibrated an AI-1500 with a 1/10 DIN RTD from a different supplier. I did a 3 point calibration up to 60deg C. the largest difference between the reference system and the AI-1500 was 0.008deg C. It may be worth getting some advice from them if you want to use an industrial transmitter.

Some of the web sites above have some excellent homework material and some of these companies will provide reference material. I'm sure they will all provide good technical advice.

Good luck

Vince Dooley
An RTD is the most accurate but I doubt that you will be able to get an RTD/transmitter combination that will give that accuracy.
How about using PT-1000 ohms instead of PT100? 1000 ohms response more accuracy than 100 and you can use transmitter.

Best regards
JH Kim
I would try using the Minco TT176 sensor. The spec are as follows:

Input: 10 ohm copper or 100 ohm platinum RTD's.
Output: 4 to 20 mA DC over specified range.
TT176: ± 0.1% of span.
TT176: ± 0.1% of span.
Ambient Temperature:
Operating: -40 to 85 °C (-40 to 185 °F).
Storage: -55 to 100 °C (-67 to 212 °F).
Ambient Temperature Effects:
TT176/TT676: ± 0.009% of span/°C.
± 0.018% of span/°C for spans less than 10 ohms.

You may find other sensors that are more appropiate for your application at http://www.minco.com. They also have matched sensors if you need them. We have used their sensors and have had good success especially in one-of-a-kind applications.
Moore Industries' THZ Smart HART Temperature Transmitter may be able to handle your application. High accuracy measurements are achieved using a 1000 ohm platinum RTD calibrated using a sensor-matching calibration bath. If you read the digital HART signal at your control system instead of 4-20mA, you can increase your measurement accuracy even further by eliminating the analog output error.

You can view the data sheet for the THZ at: http://www.miinet.com/products/sg_temperature.shtml

Steve Todd


I assume this to be RTD input for low range
recommend following make.

1] Acromag 800 series transmitter

2] Moore industries

3] Lufft - Germany 0.05 deg c accuracy.

[email protected]

Greg Brunette

On February 27, 2004, Les Bond wrote:
>Rosemount offer a transmitter 3144 that will give +- 0.24 <

Per Rosemount's 2004 catalog, Temperature section page 7, 3144P- "Reference Accuracy Example:
When using a Pt 100 (á = 0.00385) sensor input with a 0 to 100 °C
span: Digital Accuracy would be ±0.10 °C"
You can bring this HART digital signal into your DeltaV and have your required accuracy. If the highest accuracy is required you should also look into "Sensor Matching" which addresses the sensor interchangability error you will get with all RTD-transmitter systems.
Full disclosure- I sell Rosemount products. All statements above are from Rosemount's published data.

Good luck,
Greg Brunette
Write your reply...I will advice you should use RTD than TC due to the temperature range is till small. but it now depend on the area you want too use it.