Because of this, one of the most critical measurements - and the one most often taken inaccurately - is that of pipe temperature.

Difficult Measurement To Take

When the ambient temperature is higher than the pipe's temperature, a false measurement will result in a temperature reading higher than the actual pipe temperature. This, in turn, indicates a falsely higher superheat. Since most superheat charts call for accuracy of ±1 degree F, readings that are off by typically 5 degrees may cause the technician to overcharge the system, putting the compressor in danger.

No one is more aware of the problem than Adolfo Wurts. He is the senior research specialist at Fieldpiece Instruments, a Brea, Calif.-based company that manufactures measurement instruments for the HVACR market. One day his boss gave him the assignment to find a better way of measuring pipe temperature.

"He wanted heat transfer from the pipe to the sensor to be as easy as possible without exposure to ambient air. I'd seen so many temperature measurement products that I figured that I'd wrap up this project before lunch," said Wurts. "Two months later, after numerous experiments, consultation with HVAC teachers and other gurus, and several prototypes that didn't do the job, I was extremely frustrated."

Some Missteps

"The digital pocket thermometer is an inexpensive solution for taking air temperature measurements and can deliver good performance, provided the ambient conditions are stable and provided you can leave them in place long enough. But their inability to make good thermal contact with pipes makes them a poor choice for pipe temperature measurements.

"Yet one HVAC instructor I consulted insisted that I could take accurate tests with a digital thermometer by using a clamp such as those used with automobile jumper cables to hold it on the pipe," said Wurts. "It didn't make sense to me for two reasons. One, you can only make point contact with the pipe, leaving most of the probe exposed to ambient air. And, two, the clamp is metal and will conduct heat away from the thermometer to the environment."

But Wurts decided to experiment with the pocket thermometer. He put a bend in the thermometer to ensure line contact between the pipe and the sensing rod as opposed to point contact. Then he substituted a Velcro® strap for the clamp to hold it in place. Finally, he insulated the thermometer with neoprene to keep out ambient air.

"I was finally able to take reasonably accurate measurements," said Wurts, "but it took too long to set up and stabilize and the thermometer, once altered, was useless for other applications."

The ATC 1 pipe clamp.

Achieving Accuracy

"My first attempts at this gave me readings that I knew were off by 3 degrees to 5 degrees," he said. "I was surprised because this was such a common method for taking pipe temperatures. It makes sense, though, if you have a spherical thermocouple bead contacting a tubular pipe surface you end up with a point contact between the two. That slows heat transfer and leads to inaccuracy."

Wurts wondered if there was a way to accelerate heat transfer. He decided to strip some of the insulation near the bead and bring the wire in contact with the pipe.

"I took several Fieldpiece beaded thermocouples from stock and stripped back insulation from the sensor at different lengths," he said. "I wanted to see if having the wires near the bead come in contact with the pipe would improve the accuracy of the thermocouple or decrease the response time. It did both.

"After running experiments with the insulation stripped at different lengths, I noticed significant improvement when 3/8 inch of wire was exposed and brought into contact with the pipe. But after 3/8 inch, the improvement diminished with length."

Even though the temperature measurement was only occurring at the tip of the bead, the wire in contact with the pipe increased the rate of heat transfer to the bead.

To remove the influence of ambient air temperature, Wurts strapped neoprene insulation around the bead and wire.

"That gave us the accuracy and response we were looking for," he said. "We immediately began producing K-type thermocouples with 3/8 inch of insulation stripped back, and we amended the operator's manual to explain in detail how to get the best pipe temperature measurements using a beaded thermocouple."

The ATC 3 pipe clamp.

The Clamp Thermocouple

"We already manufactured a clamp thermocouple at Fieldpiece, but it had room for improvement in accuracy and response time," he said.

"I thought the lessons learned from the beaded thermocouple could be applied here."

The original design of the Fieldpiece clamp thermocouple consisted of thermally isolative material with a beaded thermocouple in its center.

This could only measure temperature to within 5 degrees of the pipe temperature, much too inaccurate for many of today's applications. Wurts wondered if he could change the shape of the contact point of the thermocouple by using a different material as the junction.

The problem with thermocouple material is that it is brittle and makes a poor spring. The shape of the sensor was much more important than the material from which it was made. The most important characteristic of the shape is that it must provide a line contact between the thermocouple and the pipe while also isolating the sensor from the clamp.

"I started experimenting by forming brass into different-shaped sensors," he said.

"I was getting closer to what I wanted, but the results weren't accurate enough. Finally, one shape gave me the results I needed. I used a thin piece of brass in the form of a curled cylinder with small metal flanges as standoffs. Not only did this provide adequate heat transfer, but the flanges were an effective standoff between the sensor and the clamp."

Unfortunately, while this approach gave Wurts the accuracy and response time he needed, it proved too difficult to manufacture. But it led him to another approach.

"I thought, why not use the thermocouple's own wires as standoffs. I drilled four small holes in the base of the clamp and ran wires up and back, knotting them at the ends so they wouldn't slip back - no welding or soldering. Over the wires, I placed a thin brass band on which the pipe could rest without damaging the wires. The junction in the thermocouple would only exist when the clamp jaws were placed around a pipe, causing the band to come in contact with the thermocouple wires and close the circuit."

This approach worked beautifully. The thermocouple wires created a small pocket of still air underneath the band, enough to keep the band at the temperature of the pipe. It was so effective at giving accurate measurements in a short time, Fieldpiece has applied for a patent. In its final production form, the wires were welded in place rather than knotted and a more resistant material than brass was used for the band.

For Fieldpiece, this has resulted in three new products. In the first two clamps, the ATC1 and ATC2, the same principles described above were implemented, with the thermal junction between the two wires being this thermally conductive band.

Because pipes often have condensation on their surfaces, these clamps feature a metal band that is highly conductive, very thin, and resistant to moisture.

The band allows for excellent thermal conductance between the sensor and pipe. The thin wires serve as standoffs and create a thin layer of still air while sealing out ambient air. To ensure maximum accuracy, the technician can also wrap the clamp and pipe in neoprene insulation.

"After we put our first two clamps into production, we kept up the research," said Wurts.

"The result was our newest clamp thermocouple that features a special spring to suspend the metal band. It has tiny holes drilled into it and leaves the sensor surrounded by completely still air, effectively sealing out ambient air. This clamp, the ATC3, is so effective, it provides high accuracy with an extremely fast response time - 20 to 30 seconds. We actually use this clamp as a standard to test the others."

Wurts offered a couple of hints on taking accurate measurements.

"No matter whose temperature measurement system you are using, it is critical that you calibrate it as soon as you take it out of the box," he said. "Our research shows the easiest and most effective method is to use a bucket of ice water. Stir the water and ice to ensure consistent temperature throughout the contents, then insert the sensor into the mix.

"After a short time, dial in 32 degrees on the calibration pot. When this is done, all of the inaccuracies associated with the thermocouple and the meter are eliminated and an accuracy of 1 degree is achieved.

"Also, when measuring pipe temperature on a horizontal pipe, place the sensor at a 45-degree angle or greater from the bottom of the pipe. This is because all systems contain some amount of oil in the refrigerant and it tends to rest at the bottom of a horizontal pipe, giving inaccurate measurements."

With the challenges presented by new systems and environmental standards, Wurts and his team appear to have found some elegant solutions.

Sine is a freelance writer specializing the HVACR market. He is based in Beacon, N.Y.

Publication date: 06/07/2004