Hoppa till innehåll
Free shipping on orders over 3000 SEK excl. VAT! Volume discount: 5 pcs 2% · 10 pcs 3% · 25 pcs 5%
Volume discount: 5 pcs 2% · 10 pcs 3% · 25 pcs 5%The air in the fridge doesn’t show the temperature of the vaccineThe Air in the Fridge Lies: How to Monitor the Real Temperature of Your Goods Without False AlarmsHow to show in seconds whether the load stayed at the right temperature and humidity during transportThe air feels heavy long before anyone measures it: how to make carbon dioxide levels visibleThe freezer showed -18 °C on Monday morning, but what happened over the weekend?The core temperature is what counts, and it only shows if someone has logged itTwo measuring points and the difference between them: how to make the temperature differential in a process visibleFour Measuring Points and a Time-Stamped Log Show Where the Heat in the Process Actually GoesYou Can’t Guess the Core Temperature of a Product: How to Document It Without Standing Next to ItThe fridge looked like it was holding the right temperature, but what happened overnight?Volume discount: 5 pcs 2% · 10 pcs 3% · 25 pcs 5%The air in the fridge doesn’t show the temperature of the vaccineThe Air in the Fridge Lies: How to Monitor the Real Temperature of Your Goods Without False AlarmsHow to show in seconds whether the load stayed at the right temperature and humidity during transportThe air feels heavy long before anyone measures it: how to make carbon dioxide levels visibleThe freezer showed -18 °C on Monday morning, but what happened over the weekend?The core temperature is what counts, and it only shows if someone has logged itTwo measuring points and the difference between them: how to make the temperature differential in a process visibleFour Measuring Points and a Time-Stamped Log Show Where the Heat in the Process Actually GoesYou Can’t Guess the Core Temperature of a Product: How to Document It Without Standing Next to ItThe fridge looked like it was holding the right temperature, but what happened overnight?

Galvanically Isolated Thermocouple Inputs — Four Sensors on the Same Machine Don’t Have to Share a Potential

Published 22 August 2026 at 02:19

A thermocouple measures a voltage of a few tens of microvolts per degree. That is a very small signal — and when the measuring tip is also welded or clamped onto metal that is part of a machine, it becomes sensitive to something that rarely appears in the measurement protocol: the electrical potential of that point.

If you attach four sensors at four different locations in a facility — a tool, a motor housing, a pipe, a furnace wall — those points won’t necessarily be at the same potential. If they’re connected to the same measuring circuit without isolation, a current path forms between them through the instrument. The result is channels that interfere with each other, readings that drift without the temperature changing, and in the worst case, damaged electronics. The fault looks like a measurement problem but is actually an electrical problem.

TandD MCR-4TC four-channel thermocouple logger with touchscreen, SD card and galvanically isolated inputs

What it is

TandD MCR-4TC is an industrial four-channel thermocouple logger with galvanically isolated inputs, a built-in touchscreen and SD card storage. It supports type K, J, T, S and R and covers −199 to +1760 °C depending on the sensor.

  • Accuracy: K/J/T ±(0.5 + 0.003 × t) °C at ≥ −100 °C; S/R ±(1.5 + 0.003 × t) °C at ≥ 100 °C
  • Cold junction compensation ±0.5 °C at 10–40 °C ambient temperature
  • Resolution 0.1 °C
  • Memory: up to 480,000 readings on a single channel, 120,000 with four channels running — and automatic export to SD card when internal memory fills up, so the series is never broken
  • Modular: up to four units can be linked together into 16 synchronized channels with a shared start and shared settings

Thermocouples are not included — they are chosen based on the application. A sister model, MCR-4V, measures voltage ±50 V instead.

The problem it solves

Isolation is what makes multi-channel measurement honest. When each input is galvanically separated from the others, there is no current path between the measuring points, and each channel measures its own thermal voltage without being affected by where the other sensors are located. That is the precondition for measuring on a machine in operation rather than on a test piece in a lab.

The accuracy formula is worth a closer look, since it tells you something about what to expect. The error grows with temperature: at 100 °C the term 0.003 × t is only 0.3 °C, but at 1000 °C it is 3 °C. That is physically reasonable — and it means you shouldn’t expect tenth-of-a-degree precision at annealing temperatures just because the display shows tenths.

Cold junction compensation follows the same logic. A thermocouple measures a difference between the measuring tip and the connection point, so the instrument needs to know how warm it is at its own terminals. Here it’s specified as ±0.5 °C over 10–40 °C — which also explains why a logger placed in the sun or next to a furnace gives worse readings than one kept cool, no matter how good the sensors are.

The memory and SD export ultimately make the instrument suited for what it’s built for: long processes. A hardening cycle, a drying process or a commissioning test can run for days, and what determines whether the measurement is usable is that the series remains unbroken from start to finish.

Three typical use cases

  • Heat treatment and furnaces: multiple measuring points in the load and chamber simultaneously, with type S or R for the highest temperatures.
  • Process monitoring and commissioning: mapping the temperature distribution in a machine under real operating conditions.
  • Troubleshooting: comparison between points — where the difference between channels is the answer, and must therefore be free of crosstalk.

Why it pays off

The alternative to multiple synchronized channels is measuring one point at a time and trying to piece the sequences together afterward. That only works if the process is exactly repeatable — and if it is, you rarely need to measure it at all. The interesting part almost always lies in the difference between points at the same moment in time: where is the load coldest, how far behind is the core, when does it level out. That question can only be answered with channels running simultaneously and sharing a time axis.

And isolation is what makes the answer trustworthy. A measurement series that looks neat but contains crosstalk between channels is worse than no measurement at all, because it looks like data.

Read more about the TandD MCR-4TC four-channel thermocouple logger →

More news

All news