IN Brief:
- Pickering tested its 3.9mm Series 120 against another 4mm-class reed relay under controlled thermal conditions.
- Final thermal EMF magnitude measured 46.4µV for Series 120 against 402.6µV for the comparison device.
- The measurement is relevant to semiconductor test, ATE, precision acquisition, and other low-level switching systems.
Pickering Electronics has published controlled thermal electromotive-force measurements for its Series 120 reed relay, recording a final offset magnitude of 46.4µV after 600 seconds compared with 402.6µV from another 4mm-class relay tested under the same conditions.
The difference represents approximately 88.5% lower final thermal EMF magnitude in Pickering’s experiment, but the measurement method is at least as important as the percentage. Thermal EMF is affected by conductor materials, temperature gradients, coil heating, PCB construction, and the wider measurement environment, so a bare comparison figure is of limited value without the test conditions behind it.
Pickering used a Series 120-1-A-5/2 device in a 3.9mm package intended for 4mm-pitch layouts. Both relays were wired with low-thermal-EMF copper leads and placed in a thermally insulated, draft-free enclosure, with voltage across the contacts measured using a Keysight 34465A 6.5-digit digital multimeter.
The DMM operated with a 10 NPLC integration time. Once the test arrangement had reached thermal equilibrium, each relay coil was energised at nominal voltage and one contact-voltage measurement was logged every second for ten minutes.
Pickering carried out two runs on its own relay and three runs on the comparison device, publishing averaged values from the repeated tests. The difference developed quickly after coil energisation rather than appearing only at the end of the test.
After ten seconds, average offset had reached -9.6µV for Series 120 and -92.0µV for the comparison device. At 60 seconds, the readings were -34.0µV and -338.6µV respectively, before reaching -46.1µV and -402.5µV after five minutes.
The final 600-second measurements were -46.4µV for Series 120 and -402.6µV for the comparison relay. Pickering also reports that variation between final runs was about 0.5µV for Series 120 and 3.1µV for the comparison device.
Thermal EMF becomes relevant because a relay inside a precision measurement system is not electrically invisible. Junctions involving different conductive materials can generate small thermoelectric voltages when temperature gradients develop, while heat from an energised relay coil can change those gradients over time.
Hundreds of microvolts may be immaterial when a switching system handles signals measured in whole volts. The same error can be significant in semiconductor characterisation, thermocouple routing, sensor acquisition, calibration equipment, or automated test systems where the signal under measurement may itself be only millivolts or microvolts.
The Series 120 is designed around high channel density, with its 3.9mm body allowing relays to be stacked on a 4mm × 4mm board pitch. That compact format is useful in multiplexers and switching matrices, where the number of channels that can be placed within a given PCB area can determine the size and cost of the complete test system.
Density can create its own thermal complication. Placing multiple energised coils close together can raise local temperature and increase gradients across the board, which makes relay construction, coil resistance, PCB copper, ventilation, and channel duty cycle part of the measurement design rather than purely mechanical considerations.
Pickering attributes the Series 120 result to high-resistance coil design, material selection, and mechanical construction intended to limit thermal effects. The test does not establish that every Series 120 implementation will reproduce the same offset, because board design and environmental conditions remain significant variables.
The unnamed comparison device also limits how broadly the figures can be applied. Pickering explicitly states that the results relate to this particular controlled comparison and that thermal EMF can change with PCB layout, materials, temperature gradients, and test environment.
The published methodology nevertheless gives engineers enough information to judge whether the conditions resemble their own equipment and to reproduce a comparable test if the measurement budget justifies it. In dense low-level switching systems, footprint and contact ratings remain important, but the voltage generated by the switching element itself can be just as relevant once the required resolution moves into the microvolt range.
