Omron Relay Roundup – Sizing by Real Watts, Not Contact Labels

You’re staring at a contact rating: 10 A at 250 VAC. That’s 2500 VA — plenty of headroom, right? Then why do relays fail after 20,000 cycles under a 6 A resistive load? Because the VA label hides the real constraint: watts dissipated inside the contact, the coil, and the package. This roundup looks at three Omron families — G2R, MY, G7J — and sizes them by the forces that kill relays: thermal watts at the contact, coil power, and dielectric margin under temperature. No guesswork, only numbers that change how you select.

1. Contact Watts vs. Label Amps – the AgCdO limit

All three families use AgCdO contacts except the G7J‑4A which uses AgSnO₂. That material choice directly governs resistive switching life. AgCdO has a maximum operating temperature of about 130 °C at the contact point; beyond that material transfer accelerates. The G2R‑1 is rated 10 A / 250 VAC — 2500 VA apparent. But real watts are I²R. Contact resistance for a new G2R contact is roughly 30 mΩ (illustrative). At 10 A, I²R = 10² × 0.03 = 3 W dissipated right at the junction. That 3 W is the mechanical limit. In a 70 °C ambient (G2R max), the junction rises another ~35 °C — already near the 130 °C ceiling. Push to 12 A and you’re at 4.3 W, junction over 145 °C, and contact life drops from 100k cycles to ~15k cycles (about 85% reduction). The worked consequence: if your load is 8 A continuous in a 60 °C cabinet, the G2R‑1 still has ~40 % derating margin; but in a 70 °C cabinet you need to drop to 7 A. Reversal: for intermittent duty (if the coil voltage is correct (more on that next).

2. Coil Power – the hidden heater that steals contact margin

A relay’s coil is a resistor. The G2R‑1 with 12 VDC coil draws about 90 mA (approx 1.08 W). That wattage heats the inside of the relay. In a sealed PCB package, that 1 W raises internal temperature by about 8–10 °C above ambient (illustrative per package thermal resistance ~8 °C/W). Now the contact junction is 8 C hotter — meaning the 3 W at 10 A pushes you to the 130 °C limit sooner. The MY2 with 12 VDC coil draws about 70 mA (~0.84 W); less internal heat, so the 5 A rating is more stable. But the MY2 also has only 5 A contacts — the ratio matters. Worked decision: for a 24 VDC system, the G2R‑1 coil (24 VDC) draws about 45 mA / 1.08 W (same power); the MY2‑24VDC draws ~30 mA / 0.72 W. In a panel with eight relays on a row, the G2R group adds 8.6 W of coil heat; the MY group adds 5.8 W. That 3 W difference shifts the internal cabinet temperature by about 2 °C — enough to require derating the contact by ~0.5 A. When this flips: if you use the G7J‑4A (40 A rating, coil 24 VDC at 2.5 W) for a 10 A load, the coil heat is 2.5 W but the contact dissipation is only ~1.2 W at 10 A — total 3.7 W, well within the package capability (−40 to 85 °C ambient). The G7J is overkill for 10 A, but the coil heat penalty is negligible relative to its thermal mass.

3. Dielectric Strength & Temperature Range – when real watts test the insulation

G2R and MY series both spec 1500 VAC dielectric; the G7J series 2500 VAC. Higher voltage switching (e.g., 480 VAC line) requires bigger air gaps. But the real‑watts issue here is partial discharge under heat. At 70 °C ambient, the dielectric strength of air decreases by about 4 % per 10 °C (illustrative). If your contact is dissipating 3 W and raising internal air temperature to 90 °C, the effective dielectric margin drops from 1500 VAC to about 1350 VAC. For a 277 VAC line (phase‑to‑ground) that’s still fine — 5× margin. But for 600 VAC switching (common in industrial controls), the margin shrinks to 2.3×, and any inrush transient can flash over. Worked consequence: in a 600 VAC application with a resistive heater load of 8 A, the G7J (2500 VAC dielectric, 85 °C ambient) gives 4× margin even at 85 °C; the G2R gives only 2.5× and is not recommended above 250 VAC. Reversal: for signal‑level switching (

4. Mounting & Package – how PCB vs. socket changes thermal path

G2R‑1 is PCB‑mount only; MY2 comes in PCB or socket versions; G7J‑4A is panel‑mount. The socket adds a thermal break — the relay sits in air, and heat conducts through the socket pins (small cross‑section). A G2R soldered into a PCB can sink ~0.5 W through the pins into the board (illustrative). At 4.1 W total internal heat, the PCB removes about 12 % of the heat; the rest must go by convection. In a sealed enclosure with no airflow, that 4.1 W raises the inside of the relay to ~85 °C (ambient 70 + 15 °C rise). That’s above the AgCdO limit. Worked rule: if you use a G2R‑1 in a sealed panel and the load is >8 A continuous, you must either add a heatsink pad under the relay or derate to 7 A. The MY in a socket has a similar issue — the socket traps heat, but the total internal heat is only 1.5 W, so the rise is only ~5 °C. Reversal: in a ventilated panel with forced air, the G2R can run at full 10 A because convective heat transfer doubles. The G7J‑4A, panel‑mount, has a metal base that can be bolted to a chassis — its thermal path is vastly better. At 7.3 W total, the chassis can sink 4 W, leaving only 3.3 W to the air. That’s why the G7J survives 85 °C ambient at 40 A.

5. The Rule: size by total watts per relay, then add ambient

Here’s the only threshold you need. For any Omron relay in this roundup:

• Total internal heat = I²R_contact + V_coil × I_coil (coil power from datasheet).
• If total heat ≤ 2 W, any mounting works at up to 70 °C (MY, G2R at ≤5 A).
• If total heat > 2 W but ≤5 W (G2R at 10 A), require ventilated panel or forced air if ambient > 50 °C.
• If total heat >5 W (G7J at 40 A), panel‑mount with chassis conduction mandatory; do not use socket.

That rule, derived from the thermal dynamics above, replaces any “10 A relay will handle 10 A” assumption. The G7J‑4A at 40 A dissipates 4.8 W in the contact alone — but the AgSnO₂ material and metal base keep it safe. The G2R at 10 A dissipates 3 W in the contact — but with AgCdO and PCB mount, it needs airflow. Pick by real watts, not by label.

Topology/standards per the cited standards; all product ratings are manufacturer-stated values from the cited datasheets, current to 2026-06; derived/illustrative figures are labelled as such. This is not an independent head-to-head test. Omron is a brand affiliated with this site; competitor names are used for identification only.