SFH615AY-X019T — 5300Vrms Phototransistor Optocoupler (SMD DIP-4) from Vishay

What Are the Technical Specifications of SFH615AY-X019T?

Key numbers that matter: the CTR grade is the selection driver — per the Vishay SFH615A datasheet (Doc 83671, Rev 2.2), a -1 part (40%) and a -4 part (320%) are the same package, and the grade is what your circuit math uses.

5300Vrms isolation with 400mV saturation is why this family stays the default feedback optocoupler in PSU designs.

When Should You Use (and NOT Use) the SFH615AY-X019T?

✅ Use SFH615AY-X019T when:

• Feedback isolation in switch-mode supplies. The phototransistor output drives the PWM controller directly — this is the classic flyback feedback optocoupler, and 5300Vrms covers the primary-secondary isolation requirement.
• MCU logic isolation. Buttons, sensors, and 24V industrial signals crossing to a 3.3V MCU — the 400mV saturation pulls the input low cleanly.
• Ground loop elimination. Any signal path between two grounds you want to keep separate — opto breaks the loop, period.
• Battery-powered boards. The LED works from 2mA, and the CTR grades are specified at low current — the data sheet even calls out low-CTR-degradation behavior for long-life designs.
• Reinforced insulation needs. IEC 60950 reinforced to 400Vrms/DC, CTI ≥175, insulation thickness ≥0.4mm — the safety paperwork is already done for you.

❌ Don't use SFH615AY-X019T when:

• Your signal is fast. Fall time is 14–15µs — that's roughly 30kHz of usable bandwidth. For CAN, high-speed UART, or PWM feedback loops in the MHz range, use a logic-output optocoupler (6N137-class).
• The circuit needs CTR above 100% and you picked a -1 or -2 grade. A positive-feedback latch or relay driver that demands >100% CTR will intermittently fail with a low-grade part — forum builds hit exactly this when swapping PC817 for SFH615A without checking the grade.
• You're feeding AC directly into the LED. The reverse half-cycle pushes past the -3V danger point — avalanche damage, higher leakage, accelerated aging. Rectify or clamp first.
• Sustained 60mA drive. That's the absolute maximum rating, not an operating point. Running there ages the LED fast and CTR collapses over time.

What Are the Alternatives to SFH615AY-X019T?

The swap decision in one line: all of these are functional optocouplers — the questions are pinout (DIP-4 vs DIP-6), CTR grade, and isolation voltage.

But why does the grade steal the show? Because the same package spans 40% to 320% transfer ratio — for a PSU feedback loop or MCU isolation, any DIP-4 part works, as long as the grade math checks out before you commit.

What Are the Typical Applications of SFH615AY-X019T?

Switch-mode power supply feedback: the phototransistor carries the error-amplifier signal across the isolation barrier to the PWM controller — the classic flyback loop. The 5300Vrms rating covers the primary-secondary requirement with headroom, and the 400mV saturation keeps the loop clean.

Industrial signal isolation: a 24V sensor or limit switch crossing into a 3.3V MCU. The LED side sees the industrial rail, the transistor side talks to the logic — no common ground, no ground-loop noise, no level-shifting worries.

Battery-powered control boards: the LED runs from 2mA, and the CTR is specified at low current — a wake-up or state signal can sit in standby for years without draining the cell.

Telecom and adapter circuits: the end-stackable 2.54mm package and the -55°C floor make it a workhorse in outdoor and industrial adapters where the PC817's -30°C limit gets uncomfortable.

Why Buy SFH615AY-X019T from ICMASS?

CTR grade verified on every lot. We check the transfer ratio grade before shipment — mixed-grade reels are the classic optocoupler complaint, because a -1 and a -4 part look identical and the circuit math silently breaks with the wrong one.

Cross-reference support for the whole opto family. PC817, SFH617A, LTV-817, VO617A — send us the CTR your circuit needs, the isolation voltage, and the package, and we'll tell you which grade of which family actually fits.

BOM consolidation for PSU builders. The classic flyback feedback set is a rectifier (UF4007-class), an optocoupler, and a TL431 reference — one shipment from Shenzhen covers the whole loop.

Same-day dispatch, 5–10 days worldwide. Orders before 15:00 CST ship same day via DHL or FedEx. For volume orders, we source directly from the Vishay production line.

Frequently Asked Questions About SFH615AY-X019T

Q1: Can I swap SFH615A and PC817 directly?

A: Yes for pinout and function, no for the CTR math without checking. Both are DIP-4 phototransistor optocouplers with the same pin arrangement. The differences: 5300Vrms vs 5000Vrms isolation, 70V vs 80V output, explicit CTR grades vs a single ~50% rating, and a -55°C vs -30°C temperature floor. For a feedback loop, redo the CTR math with your grade's minimum before the swap.

Q2: Which CTR grade should I order?

A: Work out the CTR your circuit needs, then pick the grade above it. Grades are measured at 10mA: -1 is 40%, -2 is 63%, -3 is 100%, -4 is 160–320%. A positive-feedback latch or relay driver that needs >100% CTR will intermittently fail on a -1 or -2 — a forum build exactly like this worked with PC817 and failed with a low-grade SFH615A until the grade was fixed.

Q3: How much current should I drive the LED with?

A: Design at 10mA, never dwell at the 60mA absolute maximum. The 60mA figure is the survival limit, not an operating point — one troubleshooting thread found the LED being driven at ~60mA with rapid CTR weakening as the result. The recommended margin is at least 2.5:1 over what the circuit needs, covering temperature and aging drift.

Q4: Does CTR degrade over time?

A: Yes — design with the minimum CTR and a margin. CTR drifts with temperature (roughly ±45% over -40 to +85°C for this class) and ages (about 35% over a decade). One aging test showed output high level dropping 28% after 500 hours at 85°C. The datasheet's "low CTR degradation" claim reduces the slope; it doesn't remove the need for margin.

Q5: Can I feed AC directly into the LED?

A: No — the reverse half-cycle damages the LED. Beyond about -3V reverse, the LED risks avalanche damage, higher leakage, and accelerated aging. Mains zero-crossing or AC sensing needs full-wave rectification or a clamping TVS on the input first.

Q6: My optocoupler doesn't switch — where do I start?

A: Check the LED polarity, the emitter-to-ground connection, and the pull-up. Reversed LED (pins 1/2), an emitter that isn't grounded, and a missing output pull-up are the three classic "opto doesn't work" causes. Measure the LED forward voltage with a multimeter diode test — about 1.25V means the input side is alive and the problem is on the output side.

Q7: What does the 5300Vrms isolation rating actually mean?

A: It's the test voltage between input and output, and the safety rating that follows from it. 5300Vrms is the isolation test the part passes; the creepage (7mm standard, >8mm with the 400mil lead option) and 0.4mm insulation thickness back an IEC 60950 reinforced-insulation rating to 400Vrms/DC. One catch from forum threads: the PCB layout still has to maintain that creepage between primary and secondary — the part is compliant, the board has to be too.

Q8: Can it handle high-speed signals?

A: Not really — budget for about 30kHz of bandwidth. With 3µs rise and 14–15µs fall times, this is a slow, cheap isolation workhorse. For CAN, high-speed UART, or MHz PWM feedback, a logic-output optocoupler (6N137-class) with its own CTR-free design is the right tool.

Q9: Can a surge destroy the isolation?

A: Yes — a big enough transient punches through the LED and shorts the input side. Lightning and mains-coupled surges (a 4kV-class 1.2/50µs event) can break down the LED, destroying the isolation the part exists to provide. Forum guidance is consistent: clamp the input with a TVS and limit the current so the opto never sees the full surge.

Q10: What's the reverse voltage limit on the LED?

A: 6V — and staying well under it is free reliability. Exceeding reverse voltage damages the LED junction and accelerates aging. If your input can swing negative (AC, inductive kicks), add the rectification or clamp so the reverse excursion stays inside the spec.