Lesson 7/1741%
MODULE 07

Reading the Datasheet

A line-by-line dissection of real optocoupler datasheets: absolute maximum ratings and what actually happens beyond them, electrical characteristic tables and test conditions, CTR and derating curves, safety certification tables, and package drawings with creepage callouts.

Learning Objectives

After completing this module, you will be able to:

Navigate a full optocoupler datasheet with confidence
Extract worst-case values (not typicals) for design
Read CTR vs IF, temperature, and aging curves correctly
Verify safety certification tables against project requirements
Check package drawings for creepage and clearance data
Where this fits. Module 6 defined each parameter. This module teaches you to find them, read their test conditions, and pull the right number out of a real sheet, including the curves and the safety table. To avoid reproducing any vendor’s copyrighted document, the interactive sheet below is an original, generic layout for a fictional part (“RX-817”) with values typical of PC817-class parts. The skills transfer directly to real datasheets.
Illustrations and animations are original teaching drawings, not to scale. Values marked representative are typical of common parts; always confirm against the current datasheet or standard you design to.

Datasheets are dense but predictable. Once you know the map, you can find any number in seconds and, more importantly, you can find the conditions attached to it.

Generic page map of an optocoupler datasheet: page 1 features and certifications, page 2 absolute maximum ratings, page 3 electrical characteristics, page 4 switching, pages 5-6 curves, page 7 insulation table, pages 8-9 package drawings, page 10 ordering information
Figure 1. The typical order of sections. Page 1 sells, the tables guarantee, the curves explain, and the insulation table certifies.

Interactive: Annotated Datasheet (fictional RX-817)

An original, generic datasheet excerpt. Click or tap each underlined item for an explanation.

RX-817 — Phototransistor Optocouplerfictional example part
Absolute maximum ratings (TA = 25 °C)
Forward current IF50 mA
Peak forward current IFM1 A (tp ≤ 100 µs, duty 1 %)
Reverse voltage VR6 V
VCEO80 V
VECO6 V
Total power Ptot200 mW
Isolation voltage VISO5000 Vrms (1 min)
Operating temperature−55 … +110 °C
Soldering temperature260 °C, 10 s
Electrical characteristics (TA = 25 °C)
ParameterMinTypMaxConditions
VF–1.2 V1.4 VIF = 20 mA
ICEO––100 nAVCE = 20 V
CTR50 %–600 %IF = 5 mA, VCE = 5 V
VCE(sat)–0.1 V0.2 VIF = 20 mA, IC = 1 mA
RIO5×1010 Ω1011 Ω–500 V DC
CIO–0.6 pF1.0 pFf = 1 MHz
tr / tf–4 / 3 µs18 / 18 µsVCE = 2 V, IC = 2 mA, RL = 100 Ω
Insulation characteristics (per IEC 60747-5-5, option V)
VIORM890 Vpk
VPR (method b)1669 Vpk, 1 s, < 5 pC
VIOTM8000 Vpk
Creepage / clearance (ext.)≥ 8.0 / ≥ 8.0 mm (option W)
DTI≥ 0.4 mm
CTI / material group175 / IIIa
Climatic category55/110/21
Ordering information
RX-817A / B / C / DCTR 80–160 / 130–260 / 200–400 / 300–600 %
Suffix “W”, “S”, “T”wide lead / SMD / tape & reel
Click any underlined line

Each highlighted item on the sheet explains what the number really means, the conditions attached to it, and the design trap hiding behind it.

“Absolute maximum” means exactly that: the stress beyond which the manufacturer no longer guarantees the part will survive. Nothing in the electrical-characteristics table is guaranteed there, and life shortens as you approach them. They are not design targets.

RatingWhat physically fails beyond it
IF, IFMLED junction overheating and current crowding; accelerated light-output degradation; bond-wire fusing at high pulses
VRLED reverse breakdown: localised heating damages the junction permanently
VCEOCollector–emitter avalanche breakdown; with energy (inductive loads), destruction
VECOEmitter–base junction breakdown at a few volts; degrades hFE
Ptot, PCJunction temperature beyond limits; encapsulant stress
VISOBarrier breakdown or partial-discharge damage (test rating, not working)
Tstg, Topr, solderingEncapsulant cracking, delamination, wire-bond fatigue, optical yellowing
Animation of a marker sweeping LED current from 0 to 75 mA across zones: recommended range, specified test points, allowed without guarantee, and beyond the 50 mA absolute maximum
Animation 1. Where a value sits relative to the recommended range, the specified test points and the absolute maximum. Designs belong in the green zone, with the absolute maximum derated for temperature.

Some datasheets (always for logic and gate-driver parts) include a recommended operating conditions table: the range over which the electrical characteristics apply. Typical entries: supply voltage (e.g. 4.5–5.5 V), IF(on) (e.g. 6.5–15 mA for a 6N137-class part), IF(off) or VF(off) (the input must be truly off, e.g. ≤ 250 µA or ≤ 0.8 V), and ambient temperature.

Pro tip — the off-state is specified too. Leakage from a driver pin can partially turn an LED on. A logic coupler with IF(off) ≤ 250 µA (typical figure) can glitch if the driving GPIO leaks, or if the LED has a pull-up through a long cable. Hold the input firmly off.

The electrical characteristics table has three value columns and one conditions column. The golden rule: design with the min or max that hurts you, at conditions corrected to yours. The typical column describes the median part. Roughly half of your production will be worse.

If you are calculating…UseBecause
Minimum output currentCTRmin × correctionsEvery part is guaranteed to exceed it
LED resistor for minimum currentVF(max)Highest VF gives the lowest IF
LED current maximum (stress)VF(min) or a hot, low VFLowest VF gives the highest IF
Leakage in the off stateICEO(max), at your hottest temperatureLeakage doubles about every 10 °C
Speedtr/tf max, converted to your RLTest circuit is not your circuit

Interactive: Spot the Trap

For each line, click the value you should actually design with. Wrong picks explain why novices choose them.

Score: 0 / 5

Switching times are only meaningful together with their test circuit. Phototransistor datasheets typically measure with RL = 100 Ω at IC = 2 mA in the active region. That is great for a fast number and unlike almost any real logic interface.

Two side-by-side circuits: the datasheet test circuit with a 100 ohm load and 2 mA collector current giving about 4 microsecond rise time, and a typical design with a 10 kilohm pull-up giving 60 to 100 microseconds
Figure 2. The same part in the datasheet test circuit and in a typical design. Use the “response time vs RL” curve (usually present) to translate, and remember the max column.
Worked example — translating tr
Datasheet: tr = 4 µs typ, 18 µs max at RL = 100 Ω. Your pull-up: 10 kΩ.
  1. The response-vs-RL curve (typical) shows roughly 60–100 µs at 10 kΩ.
  2. Scale by max/typ (18/4 ≈ 4.5) for a pessimistic bound: several hundred µs is possible for a worst-case part.
  3. For a 1 kHz signal (500 µs half-period) that is marginal, so reduce RL or choose a faster family (Module 5).

Curves are typical unless labelled otherwise. They tell you the shape of a dependency, which you apply as a ratio to a guaranteed number. Four curves matter most: CTR vs IF (often normalised to 1 at 5 mA), CTR vs TA, output current vs VCE, and response time vs RL. Some vendors also publish long-term CTR degradation curves; Module 9 uses them.

Reading log axes correctly

Most current axes are logarithmic. Each decade (0.1–1–10 mA) has the same width, and the minor gridlines for 2, 3 … 9 get closer together towards the next decade. A point one-third of the way through a decade is ≈ 2, not 3.3. Interpolate logarithmically: x = 10(log x1 + f·(log x2 − log x1)).

Animation: a cursor moves along a normalised CTR versus forward current curve on a logarithmic axis, showing the relative CTR value at each point
Animation 2. Reading a normalised CTR curve. Note how the minor gridlines crowd towards each decade line.

Interactive: Curve Reader

Drag across the plot (or use the slider) to read the normalised CTR, then apply a temperature curve and the datasheet minimum. Curves are the course’s representative teaching model.

Relative CTR (IF curve)–
Temperature factor–
Estimated CTRmin at this point–
–
Design trap — multiplying typical curves as if they were guarantees. The ratio from a typical curve is an estimate. For safety-critical margins, verify on parts from several lots, or specify a guaranteed CTR at your operating point with the vendor.

Derating graphs show the maximum allowed IF, collector power and total power vs ambient temperature. They are usually flat to 25 °C (sometimes higher) and then fall linearly to zero at the maximum temperature. Read them at your worst-case ambient inside the enclosure, next to whatever heats the board, not at the room temperature on the specification sheet.

Worked example
Ptot 200 mW at 25 °C, zero at 110 °C (fictional RX-817). Enclosure ambient 80 °C.
  1. Slope = 200 / (110 − 25) ≈ 2.35 mW/°C.
  2. Allowed at 80 °C: 200 − 2.35 × 55 ≈ 71 mW, just over a third of the headline figure.

The insulation (safety) table is the part of the datasheet a certification engineer reads. It is backed by the component certificates: UL 1577 (file number on page 1), VDE/IEC 60747-5-5 (certificate number), sometimes CQC or CSA. Key entries and how to use them:

EntryWhat to check
Installation class / OVC per mains voltageWhich overvoltage categories the part is rated for, at which rated mains voltage
Climatic category (e.g. 55/110/21)Min temp / max temp / damp-heat days of the test
Pollution degreeThe PD the internal and package ratings assume (usually PD2)
VIORM, VIOTM, VIOSM, VPRAgainst working, transient and surge voltages (Module 6)
Creepage, clearance (external)Against the end-product requirement; wide-lead options often exist
DTIMany standards require ≥ 0.4 mm for supplementary/reinforced solid insulation
CTI / material groupSets the creepage requirement for the package surface
Safety-limiting values (IS, PS, TS)The maximum input current/power and temperature under fault conditions for which insulation is still guaranteed

Interactive: Certification-Table Decoder

Checks the fictional RX-817 option W table (VIORM 890 Vpk, creepage/clearance 8.0 mm, DTI 0.4 mm) against your application. Uses representative IEC 60664-1-style values: confirm with the governing standard table and your end-product standard, which may require more.

Peak working voltage vs VIORM 890 Vpk–
Required creepage vs table 8.0 mm–
Rated impulse → required clearance vs 8.0 mm–
DTI 0.4 mm (reinforced solid insulation)–
–

Package drawings give dimensions with tolerances, the recommended land pattern and sometimes explicit creepage/clearance callouts. Use them to make sure your footprint does not quietly reduce the certified distances, and to choose between lead-form options.

Top and side views of a generic 4-pin SMD optocoupler on a PCB, annotating lead span, body width, external clearance, external creepage and the land-pattern pad gap
Figure 3. The distances on a package drawing that matter for isolation. The certified values come from the insulation table; the drawing shows whether your pads preserve them.
Design trap — extending pads under the body. Enlarging SMD pads for hand-soldering or “better fillets” moves copper towards the other side and can cut board creepage below the certified package value. Keep the vendor’s land pattern or check the gap explicitly (Module 11).

The ordering section turns one datasheet into dozens of real part numbers: CTR rank letters, lead-form options (DIP, gull-wing SMD, wide-lead), tape-and-reel, and certification options (e.g. a suffix for VDE-tested parts). Each option can change guaranteed values.

Code elementExample meaning (fictional RX-817)Why it matters
Rank letter A/B/C/DCTR 80–160 / 130–260 / 200–400 / 300–600 %Only a ranked part number guarantees the narrower window
“W” wide lead10.16 mm lead span, ≥ 8 mm creepageNeeded for reinforced 230 V at PD3, medical, etc.
“S” SMDGull-wing leads for reflowDIP versions are often not reflow-rated
“V” / “-000E” etc.VDE-certified, RoHS or lead-finish variantsThe VIORM claim may apply only to this option
Design trap — a generic BOM line. “PC817” alone lets purchasing buy any rank, any lead form, any certification option. Put the full ordering code (and approved alternates with verified insulation tables) in the BOM.

Two datasheets, two mindsets. A phototransistor sheet is organised around CTR; a logic-output sheet around thresholds and timing. Here is what to look at first in each (generic, not vendor-specific):

Look for…PC817-class phototransistor6N137-class logic output
Transfer specCTR min/max by rank at IF = 5 mA, VCE = 5 VIFH/IFL threshold (e.g. ≤ 5 mA); recommended IF(on)
OutputVCEO, IC, VCE(sat) at a test pointVOL at IOL, IOH leakage, pull-up range
SupplyNone (pull-up only)VCC range, ICC, mandatory 0.1 µF bypass
Timingtr/tf at RL = 100 Ω; response vs RL curvetPHL/tPLH, PWD, skew, data rate
ImmunityOften no CMTI specCMH/CML in kV/µs
ExtrasCTR rank lettersEnable pin, output type (OC vs totem-pole)
CurvesCTR vs IF/TA, deratingtP vs IF/TA, VOL vs IOL
Know the map

Features → abs max → electrical → switching → curves → insulation → package → ordering.

Abs max is not a design point

Stay in the recommended range; derate absolute limits with temperature.

Min/max, with conditions

Use the limiting column and translate its test conditions to yours.

Test circuits differ

Convert tr/tf to your RL with the response curve.

Curves are typical ratios

Read shapes on log axes carefully; apply them to guaranteed numbers.

The insulation table certifies

VIORM, creepage, DTI, CTI, PD test: compare with the end-product standard.

The drawing protects the gap

Check land patterns do not erode certified distances.

The ordering code is the product

Rank, lead form and certification option must be in the BOM.

Design checklist

Tick items as you review a design. Your ticks are saved in this browser only; use your browser’s print function for a sign-off copy.

Four parts: multiple choice, true/false, a matching exercise and engineering scenarios. Score 80 % on the multiple choice to pass the module.

Part A — Multiple Choice

Question 1 of 10  •  Score: 0/10

Part B — True or False

Score: 0/8

Part C — Match the Term to Its Meaning

Drag each term onto its definition, or click a term and then click a definition.

Placed: 0/6

Part D — Engineering Scenarios

Scenario 1 — The “typical” design that failed in production
An input was designed with CTR = 300 % (the value read off the typical CTR curve at 5 mA). In production, 8 % of boards fail the functional test. The datasheet lists CTR 50–600 %. What went wrong?
Analysis. The curve is typical, and the only guarantee is the 50 % minimum (unranked). About half the parts fall below the median and the tail reaches 50 %. Fix: design with CTRmin of the rank you buy, corrected for IF, temperature and ageing (Module 6). Order a ranked part number and increase IF or reduce the required IC as needed.
Scenario 2 — Wide-lead option or not?
A 230 V product for industrial cabinets (PD3, material group IIIa, reinforced) uses the fictional RX-817 in standard 7.62 mm DIP (7.0 mm creepage). The insulation table lists 8.0 mm for option W. Which do you order?
Analysis. At 250 V, PD3, group III, representative basic creepage is 4.0 mm, so reinforced needs 8.0 mm. The standard DIP (7.0 mm) fails; option W (8.0 mm) just meets it. Fix: order option W, confirm that the VDE certificate covers option W, and keep the PCB creepage at ≥ 8 mm too (slot if necessary).
Scenario 3 — Is 18 µs fast enough?
A designer reads tr = 4 µs typ, 18 µs max and concludes a PC817 with a 22 kΩ pull-up can pass a 10 kHz signal (50 µs half-period). Check the reasoning.
Analysis. The 18 µs max applies at RL = 100 Ω in the active region. With 22 kΩ the Miller term dominates: typical rise time is of the order of 100+ µs (response-vs-RL curve), longer than the half-period. Fix: lower RL to ≈ 1 kΩ with enough IF to saturate at end of life, or use a 6N136-class part.

Additional Resources

Practise on real sheets

Pick a PC817-class and a 6N137-class datasheet from two different vendors and locate every item in the annotated example. Compare their insulation tables and rank codes.

Standards referenced

IEC 60747-5-5 (insulation table) • UL 1577 (VISO) • IEC 60664-1 (creepage/clearance) • J-STD-020 (MSL, reflow).

Keep learning

Module 8 turns these readings into a selection method; Module 9 applies ageing curves; Module 16 explains the certificates behind the insulation table.