Quick answer: On most diodes the marked end is the cathode — a painted band on an axial part, a band or dot on SMD packages, and the +/– terminals on a bridge. When the mark is unclear, multimeter diode-test mode confirms it. Most diode failures in hobby and repair work are not electrical failures at all: a good 1N4007 works one way and not the other, the difference coming down to that band. Beginners learn “the stripe is the cathode,” which carries them far — until a glass diode’s tapered shoulder, an SMD part with no band, or a bridge whose cathode is four pins must be mapped. This guide decodes each family’s marking convention, then shows how to confirm any guess with a meter.
The Striped Band on an Axial Diode: What It Really Marks
The rule you were taught is correct for the most common parts. On an axial diode such as a 1N4001G or a 10A05, the banded end is the cathode, and forward current flows from anode to cathode. In the standard diode symbol, the vertical bar at the apex of the triangle marks the cathode, and current flows from anode to cathode in the forward direction; on the physical part, the printed or painted ring corresponds to that same cathode bar. When the meter’s red lead sits on the anode and the black on the cathode, the junction conducts and you read a forward drop in diode mode.
The confusion creeps in with glass and plastic bodies that add geometry. A glass diode often has the band painted near one end, but the glass may also be molded with a tapered or pointed shoulder. That point is not a second marking — it is just where the band happens to sit. If you read the point as “the anode” you will install a good part backward. The reliable habit is to treat the band as the single authority for polarity and ignore shape cues entirely.
The same convention is visible on a real part in the 1N4002G rectifier profile, where the DO-41 body carries exactly the band this section describes. One genuine exception exists: a few axial parts, mainly Zener and specialized diodes, use a second band or printed symbol that changes the marking. The workaround is not to memorize exceptions but to treat any unusual marking as a reason to reach for the datasheet. If a part carries more than one band, or the band shares an end with a printed letter, stop guessing.
Plastic-bodied axial rectifiers behave the same way, but the band can be harder to see because the body is opaque and the print is small. Tilting the part so the light falls across the body, or holding it so the band catches a reflection, takes two seconds and makes an otherwise unreadable mark obvious. When the band is genuinely gone, the meter test at the end of this article settles the matter without any visual guesswork.
SMD Diode Marking: Reading Direction Dots, Bands, and Body Codes
Surface-mount diodes rarely print the word “cathode.” Instead the cathode is marked with a band, a dot, or a silkscreen line on the PCB, and the three conventions are easy to confuse because the marking is tiny and sits on one edge.
The most common SMD signal diode, the SOD-323 and its relatives, uses a single band or stripe at one end of the body. That banded end is the cathode, exactly as on an axial part. If you hold the diode with the band on the left, forward current flows right-to-left through the body. The same band convention carries across SOD-123, SOT-23 dual diodes, and the tiny SOD-523 ESD packages.
Two details trip people up on SMD parts. First, the alphanumeric code printed on the body identifies the part number, not the polarity — the polarity cue is the separate band or dot. Second, some packages place the cathode mark at the top of the body rather than at the side, so the mark’s position relative to pin 1 matters. Manufacturers differ on whether the cathode is at pin 1 or pin 2, which is why the robust approach for any SMD part is to check the orientation dot or band against the datasheet’s package drawing.
When a part has no band at all — common on tiny DFN packages and some bare-die or no-silkscreen designs — the datasheet package drawing is the only safe reference. If the board assembler cannot visually orient the part, the drawing and the pick-and-place data decide the rotation. Getting this right up front is what separates a board that works from one that fails on the bench because a diode was rotated 180 degrees.
Bridge Rectifier Pinout: Finding +, -, and the Two AC Tabs
A bridge rectifier is four diodes in one package, which means the polarity question becomes a pinout question. Most single-phase bridge modules follow the same convention: the positive output terminal is marked “+”, the negative is marked “−”, and the two AC inputs are usually unmarked or marked “~”. The classic mistake is assuming the “+” and “−” corners align with the diode bands inside. They do not, because the bridge rearranges the four diodes so that two corners are the AC inputs and the other two are the DC outputs.
The mental model that prevents errors: the AC input goes across one diagonal, and the DC output comes off the other. If the “+” is at the upper right and the “−” at the lower left, then the AC terminals are the upper left and lower right corners. Reversing the AC input does not matter because it is AC, but swapping the DC output polarity does. The datasheet for the bridge — for example the standard bridge rectifier family — includes a pinout diagram that should be your first check before the part goes on the board.
If you have an unmarked or salvaged bridge, the meter test settles it. In diode mode you should measure roughly 0.5–0.8 V between an AC tab and the “+” output in one direction and OL in the other, and the same pattern for the “−” output. The “+” terminal reads as a cathode to both AC tabs, while the “−” terminal reads as an anode to both. That asymmetry is the fingerprint that tells you which corner is which, and it is the same logic used in a systematic terminal check on a bridge.
Three Quick Verification Methods Without the Data Sheet
When a marking is ambiguous, you do not have to guess. Three checks confirm polarity with a meter or a known-good part, and you should use at least one whenever the band is faint, the package is unfamiliar, or the part came from a drawer of unknowns.
Diode mode is the first and best check. Set the meter to the diode symbol, place the red lead on one end and the black on the other, and read the forward voltage. When the meter displays a forward drop near 0.5–0.8 V for silicon, the red lead is on the anode and the black is on the cathode. Swap the leads and you should read OL, meaning the junction is reverse-biased. That forward-reading lead assignment is the definitive answer and needs no data sheet.
The known-good comparison is a second method that works when you have a diode whose polarity you are sure about. If both parts share the same package style, the one with the band on the same end as your known-good sample is oriented the same way. It is crude but fast for a batch of identical-looking parts. The third method applies when the marking is unclear but the part has a printed code: read the code, look it up, and let the datasheet’s package drawing confirm the cathode location instead of trusting an unreadable band.
Each method has a boundary worth stating. Diode mode gives a definitive answer for a single, isolated junction. For a bridge module or a dual package, the lead assignment changes because there are multiple junctions, so “red lead is anode” applies per junction, not per part. And in-circuit readings can be corrupted by surrounding components — a diode measured on a populated board may read differently than it does out of circuit. When polarity is critical, desolder the part or verify against the schematic rather than trusting an in-circuit reading. The same three-reading discipline for judging a diode as good, open, or shorted is covered in the diode test mode guide.
How Marking Confusion Causes Real Field Failures
The cost of a polarity error is not symmetric. On an axial rectifier, installing the diode backward means the diode blocks when it should conduct, the capacitor never charges, and the circuit does not work — annoying but often repairable. The more damaging case is the reversed bridge rectifier. If the AC and DC terminals are crossed, the bridge can present a near-short to the AC line on the first power-on, blowing the fuse or stressing the downstream capacitor until it fails. The failure chain runs from a misread pinout to a blown fuse to a damaged electrolytic, and it happens on a part that was electrically fine.
The same class of error shows up on SMD boards. A rotated SOD-323 on a signal line may work intermittently or not at all, and because the part is tiny the mistake is easy to miss during inspection. On dual diodes such as a common-cathode or common-anode pair, misidentifying which pin is the shared cathode turns a working rectifier into a shorted one. These are all avoidable failures, and they are avoidable with the same habit: treat every marking as a claim to verify, not a fact to trust, and confirm with a meter before the part goes into a design you care about.
For designers choosing real parts, the polarity convention matters at the package-selection stage as much as at assembly. Package families differ in whether the cathode is marked by a band, a dot, or a silkscreen line on the board. The SMD rectifier package guide covers how these conventions interact with PCB layout and reflow, while the axial rectifier guide does the same for through-hole parts. When you select between package options, checking the marking convention early avoids a board spin later. And because a bridge’s terminals follow a fixed pattern, the bridge rectifier guide is a useful cross-reference when laying out the AC and DC sides.