Small-signal diodes are everywhere in a board — logic inputs, I/O protection, level shifters, and switching trains — and their datasheet rarely tells you the thing you actually need to know. The reverse-recovery time is quoted, but what matters in a logic or switching circuit is how the diode blurs or clips a real signal, which depends on capacitance, stored charge, and the speed of the surrounding edges. The small-signal family — the 1N4148, the BAS16, the BAV70, the 1SS355 — differs in exactly these parameters, and choosing among them is a balance of speed, capacitance, and leakage. This article puts the switching-speed question on the numbers that predict signal behavior, then maps each family to the circuits where it earns its place.
What Switching Speed Really Means for Small-Signal Diodes
For a small-signal diode, switching speed is not about how fast it carries amps — it is about how fast the junction changes state when a signal edge arrives. The relevant quantity is reverse-recovery time, the interval the junction spends sweeping out stored charge before it blocks. During that interval the diode conducts in reverse, and on a fast signal edge that reverse conduction blurs the waveform and adds a glitch the downstream circuit sees.
The datasheet’s recovery figure is only part of the story. The real edge behavior also depends on the diode’s capacitance and the slew rate of the driving signal: a diode that recovers quickly in the datasheet’s test can still distort a very fast edge if its capacitance is high. That is why the family comparison in the next section is read on capacitance and charge as much as on recovery time.
For most logic and I/O circuits, the practical question is simpler than the physics: will the diode’s recovery and capacitance introduce a glitch or a delay large enough to matter at my signal speed? At a few megahertz, almost any small-signal diode is fast enough; at tens of megahertz and above, the fast variants start to separate from the general-purpose parts.
Comparing the Small-Signal Families by What Moves Signals
The small-signal family is a set of near-equivalents with real differences hiding in the fine print, and the comparison below is read on the parameters that predict signal behavior — recovery speed, capacitance, reverse breakdown, and the physical package.
| Family | Reverse recovery (typical) | Capacitance (typ.) | Peak reverse voltage | Typical role |
|---|---|---|---|---|
| 1N4148 | Fast | Low-moderate | 100 V | General switching, logic |
| BAS16 | Fast | Low | 75–85 V | Fast logic, SMD |
| BAV70 | Fast, dual | Low | 70–75 V | Dual pair, OR-ing, clamping |
| 1SS355 | Fast | Very low | 90 V | High-speed signal path |
The table hides the within-family spread that the datasheet curve makes visible — the four parts are functionally interchangeable in many circuits, and the difference shows up only at speed or in capacitance-sensitive paths. The small-signal diode family on the site lists the real production variants, and the switching diode profile develops the fast-recovery behavior this table summarizes.
A worked edge example makes the numbers concrete. A logic line carries a 10 ns clock edge. A general-purpose small-signal diode with a recovery time of a few microseconds recovers long after the edge passes, but its low capacitance means the edge itself is not badly rounded — the transient it adds is brief and dies before the next edge. Push the clock to a few hundred megahertz and the same diode’s recovery window now overlaps hundreds of edges, and the accumulated blur becomes a real timing error. At that speed the very-low-capacitance, fast-recovery variant keeps the edge clean; the general-purpose part does not. The threshold where the difference appears is set by the edge rate versus the diode’s recovery window, and computing that ratio before selecting the part is the crux of the whole decision.
The datasheet number that predicts this best is not always printed plainly. Many small-signal datasheets quote reverse recovery at a modest di/dt and skip the capacitance column entirely; the practical workaround is either the family’s typical curve or a quick bench measurement of the diode’s effect on a known fast edge. A scope on the receive side of the diode, driving a square wave through it, shows the blur directly — the same scope-based method the VF measurement article uses for power parts, applied here to the small-signal edge.
Capacitance and Charge: The Data Points That Predict Signal Blur
Two datasheet numbers predict signal blur better than recovery time alone: the diode capacitance and the stored charge. Capacitance matters on a fast edge because the junction’s parallel-plate behavior couples the signal to ground and rounds the corners; stored charge matters because it is the physical quantity the junction must sweep out during recovery, and it sets the size of the reverse glitch.
The design rule is to match the diode to the signal edge. A low-capacitance, low-charge part like the 1SS355 keeps a fast digital edge clean; a general-purpose part with higher capacitance rounds the same edge. On a slow logic signal the difference is invisible, which is why the family selection is a “when it matters” decision rather than a universal rule.
The same two numbers drive the low-capacitance ESD protection choice, where a protection diode’s capacitance directly adds to the signal path — the identical trade that governs a small-signal diode’s blur, applied to a protection role.
Where to Select Fast Variants: Logic, I/O Protection, and Level Shift
The fast variants earn their place in three circuits. The first is high-speed logic and clock paths, where a low-capacitance diode keeps the edge clean and the timing tight — the 1SS355 or a BAV70 pair is a natural fit. The second is I/O protection, where the diode must clamp an overvoltage without loading the signal; the low-capacitance, low-leakage trade of the fast variants matters there exactly as it does for a dedicated ESD part.
The third is level shifting and clamping, where the diode’s speed decides whether the clamp responds before the overvoltage propagates. A fast variant clamps a fast transient cleanly; a slow one lets the edge pass before the diode acts. In each case the selection rule is the same: read the signal speed and the capacitance budget, then pick the variant whose numbers fit the path rather than the cheapest general-purpose part.
The switching diode families and small-signal Schottky family cover the two poles of this choice, and the SOT-23 switching diode profile shows the SMD form these parts take on real boards.
The selection checklist assembles the article into an executable order. Step one: write the signal’s edge rate and the capacitance the path can tolerate. Step two: check the standby leakage budget — a battery or always-on input limits how leaky the fast variant may be. Step three: compare the candidates on recovery and capacitance against the edge rate, and reject any part whose recovery window is a measurable fraction of the signal period. Step four: verify the package fits the board — the SMD form factors differ, and a fast part in the wrong physical family buys nothing if the assembly can not use it. Step five: prototype the fast edge through the selected part on a scope and confirm the blur is within budget before freezing the BOM.
The same checklist runs the protection variant of the selection. An I/O protector that is also a signal-path element needs the same edge-rate and capacitance accounting, because a protection diode mounted on a high-speed line adds exactly the capacitance the signal budget must absorb. The low-capacitance ESD diode and the low-leakage small-signal diode are the two published companions to this article’s selection method, and the SOT-23 switching diode shows the dominant small-signal package in production. When the signal path is clean after the diode and the standby budget holds, the selection is done.
Speed and Leakage: Balancing the Small-Diode Trade
The small-signal trade that closes the selection is speed versus leakage. The physical mechanisms that make a diode fast — a lightly doped, low-capacitance structure — tend to push reverse leakage up, and a design that needs both fast edges and low standby leakage has to find the balance point, not the extreme.
The leakage side matters in battery-powered and standby circuits, where a diode that leaks a few microamps across weeks of idle drains a small cell; the low-leakage small-signal diode article covers the family’s leakage expectations. The speed side matters in the signal path. The honest selection holds both against the circuit: if the standby budget dominates, accept a slightly slower diode; if the signal edge dominates, accept a slightly leakier one.
The final decision is a two-line check. Line one: does the signal speed and capacitance budget force the fast variant? Line two: does the standby leakage budget allow it? If both answers are yes, the fast part wins; if they conflict, the design must relax one budget or move the protection stage to a different position. That two-line balance is the whole small-signal selection in its simplest form, and it is the same trade the wider diode material comparison applies at the family level.