Model carrier travel across a device under bias. Explore length, velocity, and saturation constraints here. See time limits and bandwidth predictions in seconds today.
This calculator estimates the transport delay and a practical bandwidth limit from carrier transit time.
Use this as an engineering estimate. Real devices may include diffusion, ballistic effects, RC limits, and distributed fields.
| L (µm) | Mode | V (V) | E (kV/cm) | μ (cm²/Vs) | vsat (m/s) | τ (ps) | fc (GHz) |
|---|---|---|---|---|---|---|---|
| 1.0 | Voltage | 1.0 | — | 1400 | 1.0×105 | 10.0 | 15.9 |
| 0.2 | Field | — | 20 | 800 | 2.0×105 | 1.0 | 159.2 |
| 2.0 | Voltage | 3.3 | — | 450 | 1.0×105 | 20.0 | 8.0 |
Example values are illustrative. Your results will vary with parameters and limits.
The transit time limit links carrier travel delay to a practical speed ceiling. If carriers need a significant fraction of a signal period to cross the active region, the device cannot fully respond. A common screening metric is fc ≈ 1/(2πτtotal), where τtotal includes intrinsic and added delays.
Many high-speed devices are drift-dominated, so field strength sets the drift velocity. In voltage mode the calculator uses E ≈ V/L to form a representative field. In field mode you can enter E directly, which is useful when you have a simulated average field in a channel, depletion region, or drift layer.
At moderate fields, drift velocity follows v ≈ μE. Mobility μ (m²/(V·s)) captures scattering, doping, and material quality. Higher mobility boosts velocity at the same field and reduces transit time. This is why short devices in high-μ materials often show better frequency potential at a given bias.
At high fields, velocity becomes limited by a saturation value vsat. Once μE exceeds vsat, increasing E raises stress and power but produces little speed gain. The calculator enforces vd = min(μE, vsat) to reflect diminishing returns in fast channels and high-field drift regions.
Transit time scales roughly with the effective transport distance. Halving length can nearly halve τ when velocity is unchanged. The effective path factor helps represent longer electrical paths caused by geometry, fringing, or nonuniform conduction. For instance, a 1.25 factor implies carriers travel 25% farther than the drawn length.
Mobility often decreases with temperature due to stronger phonon scattering, reducing drift velocity. The optional scaling μ(T) = μ(300 K)·(300/T)n provides quick sensitivity checks across operating conditions. Use it to compare hot and cold corners when transit time is near the required timing budget.
Even if intrinsic transit time is tiny, interconnect and loading can add delay. A lumped parasitic term tparasitic captures extra latency from routing, contacts, and capacitance. Because fc depends on τtotal, a few picoseconds can materially lower the achievable bandwidth when intrinsic τ is already small.
Use the outputs to compare options consistently, not as a final guarantee. First determine whether operation is mobility-limited or saturation-limited by comparing μE to vsat. Then evaluate whether shortening length, changing material, or reducing parasitics gives the largest improvement. Export runs to document design tradeoffs and assumptions.
It computes intrinsic transit time from effective length and drift velocity, then adds an optional parasitic delay to estimate total delay and two frequency limits: fc ≈ 1/(2πτtotal) and a simple 1/τtotal bound.
Choose field mode when you already know a representative electric field from simulation, datasheets, or measured operation. It avoids assumptions about how voltage drops across the transport region and is often better for nonuniform geometries.
Voltage mode is convenient for quick estimates when you know bias and length. It approximates E ≈ V/L, which is reasonable for a uniform region. Treat it as a screening step before detailed field modeling.
It multiplies the physical length to represent longer electrical paths from fringing, curvature, or nonuniform transport. If carriers are likely to travel farther than the drawn length, use a factor above 1 to keep timing estimates conservative.
When μE becomes larger than vsat, velocity saturates and no longer rises linearly. Additional field increases stress and power, but transit time changes little because drift velocity is capped near vsat.
Enter an estimated extra delay from routing, contacts, and loading. If unsure, sweep a range such as 5 to 50 ps and see how strongly the cutoff frequency changes. In many fast designs, parasitics dominate.
No. They are first-order, single-time-constant estimates. Real devices may be limited by RC effects, diffusion, nonuniform fields, or package constraints. Use these results to compare design directions and then validate with detailed models.
Important Note: All the Calculators listed in this site are for educational purpose only and we do not guarentee the accuracy of results. Please do consult with other sources as well.