Bicmos Logic 1980s Measure Propagation Delay Against Pure CMOS at Matched Load
# BiCMOS Logic (Late 1980s): Step 8 — Measure Propagation Delay Against Pure CMOS at Matched Load
## 1. Setting Up a Fair Comparison
To make Steps 6 and 7's circuit-level claim measurable, this step compares propagation delay for a BiCMOS output buffer against a pure CMOS output buffer, both driving the same external load capacitance $C_L$, both fabricated on comparable process geometry. The CMOS buffer's delay is dominated by its output transistors charging $C_L$ through their own channel resistance, following the same $t_{pd} \approx 0.69 \, R_{on} C_L$ relationship that governs any single-stage CMOS driver; the BiCMOS buffer's delay instead reflects the bipolar output transistor's much lower effective output impedance while saturated, plus a small added delay from the CMOS-to-bipolar base-drive stage itself.
## 2. Where the Numbers Land
At loads in the range representative of a long on-chip bus line or an output pad — tens to a few hundred picofarads — contemporaneous measurements on production BiCMOS processes typically showed the BiCMOS buffer completing the same transition in roughly half to a third of the time the equivalent pure CMOS buffer needed, with the advantage growing as $C_L$ grows, because the bipolar stage's low output impedance keeps $R_{on}C_L$ small even as $C_L$ increases, while the CMOS buffer's resistance-dominated charging time scales directly with $C_L$ with no such compensating advantage.
## 3. Companion Diagram One: Quantitative View
## 4. Companion Diagram Two: Structural View