NivaarExam PrepOfficial exam papers ↗

23-Mechatronics-A2 Circuits and Electronics · December 2018

Question 10 of 11: Enhancement-load NMOS inverter — voltage transfer characteristic

Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)

Notes on this paper

Paper: National Exams, December 2018 — 16-Mex-A2 Circuits and Electronics. Closed-book, 3-hour paper (approved Casio/Sharp calculator only). Two parts: Part A — Circuits (Q1–Q6) and Part B — Electronics (Q7–Q11); candidates normally answer 5 of the 11 questions (3+2 or 2+3 split). Full worked solutions to all eleven questions are given below so the set can be used for study regardless of which five a candidate chose.

Reference texts: C. K. Alexander & M. N. O. Sadiku, Fundamentals of Electric Circuits (7th ed.) — resistive networks, superposition, Thévenin/max-power transfer, first-order and second-order transients, AC steady-state nodal analysis; W. H. Hayt et al., Engineering Circuit Analysis (9th ed.) — Laplace-domain circuit models; A. S. Sedra & K. C. Smith, Microelectronic Circuits (8th ed.) — op-amp T-network feedback, diode rectifiers, MOSFET common-gate stages, NMOS inverters, and BJT bias-point analysis.

Reading the figures. The paper supplies the schematics but no separate figure data; every network below is redrawn element-by-element directly from the original drawing (several figures are small hand-style schematics). All node/reference choices and source polarities are stated with the solution. Q11 uses printed variable names (e.g. “VE”) exactly as labelled on the original circuit, even where the labelled terminal is physically the collector in one sub-part — this is called out explicitly where it occurs.

Question 10: Enhancement-load NMOS inverter — voltage transfer characteristic [12, 2, 2, 2, 2]

Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.

+VDDM2M1+ vOUTvIN+
Figure-10: enhancement-load NMOS inverter — $M_2$ (diode-connected load) on top, $M_1$ (driver) on bottom, identical devices.

$M_2$ is the load, wired diode-connected ($V_{GS2}=V_{DD}-v_{OUT}$, always driven by its own drain current) with its source at $v_{OUT}$ and drain at $+V_{DD}$; $M_1$ is the driver, gate at $v_{IN}$, source grounded, drain at $v_{OUT}$. No numeric $V_{DD}$ or $V_{TN}$ is given, so every breakpoint below is expressed symbolically in terms of $V_{DD}$ and the common threshold $V_{TN}$.

Region A — $v_{IN} $M_1$ carries no current, so $M_2$ (diode-connected, always needing $I_{D2}=0$ here too since the two are in series) also carries none; with zero drain current, $M_2$’s $V_{GS2}=V_{TN}$ exactly, giving the highest output the inverter can reach: $$V_{OH}=V_{DD}-V_{TN}$$ (an enhancement diode-connected load, unlike a resistor pull-up, can never lift $v_{OUT}$ all the way to $V_{DD}$).

Region B — both transistors in saturation. As $v_{IN}$ rises just past $V_{TN}$, $M_1$ turns on in saturation while $M_2$ (diode-connected) is always in saturation whenever it conducts (its own $V_{DS2}=V_{GS2}>V_{GS2}-V_{TN}$ is automatic). Equal $K$ for identical devices and equal drain current through both (they are in series) gives $V_{GS1}-V_{TN}=V_{GS2}-V_{TN}$, i.e. a very steep transition — the hallmark of an enhancement-load inverter’s narrow, near-vertical switching region. The switching threshold $V_M$ (where $v_{IN}=v_{OUT}$, a convenient reference point) satisfies $K(V_M-V_{TN})^2=K(V_{DD}-V_M-V_{TN})^2\Rightarrow V_M=\dfrac{V_{DD}}{2}$ for identical devices.

Region C — $M_1$ in triode, $M_2$ in saturation. Once $v_{IN}$ is well above $V_{TN}$, $M_1$ is pulled into the triode region and $v_{OUT}$ drops toward its logic-low value; $M_2$ remains saturated (diode-connected devices are saturated whenever $I_{D2}>0$). $V_{OL}$ is the (small, non-zero) drain-source drop across $M_1$ in deep triode carrying the fixed current set by $M_2$’s saturation equation — it approaches, but never reaches, zero.

vINvOUTVOHVOLVILVIHA: M1 cutoff, M2 triodeB: M1 sat, M2 satC: M1 triode, M2 sat
VTC sketch for the enhancement-load inverter, with the three operating regions and the noise-margin breakpoints labelled symbolically ($V_{TN}$ not numerically given).

Noise margins. With the standard unity-slope ($dv_{OUT}/dv_{IN}=-1$) definitions of $V_{IL}$ and $V_{IH}$ (found on either side of the steep Region-B transition): $$NM_L=V_{IL}-V_{OL},\qquad NM_H=V_{OH}-V_{IH}.$$ Because the transition region is very narrow for an enhancement-load inverter (both devices have the same $K$, so the region-B slope is steep), $V_{IL}$ and $V_{IH}$ sit close together near $V_M=V_{DD}/2$, which tends to make $NM_L$ larger and $NM_H$ smaller than in a comparable resistor-load design — the load never reaching $V_{DD}$ caps $V_{OH}$ below the supply rail and directly narrows $NM_H$.

Region$M_1$$M_2$Output
A: $v_{IN}CutoffSaturation (no current)$V_{OH}=V_{DD}-V_{TN}$
B: switching regionSaturationSaturationsteep transition through $V_M=V_{DD}/2$
C: $v_{IN}\gg V_{TN}$TriodeSaturation$V_{OL}$ (small, set by $M_1$’s triode drop)