LDO and PMIC Interview Questions
An LDO is a pass element driven by an error amplifier, so treat it as a feedback loop in interviews. Know dropout, PSRR, and efficiency, and be ready to explain that the output capacitor's ESR sets a zero that keeps the loop stable. Too little or too much ESR can make the output oscillate.
Practice these questions live in a Analog Design mock interview with a verified engineer. Book a Analog Design mock interview
You are handed a schematic of a simple LDO on a bench. The board has a 10 uF ceramic on the output, and the output rings at a few hundred kilohertz. Why? And what would you change first?
Questions like this show up in analog and power management conversations because they test whether you understand the loop, not just the block diagram. This guide leans on two public Texas Instruments application notes, SLVA079 (LDO terms and definitions) and AN-1148 (linear regulator operation and compensation), plus a TI buck design report and a few TI product pages. The practice prompts are ours, not questions reported from any employer.
What is an LDO, actually?
An LDO is a pass element controlled by an error amplifier. AN-1148 describes the loop plainly: the output is sampled through a resistor divider into the error amplifier, which compares it to a bandgap-derived reference and drives the pass device. The output follows VREF times one plus R1 over R2.
Topology matters because it sets dropout and ground current. AN-1148 compares an NPN Darlington regulator (needing roughly 1.5 V to 2.5 V of headroom), a quasi-LDO, and a single-PNP LDO, where full-load dropout under 500 mV is typical. It also notes that a PNP's base current is wasted power: with a beta of 15 to 20, ground current can reach about 7 percent of load current. A P-FET avoids that, and SLVA079 treats a PMOS pass element as a resistor, so dropout is load current times on-resistance.
Efficiency is the other first-order answer. SLVA079 works a 3.3 V, 100 mA example: about 73.3 percent with a 4.5 V input and 82.5 percent with 4 V. Efficiency is mostly output over input voltage, so the headroom you give an LDO is heat you must remove.
How do the main LDO specs differ?
SLVA079 defines line regulation (output change as input changes), load regulation (output change as load current changes), and PSRR (how well input ripple is kept from the output, which depends on frequency). It points out that rejection between roughly 100 kHz and 1 MHz matters most when a switching supply feeds the LDO. TI's LDO overview page makes the same practical point: high PSRR over a wide bandwidth lets an LDO filter switching noise from an upstream DC/DC converter.
Load regulation says where the output settles. Transient response says what happens right after a load step, and SLVA079 ties that to loop bandwidth, output capacitance, and capacitor ESR.
Why can the output capacitor make or break LDO stability?
This is the part most worth rehearsing. AN-1148 explains that a PNP LDO adds a load pole at 1 over two pi times RLOAD times COUT, which moves with load current. Its worked example is a 5 V, 50 mA part with a 100 ohm load and 10 uF output, putting that pole near 160 Hz. With the internal compensation pole, that makes two low-frequency poles and, without help, an oscillating loop.
The fix in classic LDOs is a zero from the capacitor's ESR, at 1 over two pi times COUT times ESR. In the note's example, 1 ohm of ESR on 10 uF puts the zero at about 16 kHz and gives roughly 70 degrees of phase margin. Too much ESR (20 ohms in the example) pushes the crossover too high, and the note says bench data show ESR above about 10 ohms usually causes instability. Too little ESR removes the zero from useful range, and the note says large ceramics, typically under 20 milliohms, will make most LDOs of that style oscillate unless the part is designed for them. Parts like the LP2985, which the note names, build in an internal zero to allow ceramics. In an interview, say you would check the datasheet's capacitance and ESR range, including tolerance, DC bias, and temperature.
Buck converters and the PMIC context
A PMIC is usually several rails, often a mix of bucks, LDOs, and housekeeping blocks. For the buck side, TI's buck power stage report (SLVA477) gives the standard starting formulas: duty cycle near VOUT over VIN, inductor ripple current chosen at 0.2 to 0.4 times maximum output current, and output capacitance from inductor ripple divided by eight times switching frequency times allowed ripple voltage. ESR adds ripple current times ESR on top.
Startup, protection, and sequencing are where the sources we could open had the least depth. TI's public pages list enable and power-good pins on various LDOs, a foldback current limit on one part, and soft start on two bucks. For sequencing, read the datasheet of a specific part you have used rather than giving a generic answer.
A table of topics and practice prompts
| Topic | What the source says | Practice prompt |
|---|---|---|
| Dropout | PMOS pass element behaves like a resistor, dropout is load current times Ron (SLVA079) | How would dropout change at high temperature or low input voltage? |
| PSRR | Frequency dependent, important when fed by a switcher (SLVA079) | Why does PSRR usually fall at higher frequency? |
| Stability | Load pole, ESR zero, stable ESR window (AN-1148) | What happens when you swap a tantalum for a ceramic? |
| Buck ripple | Inductor ripple and output capacitor ripple formulas (SLVA477) | How does halving inductance change ripple and transient response? |
| Startup | Enable, power-good, soft start are product features (TI product pages) | What could go wrong powering two rails in the wrong order? |
Illustrative scenario: the LDO that oscillates
This scenario is ours, not a reported question. A board uses an older-style LDO with a 10 uF ceramic on the output, and the rail rings under light load steps.
- State the suspect: very low ESR means the ESR zero is far out, so two low-frequency poles can reach 180 degrees before crossover. As our own arithmetic, 5 milliohms on 10 uF puts the zero near 3 MHz.
- Check the datasheet for the required ESR window and capacitance range, including DC bias derating of the ceramic.
- Propose a test: add small series resistance, or try a tantalum, and watch the load step response.
- Offer an alternative: a part intended for ceramics.
- Admit what you cannot know without measurements, such as actual ESR at temperature.
If you get stuck, say so and reason forward, as in our guide to handling an unfamiliar question.
Where to go next
If feedback and phase margin feel shaky, work through op-amp stability, noise, and compensation first, since an LDO is an error amplifier in a loop. For what happens after the regulator, see power integrity and PDN questions. If you are weighing a path into this area, compare power electronics and IC design roles. Broader practice lives in top analog design interview questions.
When you want to say the explanation out loud, book a mock interview on MockVise.
Sources: TI SLVA079, TI AN-1148 (SNVA020B), TI SLVA477, TI LDO overview, TI buck converter overview. This article offers general technical preparation guidance, not an official interview process or reported company questions. SLVA079 and AN-1148 are older documents, so check current datasheets for specific parts. MockVise is not affiliated with or endorsed by Texas Instruments.FAQ
Q: Why does an LDO with a ceramic output capacitor sometimes oscillate? A very low ESR pushes the ESR zero far out, so two low-frequency poles can reach 180 degrees before crossover. Older-style LDOs need a capacitor within a specified ESR window unless the part is designed for ceramics.
Q: What decides an LDO's dropout voltage? Topology does. An NPN Darlington regulator needs roughly 1.5 V to 2.5 V of headroom, a single-PNP LDO is typically under 500 mV at full load, and a PMOS pass element behaves like a resistor so dropout is load current times on-resistance.
Q: Why does PSRR matter when a switching supply feeds the LDO? Rejection between roughly 100 kHz and 1 MHz matters most in that case. High PSRR over a wide bandwidth lets the LDO filter switching noise from the upstream DC/DC converter.
Q: How should I answer a question about sequencing or startup? Read the datasheet of a specific part you have used rather than giving a generic answer. The public sources behind this guide had the least depth on startup, protection, and sequencing.
Related guides
Op-Amp Interview Questions: Stability, Noise, and Compensation
Prepare for op-amp design interviews with clear explanations of phase margin, Miller compensation, noise, and a ringing amplifier debug story, grounded in public TI and university notes.
ADC and DAC Interview Questions: SAR, Delta-Sigma, and Pipeline Trade-Offs
Prepare for data converter interview discussions with a practical guide to SAR, delta-sigma, and pipeline architectures, latency, and noise, grounded in public technical documentation.
Mixed-Signal Roles Are the Hardest Interviews to Prepare For: Here's Why
Mixed-signal interviews test analog depth and digital fluency at once, which is exactly why candidates find them so hard to prep for. A breakdown of what's actually being evaluated.
What Analog IC Interviews Actually Test in 2026
Beyond the question banks: the handful of underlying skills analog IC interviewers are really probing for in 2026, and why the same handful of trick questions keep showing up.
Practice with engineers who've run these interviews
Book a 1-on-1 mock interview with verified experts from Intel, NVIDIA, Qualcomm, and Apple. Practice these questions live in a Analog Design mock interview with a verified engineer.
Find a Analog Design expertAlready running these interviews at work? Get paid to run them on MockVise, anonymously and on your own hours. Become an expert