RFIC Design


Analog/RF/mm-Wave IC Design Interview Questions

These are the questions that were asked in real interviews at companies like Apple, Qualcomm, MediaTek, Intel, Samsung, Skyworks, Keysight, pSemi, Qorvo, Anokiwave etc. for positions of senior to staff engineer.

  1. How will you design a PA with 30 dBm output power? What load impedance for PA will you choose, at what point will you start thinking of power combining? Hint: design unit PA with maximum voltage and current swings at output within reliability limits then move on to power combiners.
  3. Breakdown mechanisms in MOS. How it happens? What is gate breakdown, what is drain source breakdown etc.? Go into physics of breakdown.
  5. Have you designed any ESD protection circuit?
  7. What is the impedance looking into gate of transistor with degeneration inductor?
  9. Say you have matched LNA by doing source degeneration and a series inductor at input. What is the overall transconductance of the circuit? Hint: Trick question. GM of LNA is independent of device size at resonance. Check RF LNA Design page 5.
  11. How can you introduce real part in input impedance of CS amplifier? Hint: Inductive degeneration, drain-gate feedback
  13. How does device size affect noise in LNA design?
  15. Can you derive the noise figure of common source amplifier?
  17. How do you optimize transistor (in terms of bias, device size and layout) for ft, fmax and NF
  19. Do you know how to design bandgap, PTAT etc?
  21. There are two current mirror transistor, one mirroring current from other. How do you increase the matching between the transistors? Hint: Use high Vov, keep mirror ratio in tens not hundreds, cascode
  23. How can you match Vth of transistors?
  25. Vth increases with channel length or decreases? why?
  27. Increasing bias current of cmos increases the voltage gain or decreases it? Hint: Decreases because overall product (\(g_{m} \times r_{o} = \sqrt{WL}/I_{D}\)) decreases. Read Razavi Chapter 9 Page 354.
  29. How would you design two stage amplifier? Do you know how to compensate it?
  31. Why Miller compensation does pole splitting? You also add some resistor in Miller compensation, why?
  33. What is input impedance of CS amplifier with and without Cgd cap?
  35. Derive output impedance expression of CS amplifier including Cgd.
  37. How to increase gain of amplifier? How do you increase output impedance? Add current source load, stack devices, super transistor, how else? What about current? Is there a sweet spot for it? Hint: Maximize \(g_{m} \times r_{o}\), cascode, employing gain boosting.
  39. What is the IIP3 of two cascaded blocks? Under what conditions this cascade IIP3 formula is not valid? Hint: Cubic disortion is not the only method which creates IM3, many other mechanisms can end up creating IM3 (like 2nd order interaction: where 2nd order products from first stage intermods with fundamental of next stage to land at IM3). Classical IIP3 formula only considers IM3 from third order distortion
  41. How is IM3 affected when you increase one tone amplitude and other remains fixed? Hint: If you only increase one tone, IM3 near that tone increases 2dB/dB, IM3 on the otherside increases dB/dB. If you increase both tones togehter, IM3s on both side increase 2dB/dB. Check intermodulation of two tones for math.
  43. What is the relation between IIP3 and HD3?
  45. What is the noise figure of two cascaded amplifiers?
  47. Say you are given two amplifiers with some NF. You cannot change the amplifier design. Is there anything you can do to reduce the system noise? Hint: Increase input noise entering the system, your amp noise being relatively smaller now wont be visible as if like its NF has reduced
  49. What linearity metrics do you use for PAs? (P1,IIP3,ACLR,EVM,AM-AM, AM-PM)
  51. How does AM-AM or AM-PM effect EVM?
  53. Do you know what is memory effect?
  55. Draw Pout vs Efficiency curve of class A, B and AB amplifier
  57. If you are in class AB, how can you increase efficiency of the PA? assume passives are ideal.
  59. How do you bias PA? How do you take care of thermal runaway?
  61. What breakdown voltage in lower in CMOS and why? gate to drain or drain to source?
  63. For which impedance do you match your PA? Your signal has PAPR, your output impedance is dynamically varying. Which impedance would you choose? Hint: Loadpull simulations
  65. What is active load pull? Why do you use it?
  67. Cascode LNA. How do you choose device size of input transistor? How do you choose Lg and Ls inductors? How do you choose cascode transistor size? How does cascode affect NF? Check RF LNA Design
  69. Draw time domain waveform of output of RC and RL low pass and high pass filters if the input is square wave.
  71. What is current crowding in inductors? what are step symmetric inductors? pros and cons between signle ended and differential inductors?
  73. If you have CS amplifier, and gate is terminated with very high impedance, assume there is Cgd capacitance. What is the output impedance?
  75. How do you match 200 to 50 ohms? what is S21? Suppose the matching inductor has 1ohm series resistance. What will be the S21 now? (popular question)
  77. What is the resonance frequency, Q and BW of parallel RLC tank?
  79. Series resistor and inductor. If you adjust there values in such a way that Q remains same. How would it move on Smith chart for different values of R? Hint: Interivewer is trying to figure out if you know about Q circles on Smith Chart (those eye-shaped circles)
  81. Say a typical oscillator circuit with LC load. What is the output impedance when it oscillates? What is the output impedance when it is in transient phase where oscillations are trying to build-up? Hint: Open circuit when it oscillates
  83. Say parallel LC tank. What is the current in tank during resonance vs source current?
  85. What happens if I,Q are mismatched in RX? what about in TX? Hint: Mainly image rejection in RX and residual side band degrades in TX
  87. How do you layout a transistor? what techniques do you use to reduce parasitis? How to choose number of fingers?
  89. Pole zero analysis of typical OTA
  91. Say single stage differential opamp. If load transistors have mismatch, dc offset is produced. How would you take care it? Like how would you size the transistors, big or small? Hint: Big size is good for matching but higher Vov would serve better for this particular mismatch
  93. What is the significance of ft (transition frequency) and fmax (frequency at which unilateral gain becomes one) of transistor
  95. This is a current mirror circuit. Say noise is added at this point. What will be its transfer function to output?
  97. Input impedance of CG? What is impact of positive feedback from ro on input impedance?
  99. How does AM-AM and AM-PM affects IM3s?
  101. What is single side band modulation? Typical in TX where you only want to transmit one side band that is either flo+fbb or flo-fbb but not both together
  103. Movement of constellation points (in EVM measurements) with AM-AM, AM-PM, noise, memory effect etc
  105. If antenna impedance is varying, how do you take care of it in PA design?
  107. An ideal amplifier with gain -1, input output connected with resistor R, what is the Zin?
  109. How does the spectrum of square wave look like? What if you change the duty cycle to 25%?
  1. Where does \(\Gamma_{opt}\) lie in LNA design?

  2. Noise contribution summary in textbook common source LNA. What contributes to noise and how much?

  3. Did you do any link budget for your transceiver?

  4. Say you have a transmission line, you excite it with a pulse. What would be reflection at t=0? Trace how signal travels on line and tell me reflection at this this time or at this this place.

  5. If LNA has 20dB power gain, input and output impedances are 50ohms, and say LNA is matched. What is the voltage gain from input to output? If load impedance is say 100ohms, what is the voltage gain now?

  6. Say LNA has 20dB gain and 3dB NF. Say we put a 3dB attenuator before LNA. What is the total NF? Calculate it. It would come out 6dB, that means 3dB from attn and 3dB of LNA. But when we add attenuator, it is supposed to degrade total NF (just like an amplifier improves system NF), so what happened?

  7. Why do you do load pull? Does maximum efficiency happens at maximum power transfer? If not why not? Why do you do conjugate matching?

  8. What is loadline curve? Say you are biased in class-A. How do you chose Ropt which transfers maximum power? How do you choose Ropt if you are in class B? (questions from loadline perspective)

  9. How do you analyze NF of phased arrays?

  10. Say you have amplifier with -10dB S11. If you put a 3dB attenuator, what will be the overall S11.

  11. Ideal Wilkinson divider has infinite isolation. You put 2*Zo load resistor between port 2 and port3 in ideal Wilkinson. Say Wilkinson is lossy, what load resistor you need to put to get better isolation?

  12. If you short two ports (2 & 3) of Wilkinson and put 25 ohm load, what would be the S11?

  13. If you have LC resonant network in parallel with 50 ohms, draw how does impedance move with frequency in Smith Chart. If you reduce L and increase C such that resonance frequency is same, how does impedance move in Smith Chart now?

  14. Why do you source degeneration in LNAs? What happens if you do capacitive degeneration?

  15. You are given 1V supply, you need to deliver 27dBm power. What load impedance do you need? You found it out to be X. Say if you use this X load resistor with differential amplifier. What would be the output power delivered? How do you design matching network for it?

  16. Say you have QPSK constellation. Say you add LO noise in your system. How would it affect your constellation?

  17. How do you improve transistor matching? How do you take care of PVT?

  18. Derive s-parameters of series resistor.

  19. Match 50ohms to 200ohm and tell me voltage gain.

  20. Check slide 33 of Niknejad Matching Networks
  21. If two uncorrelated noise source P1 and P2 are combined through Wilkinson, what will be the total noise?

  22. Voltage source in series with ideal switch and output is resistive load. Switch is triggered by Sin wave. What will be the output spectrum?

  23. Say transistor is potentially unstable. How would you stabilize it? You cannot add resistors. What would you do?

  24. You have an inductor. How do you EM simulate it? Do you use two single-ended ports (one at each leg) or do you put a differential port between two legs? Pros and cons. Hint: Single-ended ports. Differential port would miss any common mode response.

  25. How do you simulate large-signal stability? Hint: Transient sims are your best bet.

  26. For small-signal stability, how is k-factor different than u-factor

  27. What is memory effect in PA? How do you know about it? Hint: If you see left and right side ACLR (or disortion) being asymmetric, this is memory effect. Check Memory Effect in PAs.

  28. In Smith Chart, \(\lambda/4\) moves by 180 degree, why?

  29. Say you have series RL circuit, switch and dc voltage source. You turn on the switch, draw transient waveforms across R and L. Turn off the switch, draw transient waveforms.

  30. How do you create S11 response like this using only passive components?
  31. RFIC interview questions
    Hint: RLC circuit. Set R to 50 ohms and LC to whatver resonance freq you want.

  32. Two lines going in parallel. One carrying input and other carrying return current. How would you change their characteristic impedance?

  33. What is electromigration

  34. Say you have a filter and LNA. Filter has some loss, LNA has some gain. Where would you put the filter, before or after LNA?

  1. What is OFDM? What are the main advantages and disadvantages? Explain it with block diagrams?

  2. What is OFDMA? What is the main difference between OFDM and OFDMA?

  3. Why do we use autocorrelation in an OFDM system? (To determine which parameters)

  4. What is cyclic prefix? Why do we need it? How do we determine its size/length?

  5. How do we make channel estimation in OFDM?

  6. What is the difference between MMSE and LMMSE?

  7. How do we define frequency spacing in OFDM? When we increase frequency spacing, what effects do we observe in time domain?

  8. What is MIMO? What is Massive MIMO? What is the difference? How do we employ these systems?

  9. What is antenna spacing? How can we properly place antennas in an antenna array?

  10. What is singular value decomposition (SVD)? How do we use it in transmit antenna selection?

  11. What is beamforming? How does transmitter know the location of receiver to select correct beamforming configuration?

  12. What is diversity? What are types of diversity?

  13. What is fading? What are the types of fading?

  14. What is coherence bandwidth, coherence time, delay spread, Doppler?

  15. How do we specify a wireless communication channel as flat or frequency selective?

  16. How do we design a waveform? What fundamental PHY layer parameters do we mostly consider?

  17. What is the difference between mean and median? Give me two examples: one for using mean rather than median, and other for using median rather than mean? Explain why you choose one instead of other.

  18. What are the variance estimation methods?

  19. Describe a communication system in term of its discrete time equivalent.

  20. What is the impulse response? Why do we use it?

  21. What is the difference between large-scale and small-scale effects of a wireless communication channel?

  22. What is the diversity combining techniques? Which one does perform better?

  23. What is Alamouti’s scheme? What is the main advantage? What is code rate of 2×2 Alamouti?

  24. What is the relation between autocorrelation and power spectral density?

  25. What is Fourier transform? Explain the duality of time and frequency?

  26. What should be the sampling frequency of the sum of two sinusoidal, 5Hz and 10Hz? How many FFT point would you need for it?

  27. Assuming the noise level is 0dB, if you are given two sinusoidal (each is 10dBm) which are multiplexed, what would be the SNR? If the noise level is 5dBm, what would be the SNR?

  28. What is ML and MAP receiver? What is the difference? Give me two scenarios for each receiver in which one performs better than other.

  29. What is transmit diversity drawback? How can we overcome it?

  1. Where is the short and the open on Smith Chart? How do you move from there if you have lossless transmission line in series?

  2. What is the TEM mode on transmission lines?

  3. Draw cross-sectional view of microstrip transmission line. How does the E-field look like?

  4. How do you setup your ports in HFSS? What are different port types?

  5. How do you make sure your HFSS simulation results are correct?

  6. Why does the wave port in HFSS need to be wider than the microstrip transmission line?

  7. What happens to characteristic impedance of transmission line when you play with design parameters like width, dielectric thickness, dielectric constant etc.

  8. What do you know about differential pair signaling? What is crosstalk?

  9. How does surface roughness affect transmission line performance?

  10. What HFSS solution modes are there? Which one do you use and why?

  11. For a real capacitor, what happens to the impedance when the frequency goes up?

  12. How can you reduce loss of microstrip transmission line? You need certain impedance, you cannot change the width. What else would you do? (Hint: increase dielectric thickness)

  13. When you want to model a device, say a transistor, what information can you get from I-V curves and s-parameters?

  14. What do you know about linearization techniques for amplifiers?

  15. How are skin depth and transmission line width important?

  16. Find SNR of following system:
  17. rf interview questions
  18. What does the following circuit look like at DC and at high frequencies?
  19. rf interview question
  20. Questions about how return current flows in microstrip transmission line, what ground path width to set and what concerns are there.