HWN - Real "Hardware Engineer I" Interview Questions

Hi fellow (and future) engineers!
In this episode we examine real interview questions from one of our community members "Jeff". The position he applied for was for "Hardware Engineer I" at a midsize company and the questions were fantastic for recent college grads. Enjoy!
Need FREE help preparing for your interview like Jeff? E-mail us at: [e-mail]
Welcome back to another episode of Hardware Ninja! Ever wondered how to get a job as a Hardware Designer in some of the top tech companies like Microsoft, Google, Tesla, Apple, Facebook, etc?
The interview process employs many application based questions whether you're a recent college grad or an experienced applicant. It doesn't matter if you want to work for Elon Musk, Tim Cook, or the up and coming start-up you heard about. We're here with curated material of real life technical interview questions.
Please consider subscribing to the channel and supporting our movement. Prospective job seekers (including yourself some day) will benefit from this resource.
LinkedIn: / hardware-ninja
Twitter: / bytor_hardware
Background Music
Merry Bay - Ghostrifter Official
/ ghostrifter-official
#analogelectronics
#cmos
#vlsi
#vlsidesign
#designengineer

[@lordpaul94]

A perfect square wave only need a 1 bit ADC to capture the signal properly

[@ramanf]

The more complete explanation will be the fact that voltage source should have low output impedance. Inorder to achieve low impedance with a voltage divider you must use low value resistors. Which will dissipate power even when the load is low. That is inefficient. On the other hand if you are driving a very high impedance load you can get away with a voltage divider.

[@HardwareNinja]

Hi, Thank you so much for contributing to the discussion and being a part of the community! You're absolutely correct. This is a natural follow up to our question to continue the interview. However, picking a voltage divider independent of what you're driving is never a good choice. PSRR is suffers and it could be very wasteful if you're trying to deliver high current to the load (whether high or low impedance).

[@romyaz1713]

PCB "taking 5V as input" is so vague, i was sure it was on purpose. my immediate question would be - what is the input impedance of the PCB input. if it is infinite, resistor divider is perfectly fine. also, are we trying to supply the PCB or just provide an input signal - that is critical info

[@AstorzTrading]

it depends *highly* on the application. Linear regulators regulate using a continuous control to the pass transistor which causes it to work in the linear mode, while buck circuits use a pulsed control. which cause the pass transistor to work in the switch mode. Switch mode can be more efficient because the transistor is either on or off. Power is P=I*V so with zero current OR zero voltage there is no power in the transistor, whereas in the linear regulator there is always at least some power in the transistor which gets dissipated as heat and therefore is wasted energy. In spite of this general view, a more detailed view shows that a linear regulator, especially a low dropout type, can be more efficient. This happens when the input/output differential voltage is low. For example, with 5.5v in and 5v out, the linear regulator is 91 percent efficient. That is not that easy to get with a buck regulator as most typical bucks are about 80 percent efficient. So using a typical buck there would cause a loss of efficiency of 10 percent compared to the linear. Buck circuits can get as high as 90 percent, but it takes some effort usually. High quality parts and careful design will get you there, but a run of the mill buck like the Simple Switcher line wont cut it. So for this application (10V to 5V) a DC/DC Buck converter SMPS is more suitable...

[@Lehpurdzzz]

My initial response to the square wave question is 1-bit to represent the two states we expect to see.

[@justinnamuco9096]

Same

[@qemmm23]

Pass P_MOS+Err amp +Vref+2 resistors+Add capacitive coupling at the input and output ends 🤔

[@coolwinder]

How does the thinking go for the solution to the first question about the ADC resolution?

[@7volatileint7]

what about a 5v zener? less efficient but quicker and less expensive for debugging, i suppose -- though i agree that the buck converter is the most efficient since the delta in i/o voltage is lost to heat for the linear reg

[@dhaneshprabhu72]

For qn 2, will zener diode of 5V be a good answer

[@amarvir7444]

You need a 69-bit ADC for your square wave : D

[@alexanderwalton4930]

What are the main care-abouts for this system? I think that’s the first question the interviewee should ask back to LDO/buck question

[@HardwareNinja]

Hi Alexander, Thank you so much for your comment and contributing to the community! You're absolutely right. This is usually the follow-up to the question of how to solve the initial problem.

[@qemmm23]

Yes

[@ikponmwosai.k.evbayiro1771]

You would use the buck converter as you won't have as much wasted power do to LDOs burning the current and kind of acting like a regulated voltage divider

[@1kuulko]

What about powering 2 boards in series? Granted their power consumption is fairly similar... :)

[@AvantGrade]

points for making the interviewer laugh 😂😂

[@imwatching2901]

For that type of job do you have to learn coding?

[@HardwareNinja]

It's not required for all positions, but it is for some. It also depends how versatile you'd want to be :)

[@justinnamuco9096]

Giving transformer as the answer would be unacceptable at all?

[@ryan_17kosh8]

transformer only works with AC. It's an understandable answer if you dont understand this, try researching.

[@vstop2176]

Choping