asked by an interviewer.. couldn't solve this

1)what are the applications of differential input

2)in digital signals how to find a pulse width,frequency

3) Design a circuit to count number of (on,off) pulses from a pwm

4) Draw the interupt vector table for an arm processor

Thanks for your valuable inputs in advance
Thanks & Regards

Sunil

There are many possible answers for 1, 2, and 3. There is no one "right" answer.

You could give a precise answer for 4, by picking an ARM chip and giving a lot of detail for that one chip - but that is not what the interviewer expects. He/she wants you to show a general appreciation of how ARM interrupts work in general.

The interview is designed to allow you to give intelligent answers, not precise answers. If you have no clue on any of those topics, then the interview has succeeded in proving that you are not suited to the job.

Either find a job that you are suited to (and skilled in) or do a LOT of learning in the appropriate field before applying for interview.

The next interview will have completely different questions, so whilst practising good answers to 1-4 above is a good exercise, it will not help at next interview unless you have a good understanding of the field for which you are applying.

The best interviews questions have no yes/no or right/wrong answers. Such as those that you list 1-4

EDIT: oh.... and most interviews I had also contained simple right/wrong questions as well. These will have a simple scoring system, but will only amount to a proportion of your overall mark.

hexreader said:

There are many possible answers for 1, 2, and 3. There is no one "right" answer.

do a LOT of learning in the appropriate field before applying for interview.

Such as those that you list 1-4

Click to expand...

Thanks could you provide me some examples (it is out of a learning interest) !!
I'm really interested in this field(hardware embedded systems)!!

But i clearly dont have enough knowledge being a fresher(i thought basics would suffice)
so ill probably take up training in embedded to ***** further interviews

sushl said:

Thanks could you provide me some examples (it is out of a learning interest) !!

Click to expand...

You are talking rubbish.... what will you learn by me providing the answer - nothing!

Lazy, lazy, lazy.

You come up with your best attempt at an intelligent reply for each question, and I, or other forum members will help you improve on your answers.

Either put in some effort for yourself, or forget ever being employed.

Looking back at my last post I see I was unfair and harsh. I apologise.

I see that this forum does not allow me to remove unkind words - but that serves me right - maybe I will learn to be nicer.

Post your attempt at answering each question and I will do my best to assist.

hexreader said:

You are talking rubbish.... what will you learn by me providing the answer - nothing!

Lazy, lazy, lazy.

You come up with your best attempt at an intelligent reply for each question, and I, or other forum members will help you improve on your answers.

Either put in some effort for yourself, or forget ever being employed.

Click to expand...

hexreader said:

Looking back at my last post I see I was unfair and harsh. I apologise.

I see that this forum does not allow me to remove unkind words - but that serves me right - maybe I will learn to be nicer.

Post your attempt at answering each question and I will do my best to assist.

Click to expand...

That's okay! here's my attempt at the first question

1) differential input cancels out the in phase noise component in a signal and doubles the signal amplitude(since inputs are similar and 180 degrees out of phase)
Application :while using long cable wires which are balanced eliminates noise and interference.
Eg: audio cables which support XLR -mono ( have 3 wires +ve , -ve and gnd ).

2) Differential inputs are also used in differential amplifiers Which have applications in volume control , AM/FM radio signal recovery

The circuit for Volume control using differential amplifier is shown below



The circuit shows a current mirror in which the Signal IN is reflected on the transistors Q1 and Q2

By changing the audio control voltages of Q1 from -Vbe to +Vbe and with Q2 grounded we can change the volume from high to low.

You could also look at IEEE magazines where they discuss topics such as how to think as a designer.

Here's one article example.
https://ieeexplore.ieee.org/document/6532377/

Your answer to question 1 is way better than my answer would have been.

I see no issue there. You have shown understanding of differential signalling operation, why it is used, and provided an excellent real life example.

Score 10/10 IMHO (unless the interviewer wanted a long list of applications, but even then score might be 9/10 because you came across as intelligent and knowledgeable, even if you did not give exactly the answer he might have wanted)

Remember these are not usually right/wrong answers. Giving intelligent answers is the important thing, which you did brilliantly.

Now for answer to question 2.....

Thank you hexreader!

attempt for the 2nd Question:

In digital signals we can find the pulse width by measuring ON/OFF times if they have equal duty cycles
in case of changing duty cycles like a rectangular wave the pulse width differs shown below

**broken link removed**

the inverse of time period is frequency ,
hence frequency of a digital signal = 1/(Ton +Toff)

Perfectly good answer in the heat of an interview when under pressure - I will give you 7/10 (attachment does not work for me)

Marks are for fun - I am no expert and no interviewer.

2a) Really simple answer is "use an oscilloscope" and/or frequency counter. Might be all that the interviewer wants, might not be, but definitely worth a mention.

2b) A possibility is to use micro-controller with inbuilt counter to count how many pulses in a second to get frequency. This does NOT give pulse width

2c) Another possibility is to use capture-compare function of micro-controller to measure pulse width. This is best suited to lower frequencies. Calculate frequency using frequency = 1/period

There are other possibilities, but maybe the above is enough to show that you have some knowledge and can give an intelligent reply. I think you covered most of the above.

Doing well so far... seems I underestimated you..... Who am I to judge anyway :sad:

Now to question 3......

hexreader said:

Perfectly good answer in the heat of an interview when under pressure - I will give you 7/10 (attachment does not work for me)

Marks are for fun - I am no expert and no interviewer.

2a) Really simple answer is "use an oscilloscope" and/or frequency counter. Might be all that the interviewer wants, might not be, but definitely worth a mention.

2b) A possibility is to use micro-controller with inbuilt counter to count how many pulses in a second to get frequency. This does NOT give pulse width

2c) Another possibility is to use capture-compare function of micro-controller to measure pulse width. This is best suited to lower frequencies. Calculate frequency using frequency = 1/period

There are other possibilities, but maybe the above is enough to show that you have some knowledge and can give an intelligent reply. I think you covered most of the above.

Doing well so far... seems I underestimated you..... Who am I to judge anyway :sad:

Now to question 3......

Click to expand...

Ohh for the attachment is shown below


Thanks for your inputs !! Your opinion does matter !!

Oops , forgot to mention an example on how we can use a scope practically to find the ON/OFF pulse width of a digital signal.

I didn't know about 2b and 2c.

I'm not sure that the 3rd question is correct but here's a try!!

Question 3 - pulse width modulated digital signal can be used as a clock to a synchronous up counter which is positive edge triggered to find the number of ON pulses similarly we can use a negative edge triggered synchronous up counter to count the negative pwm pulses. (the number of bits to be used in the counter depends on the frequency of pwm)

4 bit synchronous up counter shown below



waveform



say we observe 100 pulses in a second on the scope then well have to use a counter with 100 bits(100 flip flops).