Digital System Design

Professor. Neeraj Goel

(Refer Slide Time: 00:16)

So, this is the website. In this website, you can see here we can, there is one side where we can
say number of bits, number of bits could be 3, number of bits could be 2 or 4. So, let us say if
number of bits are 3. So, then it could be represented using 0, 1, 2, 3 and then minus 4, minus 3,
minus 2, minus 1 and similarly using 4 bits. So, there are four ways of representing here which
has been shown, let us quickly have a recap here.

(Refer Slide Time: 00:50)

These are the positive numbers, which we discussed first that 0, 1, 2 up to 15. So, this is the
positive number circle.

(Refer Slide Time: 01:03)

If we are doing it in signed manner, then we see this discontinuity is happening at two places. So
we are doing 0, 1, 2, 3, 4, 5, 6, 7 and after that, because this 100 is this is a sign notation, sign
magnitude, so basically 1 here means negative number and the rest of the 3 bits represent
magnitude.

So, you see, there is a discontinuity from 7, we came here to 0 and similarly here, this is a
negative number and if I add 1 to it, then it becomes 000, so there is a discontinuity here also. So
essentially in a signed representation, there are 2 discontinuities and representation works like
this.

(Refer Slide Time: 1:57)

In 1s complement, you again see there are two issues, one issue is that there is a discontinuity
here. So, if I add 1 here, from 7, it became minus 7. This is a continuous thing, minus 7 adding
plus 1 is minus 6, adding plus 1 is minus 5. So, that means going in a clockwise direction is
addition. But here also this is not discontinuity, but this is a kind of an error that this is also 0,
this is also 0, there are two 0s in the same number circle.
(Refer Slide Time: 02:38)

And this is how we can see our 2s complement. So, if you want to play a bit, you can play using
this particular website and have some fun.

(Refer Slide Time: 02:50)

So, let us get back to the presentation and so, now after understanding it from the number circle
perspective, let us have some theoretical background or let us have some mathematical
representation of this number circle or this particular type of number representation.
(Refer Slide Time: 03:06)

This particular number representation is called 2's complement. Why it is called 2's complement?
We will see. So, here I can mathematically define it that if a number is u, then it is magnitude is
representing 2's complement, if number is negative, then 2’s power n minus magnitude of u is
actually 2's complement, because it is to 2’s power n, so it is called 2's complement.

And since 2’s power n, 2’s power, we can also represent this using inversion or inverse or logical
invert of u and then plus 1. So, this is this is an easier way to find out 2's complement. So, we
can invert all the numbers and then add 1 that would be the 2's complement representation of any
negative number. So, as seen from our circle lines or numbers circle, we see couple of properties
here.

One property is there is single 0, the other thing is, you can see that MSB, Most Significant Bit if
it is 1, then number is negative, if Most Significant Bit is 0, then that number is positive. So, this
is also a nice property. So, that means, if I want to see whether a number is a positive number or
negative number, I need not to do any other calculation. I need not to do for example, subtract
bias or do some other comparison I can directly look at the Most Significant Bit, if that bit is 1
then number is negative, otherwise, number is positive.

The other thing is that I can always find negative of 1 number using 2’s power n minus of that
number; so which essentially means that negative of a number, I can find out using this 2’s
power n minus that number, so that means, if I do further negation that the number will become
positive. So, let us say if there is a number minus 13, I represented using this 2's complement
notation, it will become plus 13 and then again, I do negation then again, it will become 2’s
power then again it will become negative number.

The other thing is range of the number would be minus 2’s power n minus 1 to 2’s power n
minus 1 plus 1. I think there is a mistake here. So, this should have been a minus sign here, not
positive sign. So for example, for 4 bit numbers, the range is from minus 2’s power 3 to 2’s
power 3 minus 1. So, that means number ranges from minus 8 to 7, positive 7. So, this is the
range we get from 2's complement.
(Refer Slide Time 06:37)

And the other thing, which can be proven easily that addition, subtraction, multiplication all can
be done intuitively can be done using a standard binary addition method in 2's complement
notation. So, if I want to add 2 numbers, I can take 2's complement of these 2’s, these numbers
can be represented in 2's complement notation and addition of those numbers using a binary
arithmetic will give me correct results, although one of the number could be positive one of the
number could be negative or both the numbers could be positive, both the numbers could be
negative, it always gave me correct result using the simple binary addition method.

If I want to do subtraction, then the best way is that whatever number I want to subtract, I should
take 2's complement of that number. So, I should represent that number as a negative number.
Let us say that number was a negative number and I want to subtract that number. So taking 2's
complement of that negative number will become the positive number.

So, let us say the number was minus 30 and I wanted to subtract this minus 30. So, after taking
2's complement of that minus 30, it will become plus 30. So that is how this negation rule is
helping me that I can do this, the step would be first I need to take 2's complement of that
number and then add it like a regular addition.

And interestingly, even if I multiply 2 numbers using 2's complement, then the result will always
be correct. But here a caveat is that there is no overflow assume. So, that essentially means that if
I am adding 4 bit number, I am multiplying 4 bit number with another 4 bit number, the output
should not be more than 4 bit, then the result would be correct. Otherwise, it may not be correct.

So, there are always ways to fix it, so that is why multiplication is also workable or is can be
worked out using this 2's complement arithmetic. So, I can understand that it is because it is all
theory or using formula, it does not sink into our mind. So, let us take some examples and see
that how addition or multiplication will work.

(Refer Slide Time: 09:17)

So, let us take we will start with an addition. Let us say I want to add a 13 and with minus 3. So,
first I will do is I will represent 13 in 2's complement notation, because 13 is a positive number,
and here I am assuming 8 bit numbers, this method of 2's complement work for any number of
bits, but we I am showing it using 8, 8 bits here. So, it is just an example.

So first, I represent 13 using 2's complement because it is a positive number I need not to do
anything for sign bit and 13 I would represent as 1101 and I am (())(9:59) rest of the most
significant bits as 0. Now, I want to represent minus 3. So, for representing minus 3, first I have
to calculate 3 that means 1 1 and then or basically 6 0s and then 1 1 and then invert it, inverting it
will become six 1s and then 0 0 and then I have to add plus 1. So, that will give me minus 3.

1111 1101. Now because I want to add these 2 numbers, so I am adding 1 plus 1 is 0 and there is
a carry of 1. So, this because this is 0 0 carry would come directly 1 here, then again 1 plus 1 is 0
carry is 1, 1 carry plus 1 plus 1 will give me a result sum of 1 and carry is 1, because carry is 1
plus 1 will give me 0, again carry is generated 1 plus 1 0 again carry is generated, 1 plus 1 0,
another carry is generated 1 plus 1 0, another carry is generated and we finally will have a carry
here.

Now carry here, we have to ignore this carry and we can leave this carry, we can we can simply
discard this carry without worrying about the results and the remaining 8 bit gives me correct
result. So, minus 13 minus 3 essentially is 10 and that is what we can see from this number. Let
us see another number. Let us say I want to, I want to subtract a number 13 minus 5. So again,
what I will do is 13 and then take 2's complement of 5 and then add these 2 numbers.

So essentially, all subtraction would be converted into addition and there is no, so because the
negative numbers would be represented in 2's complement form, everything is addition here.
Now, in the same way because in the same way, if I will do this subtraction, then I would get a
correct result here. So, 13 would be represented by 0000 1101 and 5 minus 5 would be
represented if I will calculate. So, minus 5 is 1111 1011.

How do I got it, so, basically, 5 is 101. If I inverse it will become 1111 0 this is 1 and then 010,
010 and then I have to add plus 1 out of it, so it becomes 1011. So, this way, this minus 5 2's
complement of 5 would be calculated and if I add both of them, then 1 plus 1 is 0, carry of 1 and
this carry of 1, I add from to 1 then it is again 0 carry of 1 and this carry is added to 1, so output
is 0 here.

Now 1 plus 1 plus 1, sum is 1 and carry is 1 and now carry would be propagated here. So, this
will all be 000 and there would be 1 additional carry. Now let us see how will it work for
multiplication. So here to make my life simpler, I am taking 4 bit numbers, so there will be only
four partial products. So let us say I want to multiply 2 and minus 3, 2 is 0010 and minus 3 is 2's
complement of 3 that means 1101.

So if I multiply using this, how do I multiply? So first bit 1, I will multiply with this
multiplication, multiplicand. First bit of multiplier is multiplied with multiplicand 0010 and then
I will shift, I will put cross here, we remember all our decimal multiplication tricks. So, because
multiplier second bit of multiplier is 0, so we can simply say 0000 here. Now the third bit of
multiplier is 1, so again, this 1 would be multiplied with 0010 and I leave two places.
That means 0010 and then 2 places are left blank and for the last bit of multiplier, it is 1 again
here I will leave three places blank and then I will write the same multiplicand here 0010 and
finally, to get a multiplier output, I will add these four partial products. If I add then 0 can be
directly come here as 0 and this is 1.

And here in this row these are all 0s, so output is 0. In this row there is only 1, so output is 1. In
this row there is only 1 so output is 1. So, similarly we can do this partial addition and because as
we mentioned earlier, we are not going to consider overflow only because my input was 4 bit,
this was also 4 bit, my output I have to consider only 4 bits. So, the output is going to be 1010.

So, if I translate this 1010, it is going to be minus 6 in 2's complement notation. So, this
essentially gives us a confidence that yes, using 2's complement, we can represent any numbers
plus we can also do this this binary arithmetic addition and subtraction, multiplication using
simple operations or whenever we are doing operations, we need not to worry about the sign
weight or the underlying representation, under underlying representation of negative numbers,
underlined representation, if it is 2’s complement, it is always going to give us correct results.

(Refer Slide Time: 16:40)

So, this there is also one thing again, we have to note down that when my results are going to be
wrong, that also we should be clear. So, whenever we are going to cross that breakpoint or
discontinuity of my number circle, then I am either going in overflow condition or going under
flow conditions. So, this unless we know this overflow, underflow condition, we never know that
my results are correct or not.

So, if it is overflow that means number is outside the range of numbers that can be represented
using that many number of bits. So, for example, if it is 4 bits, and I can only represent from
minus 8 to 7 if number is if my result, addition result is more than 7, I cannot even represent
using 4 bits. So, that is why whatever results I am getting, it may be wrong, it will look wrong, it
is not actually wrong, but it may look wrong, because this number I cannot represent using that
many number of bytes.

So, how to identify this condition that is very important to know and important to practice or see
also, whether I am not reaching an overflow condition. So, overflow condition is a very can be
laid out or can be put in a very simple words, if addition of my two positive number is resulting
in a negative number. That means there is an overflow. So for example, in a number circle, I am
adding 5 plus 4, 5 plus 4 and I am representing only using 4 bits.

So, 5 plus 4 will result in 9 and using 4 bits, I can only represent up to minus 8 to 7. So, 7 is the
maximum largest positive number that I can represent, but my output is 9. So that means
whatever output my binary arithmetic will show will not be correct. I have to raise that there is
an overflow, this result is not correct.

So, to check this condition, if signing of, sign of both the operands are same and I am doing
addition and result is either of opposite sign or result is 0 that means it is an overflow. So
similarly for negative numbers adding negative numbers, so minus 3 minus 7, it is going to be an
underflow. So, this is how we can represent overflow and underflow and we also can clear
clarify one thing that if the signs are different, one number is positive, another number is
negative and we are adding these two numbers, there can never be overflow or underflow
condition.

So, the number which is larger is going to be win, going to win. So, for example, I am adding
plus number plus 5 and I am adding minus 8, plus 5 and minus 8 I am adding. So, whatever
number is larger in magnitude, your number will become of that particular sign, but it is not
going to be overflow or underflow. So, this simple condition check can tell me what is the
overflow or underflow condition.
One more topic or one more point which I would like to touch upon here that a negative number,
if I have to convert N bit number to M bit number, it is going to be slightly different here. So, let
us say I want to convert a 4 bit number into an 8 bit number. If number is positive, it is going to
be same, but if number is negative, then it can lead to a different representation, it can lead to a
wrong representation that is why we call it as a sign extended number.

So, in a sign extended number, let us say 1101 is a 4 bit number which essentially represents 1, it
represents minus 2. So, this minus 2 if I want to represent in 8 bit then I need to remember that
because it is minus 2, so it sign bit has to be 1 and minus 2 would in 8 bit would be 2’s to power
8 minus 2.

So, that means that this either I need to recalculate this number or the easiest or a trickiest
method would be that whatever is the sign bit here is that if I extend that sign bit to the larger
like rest of the MSB numbers, let us say M bit is larger and N bit is smaller, then M minus N bit I
can extend as a sign bit, then the number representation would be correct.

And this this sign extension will work even if my sign bit is 0, because then I would be extending
0s all the way, if sign bit is 1 that means, I am extending it for the negative numbers and it would
give me an easy method of converting from 4 bit number to 8 bit number or 8 bit number to 32
bit numbers, the representation is fast and otherwise, the ideally ideal method would be I had to
do for example, I need to convert from 8 bit number to 32 bit number I need to subtract that
number from 2’s power 32 minus the magnitude part. So, this sign extension trick can make it a
little fast.
(Refer Slide Time: 22:56)

So, here I would like to close and I can summarize that overall in last half an hour, 45 minutes
we have understood what could be various ways of representing negative numbers. So, one thing
you may, you may be thinking at this point that you are not going to use 1's complement, sign
magnitude, 2's complement, why we are studying all those things or why was the only
motivation to study those things were to find flaws in all those numbers?

We can, here I would say that it is not like that, they each of them has its own purpose or its own
qualities. Sometimes we do use sign magnitude number also, sometimes we do use offset or
biased method of representing negative numbers as well, we will see in some of the later classes
that how do we use them.

The other thing which we have learned in this class is how to do arithmetic operations. So, this
class would remain incomplete unless we do practice lot of questions or we do some of these
things ourselves. So, I will be posting a tutorial after this class so that you can work on those
questions. Some of those questions I would put as old questions and some of the questions which
you will try to solve yourself. Thank you very much.