I want to make a decimal to binary converter without using itoa() or vectors in C

Question

My code works up to a point. When I provide values above 1023 for num, the program goes wrong, below 1023 the program works perfectly. Someone knows what can be?

#include <stdio.h>
#include <stdlib.h>

int main()
{
    int bin = 0, num = 0, resp = 0, cont = 0;
    printf("digite um numero\n");
    scanf("%d", &num);
    do
    {
        bin = (bin * 10) + (num % 2);
        num = num / 2;
        cont++;
    }
    while (num > 0);
    do
    {
        cont--;
        resp = (resp * 10) + (bin % 2);
        bin = (bin - (bin % 2)) / 10;
    }
    while (cont > 0);
    printf("%d", resp);
    return 0;
}

Answer 1

It turns out that since it is storing the binary representation in an integer, it quickly passes the capacity of representation of it and generates a overflow.

Note the number 1200 in binary:

10010110000

See the same number in integer, visually formatted to simplify reading:

10,010,110,000

That’s 10 billion, and clearly out of the range of integers that goes up to 2,147,483,647.

So the only thing you need to do is change the type to one with more representation, like the unsigned long long:

int main()
{
    unsigned long long bin = 0, num = 0, resp = 0, cont = 0;
    printf("digite um numero\n");
    scanf("%lld", &num); //tipo lld
    ...
    printf("%lld", resp); //tipo lld
    return 0;
}

See code working on Ideone

Warning that however, you will have the same problem if you try to convert slightly larger numbers such as 1000000, because the binary representation grows too fast. To get around this problem only keeping the representation in a string, that can be as big as you want.