wcstombs producing unexpected output

Question

I am trying to convert a wchar_t* to char* using wcstombs on Linux(CentOS7 to be specific). The current system encoding is UTF-8 and the return value of wcstombs confirms that all the characters are converted. However, when I print the resulting char* string, I see unexpected values. I am pasting the sample code below. TRACE1 is a custom macro to print values.

    int length;
    int size = wcstombs(NULL,wideString,0);
    TRACE1("number of bytes needed for conversion :-------> %d" , size);
    char str[size+1];
    length = wcstombs(str, wideString, size+1);
    TRACE1("characters converted :-------> %d" , length);
    if(length == size)
    {
        str[size] = '\0';
    }

After the conversion I was expecting the char* string to have characters in ascii character set as the wideString has only ascii characters. However, I see non ascii characters in the converted char* string. wideString is something that I receive as return value from .net code and I am receiving it as a 'wchar_t' to keep it uniform across Linux and windows.I am not sure if that can have an impact here?

I have tried setting the locale to the default locale and it didn't help me. Any help is greatly appreciated. I am pasting sample wide string and output string below for reference.

Below is how the original wide character string looks like eyJ0eXAiOiJKV1QiLCJhbGciOiJSUzI1NiIsIng1dCI6

And this is how the converted string looks like ü°<92>§¥¥ý©<90><95>¡¥ý<8b><92>¦¥<8f>ý©<94><93><85><96>ý¨<92>¤<8d><8c>ý©<98>´<9d>¢ý<93><92>¦¥<8f>ü±<92><97>©<95>ý³<92><96>¥<8e>ü±<99>¶¹<89>ü¶<92><94><8d>¤ý<94><9c><86>µ<89>