floating point range

Question

Plz xplain

Answer 1

(a) Largest positive value:

This happens when all significant (mantissa) bits are 1, and the exponent bits are all 1.

$=(b)^{p-1} -1\times (b)^{X-q}$

Smallest positive value:

This happens when the LSB of mantissa is 1 and the exponent bits are 0 so that subtracting the bias gives the largest negative value possible for exponent
$=1 \times (b)^{-q}$

Comments

I have doubt regarding Floating point repsentation i cannot understand when we take the bias to 2^6=64 normalisation is said to occur when the number is in the form of 0.1bbbbb but normalisation is said to occur when bias is 63 then normalisation is said to occur when the number is of the form 1.BBBBB 

This is from the solution maual to stallings 

also many youtube videos also say normalisaton of fp is reprsentaion in 0.11111 form 

see this https://www.youtube.com/watch?v=ji3SfClm8TU

could you plz expain 

---

@Arjun sir I am not able to understand why $(b)^{p-1}$ why not $(b)^{p}$ ? Can you kindly explain?