Need Help About Multibit Phase Shifter

Hi,,

I want to design a 4-bit High/Low Pass Phase Shifter with lumped element (L-C)

for all i know 4-bit phase shifter i will produce 16 difference phase shift that is : 0, 22.5, 45, 67.5, 90, 112.5, 135, ... , 337.5 (degree)
and also the 4-bit phase shifter consist of 4 phase bit = 180, 90, 45 and 22.5 (degree), and each phase bit will have 2 phase state
and usually the phase state is consist of phase 0 (degree) as phase reference usually at high pass branch and other value of phase in low pass branch
to create phase difference : ΔΦ = Φ high pass - Φ low pass ; the phase difference is represent the phase bit.

So the problems are : when i'm using 0 degree as my phase reference i couldn't find an appropriate value for my lumped elements but i can satisfy the value of 16 difference phase shift. But if i change the phase reference into any value i can get the value of the lumped elements but i couldn't get the value of 16 difference phase shift.

so anyone has an ideas how to solve this problem...

thank you

I don't know what configuration and how many poles you use for the phase shifting reactive components, but have to know that a single pole RC or RL can give a maximum 90-deg phase shift. For 180-deg you need at least two poles, or one L and one C. In real life designers use a Pi network C-L-C to get a certified maximum 180-deg phase shift.

Hi Monolothics,

I don't understand. Can you clarify where is the problem, maybe giving an example?
Regards

Z

maybe this can explain what i want to asked :

from that picture
usually the reference is 0 and delay line is the value of phase bit itself
for example :
90 phase shift can be achieved by 0 at the reference and 90 at delay line

but 0 phase shift is hard to achieve especially using HPF/LPF phase shifter type

so from information that i get,
we can use -45° and 45° to achieve 90° phase shift ; or -90° and 90° to achieve 180 phase shift ; or -25° and 20° to achieve 45° phase shift

but what confuse me is, when those 180 , 90 and 45 phase shift is cascaded into 3-bit phase shifter i can"t achieve phase increment of 3-bit phase shifter according this table :

so this is my problem
i hope you can understand it..

Hi Monolothics,

If your 3-bit word is composed by b2,b1,b0 (b2 is the MSB and b0 the LSB), then:

b0 switches +/-22.5° (b0=0 inserts -22.5°; b0=1 inserts +22.5°)
b1 switches +/-45° (b1=0 inserts -45°; b1=1 inserts +45°)
b2 switches +/-90° (b2=0 inserts -90°; b2=1 inserts +90°)

Of course, there is a constant 157.5° phase shift, but what cares is the phase increments between states, so it is no problem.
If this constant phase shift your concern?
Regards

Z

Hi, zorro thanks for the reply.

Yes that exactly what concern me.

So, it's no problem if i have 157.5° phase shift in 3-bit phase shifter? but can you explain it why it's not a problem?

thanks

It's like a piece of cable. You will always have a constant shift in the connections (inside or outside the shifter).
Phase variations is what matters.
For example, when used in a phased array, a constant phase common to all the channels has no effect. It is the phase difference between elements of the array what makes beam steering.
Regards

Z

okay,,i understand now
thanks for your answer zorro
it's really helpful