90 degree phase shifter for a sine wave with variable frequency

N

newbie_hs

Full Member level 3
Full Member level 3
Joined
Mar 4, 2023
Messages
159
Helped
1
Reputation
2
Reaction score
2
Trophy points
18
Activity points
1,253
I need to design 90 degree phase shifter for a sine wave of amplitude 1V and frequency varies from 5Khz to 50Khz I found some solutions here itself.
In all those approaches amplitude varies with frequency.

I need constant amplitude at all these frequency range.May I know any circuit ideas.
 

Sort by date Sort by votes
Wide band analog 90 degree phase shift can be only implemented differentially, two chains of frequency staggered all-passes having output with 90 degree phase shift

Phase-shifter 90 degree

Hi, I am looking for a phase-shifter circuit, that can shift a signal with 90 degrees - over a frequency range (eg. 50kHz to 200 kHz). Any suggestion? Best regards,
www.edaboard.com www.edaboard.com

Audio I/Q phasing networks, which is better?

Hi I have found two types of BIDIRECTIONAL audio I/Q (45 degrees difference) phase shift networks, which I have enclosed in red squares. I wonder, which of the two is better, in the sense that keeps the phase difference more constant throughout the 300Hz-3KHz audio region? A third method is...
www.edaboard.com www.edaboard.com

help needed for Phase Shifter conversion to single supply

I have this circuit for a phase shifter using dual supply. I want to make one for single supply. Below is my take. Any help with supply conversion and phase shift calculations are appreciated. I don't quite understand how the frequency points are chosen. This circuit, if my math is right, has...
www.edaboard.com www.edaboard.com

90 degree phase shift for a wide range of frequency ( exactly , 90 degree !)

Dear All Hi Thank you for your time . I know how can i create 90 degree phase shift for an specific frequency , but how to create it for a wide range of frequency ? Thanks in Advance
www.edaboard.com www.edaboard.com
 

Upvote 0 Downvote
Unfortunately some old attached files have been deleted at Edaboard. There are however valuable hints in the discussion, it should help you to design a solution. You didn't yet mention phase tolerance requirements.

J-TEL All Pass Filter Designer gives e.g. this approximation for 2x3 stage, I scaled frequency range down by 10 because the tool is limited to 10 kHz

1742906397718.png
 

Upvote 0 Downvote
FvM said:
You didn't yet mention phase tolerance requirements
Click to expand...
At present I don't have any info of tolerance requirements.My aim is first to make a prototype which do pashe shift of a sine wave (5Khz to 50khz with 1V amplitude) without any change in amplitude at the output w.r.t frequency
 

Upvote 0 Downvote
Passband ripple can be decreased by increasing the order of the filter (i.e. more RC stages), but some amount of ripple is unavoidable (both amplitude and phase). How much is tolerable determines the minimum order of the filter (along with the bandwidth).
 

Upvote 0 Downvote
A microprocessor approach would be to detect the signal, calculate its frequency, and then generate the same frequency square-wave with a 90° phase-shift, which is then converted back to a sine-wave with a look-up table and an A/D.
This will give a precise phase-shift independent of frequency, with constant amplitude.
Of course that only works for a single frequency (at a time) input.
 

Upvote 0 Downvote
Question is if you can live with the differential phase approach. E.g. both all-pass chains have a total phase shift of 0 to 540 degree, but the frequency dependence is different so that you achieve a difference of 90 over the designed frequency range.

Example:


A pure 90 degree phase shift (Hilbert transform) can be practically only implemented with a digital (FIR) filter and involves a relative large delay.
 

Upvote 0 Downvote
Below is the image form J-Tek

I don,t see R4 in the designer
 

Upvote 0 Downvote
@FvM ,

Design Calculation
My frequency range is 5khz to 50Khz

I made the circuit and is given below.



When the frequency is 50Khz I am getting almost 90 degree phase shift. Please see the output below.



But when the input frequency is 5Khz output is almost 180 degree out of phase.May I know why


Please see my simulation file attached.
 

Attachments

  • Phase_Shif.rar
    3.6 KB · Views: 10

Upvote 0 Downvote
newbie_hs said:
But when the input frequency is 5Khz output is almost 180 degree out of phase.May I know why
Click to expand...
I fear you didn't yet understand the concept of generating 90° phase difference between the outputs of two all-pass chains. You are measuring only one filter chain. I believe the concept is well explained in quoted links and also J-TEK description
 

Upvote 0 Downvote
@FvM ,

You are correct.I have some confusions in understanding that tool.Please see the below image.
May I know the values in the first row of filter designer corresponds to the first stage(I mean the O/P1)?
Values in the second row for second stage and third row for third stage?

If you can kindly put the component values and show here for your previous exmaple(given below) it will help me to understand more.
 

Upvote 0 Downvote
1. The filter order n is the number of opamp stages in each branch. In this case you've chosen n=3
2. The number of paths is 2, no matter what. The 90 degree phase difference appears between the outputs of the two paths.
3. Referring to the figure you posted, the values of R1 and R2 have nothing to do with the R1 and R2 values displayed by the software (a bit confusing yes). You can arbitrarily pick R1 and R2 for all stages/paths, so long as R1=R2.
4. The R and C values given by the software refer to C3 and R4 values in that schematic (again, confusing). Each row in the table is a stage in the path. The left table (R1 and C1) is for one path, the right table (R2 and C2) is for the other path. The ordering of stages doesn't actually matter (in theory).
 

    N

    newbie_hs

    Points: 2
    Helpful Answer Positive Rating
Upvote 0 Downvote
The frequency determining components (red/green marked) are correct now. Negative feedback network of each OP would simply use same resistor value for all stages, e.g. 10 k.
 
Upvote 0 Downvote
The phase difference between the two outputs is 90 degree.Please correct me if I am wrong.
How will I connect this this to my system.
My system can take only one input which is a 90 degree phase shift of the I/P signal.
Below is my system block diagram
 

Upvote 0 Downvote
This all-pass approach will not work for how you have defined your system. The all-pass approach takes your input signal and gives you two output signals 90 degrees apart. The phase relation between the input and output is NOT 90 degrees, and will vary with frequency.
 
Upvote 0 Downvote
An RC or CR will deliver 90-degree (-ish)
phase shifts without amplifier lag. Amplitude
will scale with f , an AGC block then might
be the easy way out (at some lag, on
frequency-changes). If you're fussy about
that bit (transient tracking, vs steady state
fidelity) your options will be more elaborate
I expect.
 

Upvote 0 Downvote
You must log in or register to reply here.

Similar threads