2

I am working on a single phase PLL (phase locked loop) and I would like to make a phase shift by using orthogonal signal generator non frequency dependent. I have found many method like transport delay, inverse park transformation, Hilbert transformation, or Second Order Generalized Integrator (SOGi) but none of them fit my model. i have to shift alpha which has the same amplitude and phase to the grid voltage to 90 degree by using non-frequency dependent method. I am working in a PLECS program

enter image description here

I have tried to draw the circuit which is answerd, first after the loop filter I have got a phase shift with 180 then when I connect to flip flop I got zero output. please can you tell me where is my mistake. I have found another answer which is connection between two op-amp first shift by 180 then by 90 degrees but the problem when I apply very high frequencies i don't get a 90 degrees phase shift. please find attach the pictures of two models enter image description here

Fadi
  • 25
  • 7

1 Answers1

2

Use a master oscillator running at four times the grid frequency, and divide it down to 4 quadrature signals. Lock one of them in phase to your reference. Another will then be at exactly 90 degrees to it.

schematic

simulate this circuit – Schematic created using CircuitLab

A lot of this can be integrated into a 4046 IC, the phase detector and clk oscillator and parts of the loop filter. All you need to add is the /4 shown as the two D registers. An HC74 would be the ideal component for this.

Neil_UK
  • 166,079
  • 3
  • 185
  • 408
  • Thank you for your reply, I haven't used a master oscillator before, you meant should I insert an amplyfier with resistor and capacitor in parallel and resistor in series, then i have to divide the signal by 4 and get a phase shift 90 degrees. for me it is not clear. – Fadi Feb 06 '16 at 13:54
  • dear thank for your answer, I have tried to draw the circuit you suggest me, first after the loop filter I have got a phase shift with 180 then when I connect to flip flop I got zero output. please can you tell me where is my mistake. I have found another answer which is connection between two op-amp first shift by 180 then by 90 degrees but the problem when I apply very high frequencies i don't get a 90 degrees phase shift. please find attach the pictures of two models [![enter image description here][2]][2] – Fadi Feb 12 '16 at 14:22
  • @Fadi It would help if you drew the circuit the same as mine, then you might get the same behaviour. – Neil_UK Feb 12 '16 at 17:55
  • 'Similar' is not close enough. It needs to be identical. Next question? – Neil_UK Feb 12 '16 at 21:07
  • i have copy your circuit, what is the wrong i have done in the first picture, hope u don't be nervous from the same topic but i dont know if the phase detector is presented like i have done plus loop filter thank you – Fadi Feb 14 '16 at 22:31
  • No, I'm not going to waste my time hand-holding somebody who is not careful enough to copy a diagram accurately, and then can't be bothered to trace out what the difference is. – Neil_UK Feb 18 '16 at 14:08
  • ok sorry for bothering you. i copied but i have no account here to open loop filter and phase detector. anyway thank you for your help – Fadi Feb 18 '16 at 15:15
  • can i ask you another question : I would like to test an real inductor. it consist as you know from parasitic capacitance and resistance. the problem i face when i apply a grid voltage to the inductor i get error so i have to insert a resistor in series with the inductor but there is a voltage drop on the resistor. please can you tell me how can i measure the inductance of the inductor.i have to apply a fixed frequency because the parasitic capacitor and resistor are variables so i just want to compare first with fixed frequencies, several values. is it possible to do it in a voltage divider ? – Fadi Mar 05 '16 at 11:46
  • I'm not sure I understand your question, or your facilities. However, how to measure an imperfect component. Put in series with a resistor. Measure the three voltages across resistor, component and supply. Change the resistor value and repeat, and again. Now write down the equation for the voltage division for an arbitrary resistor and your model of the component. Fit the parameters. It if won't fit, you need a more complicated model. Repeat. – Neil_UK Mar 05 '16 at 12:33
  • thank you for response, i am working to test bench inductor and see the effect at high frequencies.the model of the inducotr consists in one component (inducotr, variable resistor and variable capaciitor with the freuqnecy). in the datasheet i have they only applied the net voltage to the inductor parts as one component and then find the impedance plus inductance. in my example i have to insert as you said a resistor in series with the inducotr (three components) but i am not sure if i do the right this which is applied the whole net voltage to this. i will try your solution , thank u – Fadi Mar 05 '16 at 20:02