Actions
Forum Thread Details
  • State Verified Answer
  • Replies 8 replies
  • Answers 1 answer
  • Subscribers 291 subscribers
  • Views 908 views
  • Users 0 members are here
Related

Impedance matching of a transformer typical solutions

Eavesdropper
I need to match impedance with the primary side of a transformer. The electricity applied to the transformer will change in frequency, so I need a way to keep the impedance matched.
 
As you know,  a transformer coil is like inductor, and a capactor was the best solution. I did this before with Thomas Edison style trial and error, and eventually found a capacitor to do the job. But I want a more calculated way to do so. Any thoughts, experiences, formulas?
 
E
  • 0
    Eavesdropper,
     
    In principle, the impedance at the primary side of a transformer will depend on the load at the secondary side. An ideal transformer will not show any inductance or capacity.
    It will just transform the impedance of the load with the root of the ratio of turns.
    An inductive or capacitive load will show as an inductance or capacity at the primary side.
    It is better to compensate a non resistive load at the load side than at the primary side of the transformer. Otherwise the powerless components of the signal must also pass the transformer.
    The core of the transformer must be suitable for the frequency.
    At high frequencies, transformers will show leak inductance and capacity between the turns of the coil. A way to compensate for capacity between the turns of a transformer is the use of transmission line transformers. A transmission line has inductance and capacity, but will behave as a resistance when used at the characteristic impedance. For further reading see http://www.home.earthlink.net/~christrask/TraskTLTTutorial.pdf or
  • 0 in reply to jvdberg@ieee.org

    JV,

     


    Thank you for the reply. What I am trying to do is get the maximum energy transfer from the source through the transformer. I will definitely look into transmission line transformers.

     


    Right now I am losing a lot of power in heat. I want to avoid this as much as possible.

     


    E

  • 0 in reply to Eavesdropper

    Hi Eavesdropper,

     

    How much does the frequency vary?As Jvd mentioned it would be better to match the impedance on the load side and you could do that using passive inductors and capacitors. A simple PI-network can allow you maintain a relatively well matched impedance over a frequency range, in that way lowering your power losses. For this to work that frequency range should be relatively narrow maybe 10% of the center frequency.

     

    Let me know if you can provide that info and I'll setup a simple schematic with some formulas for you.

     

    Best Regards,

  • 0 in reply to Jorge_Garcia

    Jorge,

     

    Not only does the frequency vary, the voltage  will as well. The Load side impedance is match, no issues there. But I  want maximum power transfer from the source side.

     

    Taking the left side as a stand along circuit, I have a varying voltage/freq source directly coupled to an inductor. I am losing power for sure.

     

    Thoughts?

     

    Thanks.

    E

  • 0 in reply to Eavesdropper

    Hi E,

     

    You can't have maximum power transfer from one side of the transformer and not the other. Remember that transformers are impedance converters, as such the impedance on the primary is a scaled version of whatever impedance is seen on the secondary, if you don't have a match on the primary you don't really have a match on the secondary. This is true if you are using a transformer that can be modeled as close to ideal.

     

    You haven't given many specifications, so I'm kind of giving you general guidelines. The varying voltage is not as critical an issue as the varying frequency, remember that the reactance of capacitors and inductors varies with frequency so a match at one frequency will not result in a match at another.

     

    In order to get an optimum match you need to categorize your source's impedance and your load's impedance within the frequencies of interest. If you can do that then designing for an optimum match will be trivial and I can help you out with that.

     

    If it's not super critical to get the optimum then you have to determine what type of load is seen at the transformer primary is it inductive or capactive. In most cases the load is inductive, so we are going to add capacitance to achieve max power transfer, this procedure is empirical but I think it's better then brute force trial and error.

     

    You'll need some means to measure the power entering the transformer as you perform the experiment.

     

    First measure the power entering the transformer

     

    Add a properly rated capacitor of some suitable value (1-10uF for low frequencies, 1-100n for high frequencies)

     

    Did the power increase? If it did then double the capacitor value and test again. If the power increases again then again double the cap value and measure. Continue this procedure until you see the power drop. Once you run into this situation the ideal compensation cap value is between the value you just tested and your previous tested value.

     

    If the power decreased? That means you added to much compensation capacitance, halve the value of the capacitor and try again you should see the power entering the transformer increase, continue halving until the power dips. Once again the ideal compensation cap value is between the value you just tested and the previous tested value.

     

    Once you have a range you can perform a similar procedure to zero in on the maximum power point. This is basically a binary search you're still brute forcing it a little, but the force is a little more directed.

     

    If the procedure is not clear let me know and I'll draw some diagrams for you.image

  • 0 in reply to Jorge_Garcia

    Jorge,

     

    That is quite extensive reply, thank you so much for the information. I tried the capacitor matching, I am getting closer to my goal. I need more caps. I agree it is a little brute force, much like how Edison experimented, but that is fine with me.

     

    You are an asset to E14.

     

    E

  • 0 in reply to Jorge_Garcia

    Jorge,

     

    I am going to start doing some capacitor matching soon. I found a transformer that may be ideal for my project. More soon...

     

    E