Building transformer but low inductance on primary coil

[szzuk]

I want to drive a piezo with this set up:

Amp : Transformer : piezo

Transformer is turns ratio 1:3
Piezo is capacitance of 1uF
piezo is operated at resonance of 100KHz

I'm currently designing the transformer I need to build but I'm finding I need a really low inductance on the primary coil. What should i do?

The calcs are these: Piezo capacitance refelected back to primary coil is 3uF. At 100KHz this implies I need a primary inductance of 0.8uH.

Thanks,

Szzuk.

 

[BobW]

You are confusing open circuit primary inductance with the reflected impedance of the load. These are two entirely different things.

The important question is: What load impedance does the amplifier want to see?

Also, does the piezo transducer still look like a 1uF capacitor at resonance?
At resonance I would expect it to be largely resistive.

 

[szzuk]

===============Duplication.=====================

 

[szzuk]

You are confusing open circuit primary inductance with the reflected impedance of the load. These are two entirely different things.

The important question is: What load impedance does the amplifier want to see?

Also, does the piezo transducer still look like a 1uF capacitor at resonance?
At resonance I would expect it to be largely resistive

I know that is correct but I'm not sure why. The amplifier is the OPA548 operational amplifier, it is rated at up to 60V and 5A, but there is no mention of a preferred load impedance. So I was working on the principal that the piezo capacitance should 'cancel' the primary coil inductance leaving as low a real resistance as possible.

The problem I'd say is that I'm using a gapped core ferrite and I can't build a primary coil that is 0.8uH, it would be about 1/3 of a turn even with a fairly small bobbin.

Perhaps there is something else too?

Thanks.

 

[RCinFLA]

If all you want is the resonate frequency of the transducer then make the secondary inductance close to parallel resonance with the transducer capacitance at resonate frequency.

Primary inductance is of little importance. Turns ratio squared is the load transform impedance ratio to the parallel equivalent real part of transducer in parallel with the parallel equivalent resistance losses of the secondary winding. You may have lower equiv parallel resistance in transformer secondary winding then transducer but that is okay. If this secondary parallel combination is 5k ohms then a 1:25 turns ratio will yield 8 ohm load at primary side. If near resonance at secondary there will be a 25x voltage swing on secondary compared to primary swing.

Op amps output impedance is not intended to be matched to load. It is what it will drive without giving excessive distortion and heat dissipation. OPA548 frequency response is marginal for driving 100 kHz. Open loop gain on OPA528 is only 20 db at 100 kHz. You can only set it up for maximum of 10 to 15 db of gain.

 

[crutschow]

If the transformer inductance it too high, then just parallel the secondary with an external inductance of the desired value to resonant the transducer.

 

[szzuk]

If all you want is the resonate frequency of the transducer then make the secondary inductance close to parallel resonance with the transducer capacitance at resonate frequency.

At 100KHz with a 1uF piezo capacitance the secondary would have an inductance of 2.53uH. I couldn't wind with a ferrite core for 2.53uH it would probably be just one turn, but maybe I could just not use a ferrite core?

Primary inductance is of little importance. Turns ratio squared is the load transform impedance ratio to the parallel equivalent real part of transducer in parallel with the parallel equivalent resistance losses of the secondary winding. You may have lower equiv parallel resistance in transformer secondary winding then transducer but that is okay. If this secondary parallel combination is 5k ohms then a 1:25 turns ratio will yield 8 ohm load at primary side. If near resonance at secondary there will be a 25x voltage swing on secondary compared to primary swing.

Yes I get that now. At resonance I only need to work with the real resisitive part of the piezo / secondary coil. I need to ensure the turns ratio is big enough to reduce this down to something which will ensure I am going to get enough power from the opamp.

Thanks.

 

[RCinFLA]

You probably want to use a powered iron core instead of ferrite. Work with a lower permeability core. Ferrite have way too high permeability for this application. Torroids have bigger selection in this range if you don't mind winding it.

 

[antknee]

---------------------

 

[indulis]

Or just make the gap bigger to get the inductance factor down so you'll have to wind more turns.