How to wind transformer?

[Astrid]

How to wind transformer on **broken link removed** core
Say the primary winding has 6 turns and 60 secondary turns.
What is the correct procedure, the primary coil former and over to secondary or vice versa?
Alternatively, both winding side by side?

 

[WimRFP]

There is no easy/short answer for that.
Some things to consider:
-Insulation requirements (I mean from a safety perspective if applicable, this affects insulation thickness and margin to get certain minimum creepage path)
-Leakage induction (interleaving gives lower leakage inductance).
-capacitance between prim and sec winding.
-presence of air gap (avoid thick wires close to the airgap as this increases eddy current loss in the wire).
-is it a transformer only, or does it need to store some energy (as in a flyback converter, you need airgap, or very low permeability material).
-interference from primary to secondary (make sure that the cold side of a multi layer winding sees the other winding).

Just primary and secondary on top of eachother is easiest, you have the lowest capacitance, but the largest leakage induction.

sec-prim-sec (interleaving) gives more capacitance, but significant reduction in leakage inductance. It is best that both windings have same width, you may use more wires in parallel for the primary so you get good copper coverage.

 

[FvM]

Just primary and secondary on top of eachother is easiest, you have the lowest capacitance, but the largest leakage induction.

And side-by-side, which was asked in addition, has even more leakage inductance. The geometry is sometimes used in safety transformers.

 

[Astrid]

OK for clarification
safety aspect is not interesting
on the primary are 10V pulses at 300kHz
secondary coil supplies capacitive load, and ideally it would be to form a resonant circuit
the core has a relatively large air gap

 

[WimRFP]

So you mean that you need a given leakage induction to have a certain resonant frequency with the capacitive load? Is it to transform the input pulses to a sine wave output? Can you provide us more info on the application?

 

[Astrid]

Primary will be excited either sine wave or pulse (on / off).
Secondary supplies vibration pads vibrating at 300kHz, spare electrical diagram is capacitor parallel RLC circuit.

 

[WimRFP]

Based on the short info given, just do some experiments:

Measure primary inductance at a frequency well below the self resonant frequency of the primary winding.
Do the same again with short circuited secondary.

Use Lshort/Lprim = 1-k^2 to calculate the coupling coeffiient (k).

When you also measure the secondary inductance, you can make a simple pspice model to do some simulations. With the inductor coupling you can couple 2 inductors to form a transformer with non ideal coupling. You can make the model better by adding parasitic capacitances, but that requires you to determine self resonant frequency or using a steep pulse. Measuring prim to sec capacitance is the easy one. Adding loss resistances is not easy as they are frequency dependent, so you may need to measure them at the operating frequency.

Start with your most favorite winding strategy. When k turns out to be too low (that means large leakage inductance), you need other winding strategy (such as interleaving).