# Two-Pi Winding

-April 22, 2014

This symbol has been the logo for my LinkedIn group, Analog Developments. The image is of a single transformer winding that I needed to make whose turns were arranged on the bobbin as two pi-sections.

This sketch illustrates how those two pi-sections were constructed so that the distributed capacitances to ground from the windings' turns was as equal and symmetric as possible from side to side around a center plane perpendicular to the bobbin's axis.

The project that needed this was a high voltage power supply of around two or three kV which used a bridge rectifier as shown below. The two capacitors shown at each end, C1 and C2, represent the distributed winding capacitances that are shown above.

The issue was that there were spikes appearing on the high voltage output voltage ripple. These spikes were eventually traced to the capacitances associated with the structure of the transformer winding.

If C1 and C2 are allowed to be of different values, when the transformer's secondary voltage is swinging from one polarity to the other, there is a time period in which none of the four diodes of the bridge rectifier is in conduction and the transformer winding floats. During that non-conduction time, the transformer winding is in a free floating condition in which it develops a common mode voltage in proportion to that capacitance imbalance.

Then, when a diode pair starts conducting again, that common mode voltage gets very quickly discharged from which action, a spike of current comes about and induces a voltage spike that appears as part of the output voltage ripple.

Arranging the transformer winding as shown balances the winding capacitance and therefore minimizes the spiking effect by minimizing the development of that common mode voltage.

It was an interesting exercise using the two-pi winding to move a grounded sheet of aluminum back and forth around the pi coils and watch the ripple spike amplitude vary as a function of the aluminum's placement. Regrettably, I never captured a photograph, but just take my word for it.

Of course, nothing is perfect. Attaining a zero imbalance of capacitances would be the attainment of perfection and that is essentially unachievable which means that a bridge rectifier used this way is going to result in some spiking, no matter how small.

In hindsight from this project, my recommendation is to not use a bridge rectifier at all for this kind of thing but to use instead, a voltage multiplying ladder for rectification in which case, the transformer winding is never allowed to float.