John,
The two equal valued capacitors connected in parallel is keeping me awake. I recall seeing a reference to that many years ago but at the time did not give it much consideration so I guess now is the time I must pay the cost! I want to address you question concerning capturing the energy that is stored on the first capacitor without being forced to accept a loss of half due to the voltage dropping by a factor of 2 after connection. You were wondering if the lost energy could be reclaimed and I have found a simple way to keep most of it. All you need is two switches and one inductor to handle the task. Place one switch in series with the charged capacitor output terminal leading to a series inductor of value "L". The other end of the series inductor is connected to the input terminal of the discharged capacitor. The ground terminals of the capacitors are connected directly together. The second switch (referred to as parallel switch) must be placed at the junction of the first switch (referred to as series switch) and the series inductor with its output terminal grounded along with the bottom of the capacitors. Begin with both switches open so no current flows in your system. Close the first switch so that current begins to build up in the series inductor. Allow the inductor current to increase until the energy contained within the first capacitor is 1/2 of its original value. At this time the voltage across the first capacitor will be .707 times the original starting value. Now open the series switch and close the parallel one simultaneously. Allow the energy stored within the inductor to transfer to the second capacitor until the inductor current instantaneously reaches 0 amps. Now you open the parallel switch and can close the series one if you like. A small transient will most likely occur due to voltage miss balance between the capacitors. Both capacitors will have an energy of 1/2 the original amount stored in the first one minus the losses due to I*I*R heating when the system settles. I assume you are mainly interested in a theoretical system in response to your question. We could go into details if you wanted to actually build one. Dave
