The pictures show a bank of (6) 2,900 uF, 400 V electrolytic capacitors. Three of these units (as the picture is shown) are wired in series for a net 966 uF, 1,200 V; six of them would provide 1933 uF; or if you require a different power density ratio, wire them in series groups of four caps for 725 uF, 1,600 V, and parallel two of these groups for 1,500 uF, 1,600 V. The white box in the picture is the original 150 uF, 1,250 V cap (not used).
The charging system for the above bank of caps is a 1,000 W microwave oven transformer. These units are available surplus or cheaper :-) and are way capable - a typical rating being: 3 KV at .5 A!! This potential is rectified through a homemade bridge made with microwave HV diodes (12-15 KV). You also may use just one as a half wave rectifier.
DOUBLE WARNING: These transformers can be quite lethal, and when built into the power supply capacitor bank, you have just built your own electric chair should your body fall conductive to it's potential.
(Note: I would most definitely recommend - require - bleeder resistors with a visible indication of non-zero charge on *each* of the lethal caps --- Sam).
The rep rate is based on the power supply charge limits and not the cooling capability of the cavity. The unit was fired at a rep rate of once every 5 to 6 seconds for better than 5 minutes in one instance with no perceived change in power and or damage to flashlamp or rod (I have both a ruby and a Nd:YAG rod that I can interchange in 10 min.) Will the bulb take 2,500+ Joules of input? No - the limits of the EG-G FX-103 flashlamp are 1,550 Joules explosion energy and when you work up your power supply, the maximum input is calculated from: repetition rate, cooling type, and percentage of explosion energy versus lifetime (# of shots). The original flashlamp lasted through many months of play with power ranging from the rod's threshold of approximately 640 VDC up to the 1,200 VDC master pulse.
Output power? who knows. I just recently obtained a power meter and haven't played with it yet and also need a new lamp as the original one expired in a final glorious pulse. The power output is amazing even without the Q-switch. I can't imagine what this unit would do with the Q-switch timed and functioning. I also have a Hughes Q-Switch Driver but I don't have the pinout and as soon as I figure it out the Q-switch will be re-installed.
Several flashlamp manufacturer's have replacement bulbs for this unit with larger tube diameter's allowing even higher input energies. ILC has one that is still 3" arc length but 9 mm outside dia instead of the 5 mm of the original. That unit will take a 2,500+ Joule input and live. I've also thought of gold plating the cavity to optimize for the YAG rod. However, I've been told that the ruby will not likely even reach threshold in the gold cavity - the YAG would improve by 8 to 12%.
View 01 . View 02 . View 03 .
View 04 .