RCKicker

Design Overview

We went with a solenoid driven kicker (will explain why eventually). Because energy in capacitors is proportional to the square of the voltage, high voltage systems are advantageous. However, high voltage, high power systems are dangerous, and it's difficult to get high voltage out of 9.6V NiMh packs. There is, however, other applications for high voltage, high power from low voltage; camera flashes. However, suitable photoflash capacitor chargers do not exist (we have higher power and shorter charge time requirements than a common camera flash).

We based our calculations on the following: (old calculations - but still kinda right)

.05 kg -- mass of golfball

35 mph -- final ball velocity

5.62 J -- energy to shoot the ball

If we assume 10% efficiency from the capacitor to the ball then

we need 56.2 J in the capacitor. So we need

1244.4 uF capacitor (E = .5*C*V^2; running at 300V)

Its current configuration has two packs of three 150uF caps in parallel giving us 900uF. We are considering adding another pack of three bumping our capacity up to 1350 uF.

Enter the Linear LT3570 capacitor charger. It is discrete, meaning it can drive (almost) arbitrarily large currents through its flyback topology simply by increasing the capacity of the FET. In it's present configuration we draw 4A on the primary giving us a theoretical charge time of less then 3 seconds (if we add another cap pack we anticipate the charge time to increase by 1.5X to 4.5 secs).

We still must deliver that power to the solenoid. FETs are not suitable because of their relatively high on resistance. SCRs are obnoxious to work with. High power logic level IGBTs are an ideal solution. Driving the IGBT is another problem, the gate of the IGBT had to be pulled high then asserted low. An IGBT driver was necessitated. However, most high power IGBTs require 15V to turn on. Ours would fry if we drove it that high. After weeks of searching a suitable part was found.

(I will add part numbers later)

Downloads

File:RCKicker.zip