Dual switchover: one leg's pulse into the cell
When a switch leg connects the cell to the supply through its load resistor, how far does the cell charge, what current flows, and what does each pulse cost?
- U
- Cell voltage after the pulse, V
- τ
- Charging time constant, µs
- e
- Effective on-time, µs
- q
- Charge per pulse, C
- J
- Average supply current, A
- W
- Average power, W
- V
- Supply voltage, V
- f
- Pulse frequency, kHz
- R
- Load resistor, Ω
- C
- Cell capacitance, nF
LaTeX
τ = R \cdot C \qquad U = V \left( 1 - e^{-e / τ} \right) \qquad q = C \cdot U \qquad J = q \cdot f \qquad W = V \cdot J
Method
- The period is one over the frequency; the on-time is the period times the duty; the effective on-time is that less the dead time, and not less than zero.
- The cell charges through the load resistor as an RC circuit: time constant RC, and after the effective on-time it has reached V(1 − e^{−t/RC}) from empty.
- The charge moved is C times that voltage. Once per period, so the average supply current is charge times frequency, and the average power is the supply voltage times that.
- The peak current is the moment the leg closes: the whole supply across the load resistor.
Assumptions
- The cell is empty when each leg fires. In the dual switchover the other leg's pulse is meant to drain it, so the model is one charge per period; if the cell holds its charge the second pulse moves almost nothing and the supply current is far lower than this.
- The switches are ideal: no on-resistance, no drop across a diode, instant edges. A real MOSFET and diode take a volt or two off the supply.
- The cell is a capacitor and nothing else; its leak through the water during the pulse is not counted here. The step-charging calculation takes the leak.
- Meyer's claim for the arrangement is that the cell sees voltage while the resistor takes the current. The arithmetic here is what that arrangement does as a circuit; it does not speak to whether it does anything else.