PWM is is basically a switching regulator without the inductance and capacitance required to smooth the output to DC. For a purely resistive circuit (like an atty), you'll get a square wave looking amperage which gives rise to the same problem as the voltage. Neither the current nor the voltage is DC. Another complication comes up when driving a highly inductive load like a stepping motor. In that case, the motor's large inductance smooths the amperage waveform and it looks more like a clipped AC wave. Pretty difficult to come up with a number for an equivalent DC current. About the only way to do it is with an integrating measurement, which is what a true RMS meter does.
For devices like the eGo, they use a lookup table programmed into the micro-controller to figure out the "duty cycle" for the output. They don't' actually measure the output, only battery voltage. For closed loop regulation, you can do the necessary integration with a Type I filter network. Not something easy to design. It would be easiest with with a micro-controller as opposed to a 555 based circuit. Personally, I would do it with a micro-controller and keep it open loop, similar to how the eGo does it, but with the addition of variable voltage and two cells. Though it would require a 3 dimensional lookup table which might be difficult to code, haven't done one before.
So far, I've been perfectly happy using boost regulation in my mods. PWM is about 5% more efficient, but about 1000% more difficult to engineer, at least for one that automatically regulates to a specific power delivery. There are simple ones that regulate open ended, but dont' compensate for changes in battery voltage. In other words, as the battery drops, so does the output. You have to continually adjust the output as the batteries discharge.
For devices like the eGo, they use a lookup table programmed into the micro-controller to figure out the "duty cycle" for the output. They don't' actually measure the output, only battery voltage. For closed loop regulation, you can do the necessary integration with a Type I filter network. Not something easy to design. It would be easiest with with a micro-controller as opposed to a 555 based circuit. Personally, I would do it with a micro-controller and keep it open loop, similar to how the eGo does it, but with the addition of variable voltage and two cells. Though it would require a 3 dimensional lookup table which might be difficult to code, haven't done one before.
So far, I've been perfectly happy using boost regulation in my mods. PWM is about 5% more efficient, but about 1000% more difficult to engineer, at least for one that automatically regulates to a specific power delivery. There are simple ones that regulate open ended, but dont' compensate for changes in battery voltage. In other words, as the battery drops, so does the output. You have to continually adjust the output as the batteries discharge.
Last edited:
