# Where does the current in a Dickson Charge Pump come from? Does the oscillator/PWM pin of a microcontroller also source current?

Well, I am making a small circuit which looks like this:

When simulated, the load current gets up to about 200-280 mA no problem, so I want to implement this circuit in real life. My only problem is, I want to use either an oscillator, a microcontroller or a 555 timer as the source of the PWM signal V1, but such a device can only source about 20mA, so I am wondering, where does the current through the load come from? The constant 3V source V2? Or does it also come from V1? And if V1 also sources current, how much does it source? Is it under 20mA?

• Have you tried looking at those currents in your simulation? Nov 2 '14 at 19:58
• Current comes from the steady DC input and also the PWM. The 'sharing' of current comes more from the DC input than from the pulsed input, for obvious reasons Nov 2 '14 at 19:59
• Please refer to en.wikipedia.org/wiki/Charge_pump, when capcaitor is charging, the current need sink through your pulse source, when discharging, your pulse source need source some current, too. Otherwise the 'charges' can't be pumped to the load. Nov 3 '14 at 4:22
• Yes, I did, Manjeko. I can only look at currents on circuit elements so I could only see the current through D1, and let me tell you that one looked really odd and I didn't understand a thing. Current spices with peak to peak distances > 3A, that's all there was to it. Nov 3 '14 at 5:07

It will come from both sources. I would suggest building a discrete CMOS inverter with a power N-channel and P-channel MOSFET transistor, and then driving this with the PWM signal from the microcontroller. This will alleviate the I/O pin from supplying power.

• I'd suggest a MOSFET gate driver. Nov 3 '14 at 7:34
• That's a good idea. Basically, a half H bridge is required. Nov 3 '14 at 8:07
• A single gate driver is a half bridge (= totem pole). Nov 3 '14 at 15:11