Charge Pump Refinements

[Primus]

If you compare this charge pump design: http://www.musicianshotline.com/archive/monthly/pedal_power/images/0304/fet-distortion.jpg

To this charge pump design: http://www.geofex.com/circuits/voltpmp1.gif

You can see that they have a lot of parts in common. On the former, there is also a FET booster circuit to the right side of the diagram. There are a few extra componenets in the charge pump layout of the former though. Specifiically, a 10 uF cap from groun to +9V, some diode I can't read from ground to +9V and a .1 uF cap from ground to +9V. What are these three extra components for?

[R.G.]

a 10 uF cap from groun to +9V, some diode I can't read from ground to +9V and a .1 uF cap from ground to +9V. What are these three extra components for?

The 10uF and 0.1uF capacitors from +9V to ground are to ensure a low power supply impedance from the battery. It's a good idea, but not essential for the circuit to operate. The diode from ground to +9V is a polarity  protector diode, which has no function in normal operation. If (when!) you hook up a battery backwards, the diode clamps the voltage to only 0.7V and keeps the reversed battery from killing the charge pump IC.

[Primus]

Thanks, RG!. Small bear has a 1N5818 Shottky. I think you specced a 1N5817. Is this an okay replacement? I think it's fine, but I want to make sure i'm not stupid.

[R.G.]

I think the 5818 and 5817 differ only in reverse voltage specification. The 5817 is 20V, the 5818 is 30V. In this application, there's only about 9-10V across each diode in reverse operation, so either one should work.

[Sir H C]

RG-

Why would they tie Vout to ground?  That seems to be odd to me.

[R.G.]

The "Vout" pin is only the output pin for voltage inversion applications. Grounding this pin is right out of the Maxim 1044 application note.

What you are doing when you use the 1044 as a doubler is alternately filling a capacitor from +V and ground, then connecting the more-negative terminal to V+, which pushes the charge in the capacitor above V+. The nominal Vout pin for voltage inversion does not play a part in this.

It is also possible to both double the input voltage and invert it at the same time with the 1044, as shown in their application note. This gives you +2 times the input voltage and -1 times the input voltages, minus diode drops.

[Sir H C]

Cool.  Was thinking more about it, and I guess grounded you don't get the miller effect on those devices so you save some charge/discharge current.