555 Voltage Doubler

[spacecommandant]

I have a circuit which, with 9V, draws around 31mA. I've been using this voltage 'doubler' (image attached) with a NE555P chip to bring the voltage to 15.9V and the current to 75.5mA.
It sounds fine, no issues I can detect, but I don't have much experience using doublers so I'm curious if this looks like a good circuit to consistently use with that amount of current.
I'd likely use an SMD version of a 555 chip based upon availability.
Thanks in advance!

[luix]

Let's use a dedicaded chip like the Microchip TC1044SCPA and ceramic 10uF capacitor, the expense would be a little bit higher but since you are not in an industrial production it will not be a finacial problem.

Using a NE555 you will see the voltage rapidly drop down with load because of the high internal impedence, sorry but the net is stuffed with bullshit, you should add a MOSFET like a BS170 or some better SMD choice (lower rds) but in this way the cost will be the same as that of TC1044SCPA and the circuit will complicate.

Don't buy TC1044, ICL7660 or other chip like those on ebay chinese sellers because they are almost all fake, they work with frequency lower than 20kHz (tipically 4kHz) and you will hear the whistle out of your effects, expecially if used with hi-gain distortion.
You can buy 10uF 50V ceramic capacitors from them because they are out of tollerance in the negative range of capacitance (8uF) and I bet they are also not suitable to work with the highest voltage declared, but series resistance is OK and for 9 to 27V circuits they are fine.

[plush]

MC34063

just add pi filter on the output.

[luix]

MC34063

just add pi filter on the output.

This is not a charge pump, it's a boost converter and it will work in discontinous mode for light load, so if the load don't draw 75mA constantly (as I've assumed, NE555 wont work well with all that current) it's not a good choice, even with CLC filter.
The less is the current the higher is the inductance, you also need to shield the inductance or use a toroidal one that is usually bigger than the classical rod one.

[plush]

MC34063

just add pi filter on the output.

This is not a charge pump, it's a boost converter and it will work in discontinous mode for light load, so if the load don't draw 75mA constantly (as I've assumed, NE555 wont work well with all that current) it's not a good choice, even with CLC filter.
The less is the current the higher is the inductance, you also need to shield the inductance or use a toroidal one that is usually bigger than the classical rod one.

Yep, I know, did dozens of them in my early years. IMO it performs better than capacitive charge pumps and provides way more stable supply during transients. In fact, it can perform quite well at low loads with a quite tolerable output ripple.
As for the inductance - yep, you are also right, it will require a large primary of about 200-220uh (around 6mm in diameter, 10mm in height - a size of your typical 35v 100uf cap), which you can get pretty much anywhere for cheap. Speaking about toroids - I've did my research and came to the conclusion that it really does not emit that much. Fur sure, if I'd be an undustrial designer, I'd put shielded inductor to meet emi standards and stuff, but for enclosed audio applications in won't make any noticeable difference when using screened toroid vs unscreened radial leaded. Still, there might be some coupling via radiated emisson, but it will gradually fade around 1-2cm from the coil and won't damage anything or make something oscillate.

If was OP, I'd build something around industrial grade stuff like TPS61170, which I've used earlier. Jokes aside, 3mvpp output ripple max when boosting from 3.4 to 12 volts @1-200ma load with hardly noticeable emi signature. The only downside is that this chip, along with many others, is very demanding on pcb design skills.

[Paul_5]

I'm toying with using a doubler for a 12v supply (to provide roughly 24v) for a class D power amp that draws 20mA and use the 12v to run the Baja Orange AD200B preamp. Any reason why this is a bad idea?

[luix]

I guess that class D power amp draws 20mA only without audio signal, with load you can brutally calculate the DC current required using the power formula P = I²*R_load -> I = SQRT(P/R), for example to have 10W on 4ohm the DC current needed will be about 1.6A, stay safe increasing a bit that current.

For this application the best way to get 24V is using a Boost converter as said by plush and to regulate the preamp power supply with CLC filter before a classic LM317 regulator, CLC can be overkilled but I would prefer to design a PCB with the space for this components and than not use them.

You can choose plenty of IC controller, be sure that the switching frequency is higher than 20kHz, higher frequency means lower inductance value and so smaller item... search the net for Texas, Microchip, National Instruments or other big company papers, they are the best way to start with SMPS.

[Manfred]

Have a look at this topic, I built a voltage doubler for higher load current.
viewtopic.php?p=287089&hilit=Morley+sldw#p287089
The current compensated choke coil was absolutely necessary to avoid interference noise via the supply lines.