Getting ±3V from an AA battery

[microbug]

I'm building a small headphone amplifier. The op-amp needs at least ±3V, and the power supply is a single AA battery (0.8-1.5V). Suggestions? TI and LT's automated systems can't find anything. The whole system will only use 20mA max so a single 6V rail with a resistor divider for input bias is an option.

If this is a lost cause I could go for a lithium-ion battery or 2 AAs, but I'd like to keep it compact and the limiting factor in that is the battery size.

Thanks in advance!

[IanB]

I have a bunch of cellphone boost chargers that produce a 5 V output from a 1.5 V cell, but these really want to work with a lithium cell not an alkaline. I could maybe see what chip they use and let you know. I think it's quite a stretch getting 6 V from an AA battery especially down near the 1 V end of the scale. Only needing 20 mA should help, though.

[uncle_bob]

Hi

A very common solution is to run the op-amp off of +6 rather than +/-3. AC couple in and out. You generate a "fake ground" at +3 either with another section of the op-amp or just with a couple of resistors.

Bob

[microbug]

Thanks Bob. If it's possible to get up to 6V on a single AA cell then I'm all ears, but as IanB said it seems unlikely or at least difficult. I'd like to keep this thing fairly efficient too; seems like a dual AAA solution is the way to go (one 3V boost converter per cell with the negative terminals connected). Shouldn't be too bad for size.

[IanB]

Possibly you could use a pair of back to back lithium CR2 photocells? They are smaller than AA cells, would need no boosting, and would have a really long life.

[microbug]

That's great, thanks IanB! I'll do that.

Edit: Having thought about it a little, I changed my mind: I'll do a fake ground solution with an 18650, since I have some lying around. They boost to 6V easily. That also means I don't have to buy more CR2s when they run out.

[uncle_bob]

Thanks Bob. If it's possible to get up to 6V on a single AA cell then I'm all ears, but as IanB said it seems unlikely or at least difficult. I'd like to keep this thing fairly efficient too; seems like a dual AAA solution is the way to go (one 3V boost converter per cell with the negative terminals connected). Shouldn't be too bad for size.

Hi

You can easily build  / buy / design a boost converter to take an AA up to some crazy voltages. A lot depends on how much current you need. There are micro power solutions and there are not so micro power solutions.

http://www.linear.com/product/LT1613

Is a randomly picked one of many solutions.

Bob

[microbug]

I added symmetrical charge pumps to the boost converter in my design, as suggested by blueskull. You can find the most recent design here in my latest post on the main thread for my headphone amp.

[chris_leyson]

SEPIC plus Cuk is good for balanced supplies, Analog Devices AN-1106 http://www.analog.com/media/en/technical-documentation/application-notes/AN-1106.pdf
You could use something like LTC3539

[microbug]

Looks interesting - I'll save that app note. If the symmetrical boost works, I'll stick with that for now, for the sake of simplicity and fewer inductors.

[tsaG]

Start at about 21:30. Maybe this could help you.

[microbug]

Thanks, that's useful. The bipolar supply should do it for this project but it's useful to know about inverter chips.

[Audioguru]

On another forum recently, an old ICL7660 voltage converter was used to make a -3V supply for a headphones amplifier but it produced a nasty whine noise because it switches at 10kHz.
Maxim has their newer MAX660 voltage converter that has a pin that can increase the switching frequency to an inaudible 35kHz to 80kHz.

[microbug]

Yes, after watching Dave's video on the MC34063, alarm bells started ringing when he called the 7660 'classic'. Good to know that there are better alternatives - I might just use that MAX660.