Using battery to power logic circuits, feasiblity and pitfalls?

[MuriTheMyth]

Hi folks,

I've been working on my project that involves a 100V 20A high-power part, which were driven by +5V logic circuits (plus a -5V rail for op-amp).

Initially I was planning to draw power from the 100V rail and pass through a buck converter followed by LDO to power the logic devices. The -5V is generated by a charge pump followed by LDO.

Though, as a initial prototype, I felt like the converter part is too complex. I'm wondering if it's possible to power the logic rails with lithium batteries. I will need 500mA for the positive rail because gate driver draws, so AA batteries may not work well.

I proposed a topology in the attachment. Please let me know your thoughts on it. Also I would appreciate any possible pitfall you noticed.

Thank you very much for your time!

Regards,
Anyu

[ledtester]

I will need 500mA for the positive rail because gate driver draws, so AA batteries may not work well.

It sounds like the cells powering the positive rail will get depleted faster than those powering the negative rail.

How much current do you need for the -5V rail? If it's not too much perhaps you can generate the negative -5V using a charge pump or a small DC/DC regulator. Then you would only have one battery pack with only one charging/protection circuit.

[MuriTheMyth]

That could be done! Thank you for the advice!

[pcprogrammer]

Why not use a switch mode power supply fed from the 100V rail. Most of these work from 80V to 260V AC, but can also be fed with DC. A simple USB phone charger could be used for this. The negative 5V can be generated with an additional DC-DC converter.

Or search for a SMPS that has both +5V and -5V as an output.

Will be more reliable than using batteries, which need to be charged once in a while, and with 500mA current drawn this might be rather often.