I am putting a 9v powered circuit into an electric guitar. It will be powered by a 9v power supply through a cable, and also have a backup 9v battery on board the guitar. I am wondering if I could run the power supply and battery simultaneously in parallel, to accomplish 3 main goals:

  1. Although the power supply has filters, it needs to be as ripple-free as absolutely possible, as any ripple will be audible as noise. will the battery help to smooth the power supply's delivery? This is the main goal.

  2. I am hoping the power supply will keep the battery charged, will this be the case?

  3. I want to simplify switching and avoid having to manually switch power supplies.


Edit: essentially, I am trying to replicate a car charging system, where the alternator and battery run in parallel - the alternator float charges the battery, while the battery smooths any voltage spikes from the alternator, and if you disconnect the alternator the battery will keep the system running.

  • \$\begingroup\$ A battery has infinite capacitance so it can be used as a ripple remover. If the battery is a rechargeable one and the main supply voltage is higher than battery's then yes, you can. If not, putting series diodes to both rails will be safer. \$\endgroup\$ – Rohat Kılıç Jun 11 '18 at 3:59
  • 2
    \$\begingroup\$ No, no, and no. Battery backup circuits are far more complicated than that. \$\endgroup\$ – Chris Stratton Jun 11 '18 at 4:30
  • 1
    \$\begingroup\$ I would not use a battery as a ripple remover, \$\endgroup\$ – Voltage Spike Jun 11 '18 at 4:42
  • \$\begingroup\$ Can you tell me what the problem is with using a battery as a ripple remover? I know you can sometimes do it with car batteries in automotive applications. \$\endgroup\$ – Kendric Jun 11 '18 at 4:46
  • \$\begingroup\$ @Kendric yes, we do it that way in our factory during the tests. We put two series-connected car batteries right across the output of the lab-type linear power supply (27.8VDC output) when we test especially audio circuits and MCU-based products. We also use some extra circuitry to generate spikes (up tot 0.5 times the supply voltage) on supply rail for testing and the battery is always connected across the output. We've never encountered any problems for years. I'm also sure that a car battery is connected in that way in a car (i.e. alternator -> regulator -> battery). \$\endgroup\$ – Rohat Kılıç Jun 11 '18 at 5:18

Forget the recharge nonsense. Use 4 each 9 volt batteries and sum them together (in parallel) with 1N5817 Schottky diodes with low Vdrop.

Keep plenty of fresh batteries on hand and replace weak ones before each gig.

Buy a dirt cheap voltmeter to check each battery. The diodes will isolate each battery from the other, but sum their currents for a long life about 4 times normal.

  • \$\begingroup\$ This is an interesting suggestion and I may use this setup if the power supply is too noisy. However, I am using a very power-hungry circuit and want to minimise the expense of batteries if possible. \$\endgroup\$ – Kendric Jun 11 '18 at 5:45
  • \$\begingroup\$ You can recharge the batteries. Having multiple (packs) if batteries that you can switch quickly and recharge externally should be a plus. \$\endgroup\$ – Goswin von Brederlow Jun 11 '18 at 13:37

While this is not the answer to your explicit question, I must address the larger question: how can I minimize noise in my circuit? Yours is a good example of an xy problem. You need to do x, you've settled on y as a solution, and want help in doing y. What you really need is to address x.

And the most basic answer I can give is: get somebody who knows how to design amplifiers (or whatever your circuit does) have a go at fixing yours. There is simply no good reason nowadays for power supply ripple to show up in an output.

As an alternative answer: Get somebody who knows something about power supplies provide you with a new 9-volt power supply. One with negligeable ripple. Trust me, it's not hard to do.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.