I am looking for a circuit to drive a speaker to make a fairly loud simple tone - a square wave driven from a microcontroller (Arduino/ATMega).

There are obviously a billion chips and amplifiers out there that talk about "gain" and stuff, but I want to do something much simpler. I currently have a darlington pair transistor stack which will dump a couple of amps from a 12 volt source through the (4 or 8 ohm) speaker, from the uC's digital output pin.

For a "Standard 4 ohm speaker" - I am not sure how to drive this. Is it okay to just dump a "12 volt / 3 Amp" square wave into the speaker - or should I step up/down the voltage/signal with a transformer?

If the speaker was purely resistive (which it is not) one would argue that I need to only provide adequate current so that with the 4 ohm load I am not exceding the maximum wattage of the speaker, per (W=IIR) (Where I is controlled by the transistor, R by the speaker, and W is the maximum wattage).

Is it as simple as this, or am I missing something?

  • 1
    \$\begingroup\$ That will work, in a rather ugly fashion; don't forget the 48W flyback diode! \$\endgroup\$
    – pjc50
    Apr 22 '13 at 16:05

Speakers, even bad ones need to be driven with an AC current or AC voltage - this is because there is usually a magnet that has a moving-coil wound round it. The moving-coil attaches to the diaphram that makes the sound. and the average position of the coil-diaphram is its undriven rest position. All is well when this occurs.

You are proposing to drive the coil with an intermittent DC voltage and this may work but it will also likely damage your speaker. It may damage it instantly or it may damage it gradually or you may be lucky and it won't get damaged.

Your signal is 12Vp-p but there is a standing 6Vdc voltage on this signal because of your proposed solution. This contributes nothing other than warming the coil in the speaker - this produces no sound and is wasted energy. The speaker doesn't know how to deal with this other than force its cone or diaphram in one direction or the other AND dissipate probably in the other of 12W - I remember testing an 8 ohm speaker - it had a dc resistance of 6ohms and for your 4 ohm speaker I calculated 12W on that experiment.

  • \$\begingroup\$ I would be driving between +12v and 0v. So when the sound was "off" there would be zero voltage. When the sound was "on", it would be "AC" because I would be sending a square-wave between +12 and 0. So the only "difference" between it and a "normal" amp is perhaps that my signal is only going positive - not negative. Would this be an issue? Would/should I compensate by only effectively driving it to half it's rated wattage (because I'm essentially sending a rectified half-wave)? \$\endgroup\$
    – Brad
    Apr 23 '13 at 12:49
  • \$\begingroup\$ It might work how you are driving it or it may smoke because of the dc content. You want to do something "much simpler" and I'm highlighting the risks. \$\endgroup\$
    – Andy aka
    Apr 23 '13 at 13:04

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.