An early Australian solid state portable receiver manufactured by AWA used a 30 ohm centre tapped speaker that was directly driven by a push pull PNP germanium output stage. The orthodox arrangement of the day was to use an audio output transformer which increased cost and reduced sound quality.

In these old times PNP/NPN output stage were unobtanium. Matching two transistors of the same type is still much easier today. This direct drive push pull scheme could give very low distortion.

Why is this not done today? Is a centre tapped voice coil difficult to manufacture? Does the reduced copper utilisation affect efficiency much when considering other factors like output transformer losses?

  • \$\begingroup\$ I haven't seen what you are describing before. But it reminds me of the old power supplies using a vibrator relay to create the necessary plate voltages used for ham radio equipment that would be placed in the trunk of a car, back in the day. May I assume the speaker's center-tap was grounded (probably not tied to V+, given the PNPs?) \$\endgroup\$
    – jonk
    Nov 28, 2019 at 4:41
  • \$\begingroup\$ In germanium-pnp days the circuit ground was commonly the + terminal of the DC power supply. An example: audioservicemanuals.com/g/grundig/grundig-n-z/grundig-solo/… But the mentioned center tap speaker really would be usable in this type push-pull output stage. Note: No feedback to linearize the output stage => quite trashy sound for listening music. And without an output transformer such feedback even couldn't be inserted. Emitter follower output stage would be different case. But the 1st transformer should generate the needed voltage level. \$\endgroup\$
    – user136077
    Aug 28, 2022 at 21:05

1 Answer 1


Does the reduced copper utilisation affect efficiency much when considering other factors like output transformer losses?

Yes. With a center-tapped coil only one half is utilized at a time. To get the same number of turns in half the space the wire must be thinner so its resistance is higher, requiring more voltage for the same current and cone movement. In most speakers the voice coil resistance is not much lower than the nominal impedance. That resistance wastes power as heat, so a center-tapped speaker would probably be about half as efficient as a normal one (so perhaps 0.5% instead of 1%).

The resistance of an audio transformer winding is usually only a small fraction of the load impedance, so the reduction in efficiency due to having a center-tapped primary on the output transformer should be much less than using a center-tapped speaker.

Is a centre tapped voice coil difficult to manufacture?

It certainly would be more difficult than a single coil. An extra terminal, extra wire going to the cone, and a way to join it to the center tap. Winding and assembly would be more difficult and reliability could suffer. Having to manufacture special center-tapped versions of each speaker could reduce economy of scale and complicate inventories.

Another disadvantage of a center-tapped speaker is having to run 3 wires to it. This is a problem if you want external speakers (50% more wire, 3 terminals instead of 2, possibility of wrong connection) or a headphone jack (would need special 'center-tapped' headphones, or a separate headphone amplifier). All-in-all a lot of complication just to eliminate one or two transformers.

Why is this not done today?

With modern ICs, producing push-pull output without transformers is no problem. 'Quasi-complimentary' designs that do it with NPN output transistors have been made for many years. A class D bridge amp can produce more than 4 times the sound output on the same voltage with much higher efficiency, takes up less board space and doesn't need a big heat sink - or a center-tapped speaker.


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.