26

Back in the early 2000s, I remember that, what were at the time older, PC cases had actual loudspeaker instead of the buzzer, which was popular as the PC speaker on the PC cases of the time period.

I'm interested in finding out the actual technical specifications for the speaker, like for example, impedance, power rating, sound pressure level (assuming it was specified) and so on.

  • I don't think buzzers were ever used. A buzzer vibrates on its own when a voltage is applied, where as all PC speakers have been actual speakers driven by a square wave generated by the CPU. The only variations I know of are adding volume control (e.g. Amstrad PC1512) or diverting the square wave to the sound card for mixing. – user Jun 9 at 13:16
  • Perhaps you mean when did they switch from a full size speaker mounted on the case to a one mounted directly on the motherboard. My guess would be when sound cards became common, probably some OEM that shipped a sound card. – user Jun 9 at 13:19
  • @user No, there are self-oscillating buzzers and there are buzzers which expect AC input waveform, and no, I really mean when the switch was made on actual cases. There was a period, where the motherboards didn't have a buzzer, and the PC cases did have one. In fact, I never had a motherboard with a buzzer, I skipped that generation, and posted the question when I was looking for a buzzer for my buzzerless motherboard on my new computer. – AndrejaKo Jun 9 at 17:58
28

The genuine IBM-PC Speaker (Part#8529143) is rated as 8 Ohm 0.5 Watt. Back then a very common small speaker for cheap radios, cassette recorders and alike. I guess still easy to come by today.

It was driven by a SN75475 (or MC1472 see here at p.5-42) line driver (*1) capable of providing up to 300 mA which equals to a static load of up to 1.5W. The 33 Ohm series resistor limits current to 150 mA in case of shortcut or ~120 mA with speaker, allowing an effective output of 0.55 Watt, so still (basically) within spec.

enter image description here


*1 - While the basic PC design is quite crappy, individual parts are well made, not at all cutting cost. The IBM solution is quite reliable and safe. The Apple II speaker in contrast is driven by a Darlington transistor prone to burn out or unsolder itself - heck, I even managed once to kill its equivalent within the IOU of an Apple IIe.

| improve this answer | |
  • How did you kill the circuit in the IIe? – Spud Jun 6 at 22:52
  • @Spud Don't ask me. happened 3 times. Twice on a II+ and once on a IIe. I never really figured it out. The fantastic part about the IIe is that the whole rest of the IOU still worked fine. Just no sound :/ – Raffzahn Jun 6 at 23:29
  • That's crazy! They are fun machines in that way. Others have more fancy screen capabilities, but the Apples always made great dev workstations. – Spud Jun 7 at 1:38
  • @Spud Oh, no doubt. My II+ is not only still set up and operational, I do use it from time to time to play with (which is BTW the one were I had to replace the transistor twice). – Raffzahn Jun 7 at 1:41
  • 1
    This basic circuit is also used by the contemporary raspberry pi for audio. Note that the speaker provides some intrinsic analog filtering that would smooth out the square edges of the PWM, raspberrypi.stackexchange.com/a/72599/71180 – crasic Jun 8 at 22:42
7

https://www.allelectronics.com/item/sk-85/2.25-8-ohm-0.5-watt-speaker/1.html

They were very similar to the speaker I linked. These paper cone, 8 ohm quarter to half watt speakers were found in small portable radios, computers, such as the Apple 2 and IBM PC, and some game consoles that did not deliver sound through the display.

http://www.applerescueofdenver.com/products-page/case-internal-parts-apple-ii/apple-iie-speaker/

That's a period correct one seen in many devices, including personal computers.

You may just want to get one and more fully characterize it.

| improve this answer | |
  • 1
    I don't think it's necessary or useful to try to "fully characterize" a speaker as one would e.g. a hi-fi loudspeaker. I think the primary significance of "8 ohms 0.5 watts" was that it takes about 2 volts to drive and can tolerate 250mA [0.25 amps squared times 8 ohms yields 0.5 watts]. I don't think the people building speakers into computers of that era generally made much effort to ensure any kind of consistency. – supercat Jun 8 at 16:05
  • This really depends on what the OP wants to do, or understand. I agree with you, given these speakers and most use cases. But, they were asking for response curves basically. (sound pressure) So, I was offering benefit of the doubt commentary. If it were me, I would sweep one and work from there. It's not too hard to do. And yes, these little guys vary a lot! We need more detail on the goal to speak better to any of it, IMHO. – Spud Jun 8 at 21:00
6

I still got a few of them in here. They all are:

8 Ohm 
0.5 Watt

Usual Identification: YD58-1B or YD58-2B

Here first image I found with google (source):

speaker

Its a wide-band (general purpose) speaker. They where popular back in the days however pocket radios and stuff usually used 0.25 Watt version. In all the PCs I saw there was the 0.5 Watt versions used.

The fact that the speakers where wide-band allowed the PC to reproduce PCM sound playback (using nasty PWM technique) even without the sound card present. For more info see

The buzzers have much worse audio properties and are not very good for sound reproduction they good for beeping in specific ranges of frequencies instead. They especially bad for human voice reproduction (when PCM to PWM playback is used so my bet the higher frequencies are problem). I tried to use those for TTS text to speech on MCU lately instead of LCD output for fun. Speakers where fine but buzzers was unrecognisable giberish/noise. However playing melodies is fine with them.

| improve this answer | |

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.