Question time ?

[Audio Al]

Hi

Does anyone know why speakers are built with different Ohms values ?

4 Ohms 8 Ohms and 16 Ohms making them harder to drive with different amp requirements

Sorry if its a numpty question

[Firebottle]

I have a very interesting book called 'Audio and Acoustics' by G. A. Briggs.
An interesting quote is 'Another major breakthrough to good audio was the production of the moving-coil speaker, the pioneers being Rice and Kellogg in America around 1925, although Voigt had powerful moving-coil models working in 1924'.

The only reference I can find to speaker impedances is in the chapter dealing with 'Schools and constant-volt lines'.
For PA and sound distribution systems it was (is?) common to use a 70 or 100 Volt Line, this involves using matching transformers wherever a speaker is positioned. Tappings on the transformer allowed different power levels to be drawn from the constant-volt line system.

The impedances listed are 3, 15, 327, and 670 ohms for a 15W amplifier power, plus 3, 15, 160 and 333 ohms for a 30W amplifier.

Without getting too technical this is all to achieve low losses where speakers are distributed through out a building.

Why 15 ohms was originally chosen in the good old valve era I don't know. My guess is that it enabled cheaper output transformers because the ratio of the windings could be lower, plus 15 ohms for a speaker voice coil was probably a compromise for ease of manufacture.

3 or 4 ohm speakers are/were made so that more power output could be achieved with a lower supply voltage, particularly relevant to car audio, before the likes of semiconductor BTL (bridge tied load) amplifiers.

When solid state amplifiers became more popular then 8 ohms became the compromise loading, meaning good power levels without having to have a particularly high supply voltage. Early transistors hadn't got very high voltage ratings

Alan

[Reffc]

What Alan said but also worth bearing in mind that loudspeaker driver units are FAR from linear devices, so there's really no such thing as an 8 Ohm speaker. The name really refers to the approximate nominal impedance figure which you see quoted by manufacturers.

What I mean by speakers being non-linear is that their voice coils behave very much like inductors, so impedance actually rises with frequency. A nominal 8 Ohm speaker typically might have a minimum impedance of 5 or 6 Ohms, a nominal impedance above lower resonance in the lower frequencies which averages say 6 Ohms where most of the power demand might be before climbing much higher. A typical 8 inch 8 Ohm woofer can vary from say 6 Ohms to 35 Ohms (or more), although its output obviously tails off acoustically once it has reached upper resonance.

The current requirements vary with frequency when you consider the non-linear nature of the impedance with frequency, so if you consider a fixed voltage device (amplifier) driving an "8" ohm woofer, most of the current demand will be at lower frequencies (from the basic equation V=I*R).

[Naughty Nigel]

I have a very interesting book called 'Audio and Acoustics' by G. A. Briggs.

3 or 4 ohm speakers are/were made so that more power output could be achieved with a lower supply voltage, particularly relevant to car audio, before the likes of semiconductor BTL (bridge tied load) amplifiers.

When solid state amplifiers became more popular then 8 ohms became the compromise loading, meaning good power levels without having to have a particularly high supply voltage. Early transistors hadn't got very high voltage ratings

Alan

Early transistor amplifiers used output transformers in the same way that valves do! This was partly because early [germanium] transistors had very low VCB (voltage from collector to base) ratings, with 20 or 25 volts being fairly typical. This meant that supply voltage had to be limited to about 15 volts for safe working, so driving an 8Ω or 15Ω loudspeaker to provide a reasonable output was difficult. Running these early amplifiers without a loudspeaker connected was also a recipe for disaster as open circuit voltage across the output transformer was more than enough to destroy the output transistors!

There was an additional problem in that early Germanium output transistors were only available in PNP varieties, so for many years a complimentary NPN-PNP pair was out of the question. This all meant that an output transformer was the only option if reasonable output power was to be achieved.

As a point of interest I should add that until about the mid 1960's car radios were all-valve designs, using a mechanical 'multivibrator' in an inverter circuit to generate an HT voltage of 220 volts or so to provide power for everyday radio valves. These were usually multi-element valves such as an ECL86 output valve (triode-pentode) in the amplifier section, and ECH86 as a frequency changer and IF amplifier.

(Some older members may remember these radios 'humming' or 'buzzing' thanks to the multivibrator.)

Later car radio designs used output transistors, but still driven by valves in the RF stage. 'Gas filled' valves were also developed to work with low HT voltages, but these soon gave way to new all-transistor designs apart from some military applications, which still use valves. (This is because the intense electromagnetic waves generated by a nuclear strike would destroy most semiconductor devices.)

As a further point of interest, and whilst reminiscing, the 'E' in a valve's code indicates a heater voltage of 6.3 volts; so for example, ECC83, EF86, EL84, EL34 all have a heater voltages of 6.3 volts. 'G' as in GZ32, GZ34, etc. denotes a heater voltage of 5.0 volts.

Television valves generally had a heater current of 100 mA, (denoted 'U') and a high heater-to-cathode insulation voltage, so they could be connected in series and run directly from the mains with only a huge 'mains-dropper' resistor for regulation. All this was done to avoid the cost of a mains transformer! Some of those old designs were truly horrible, and I am sure would be banned on safety and efficiency grounds today.

I should add that some early television loudspeakers were wound with very high impedance voice coils to avoid using an output transformer altogether (OTL Valve amplifiers anyone?), but I suspect these were too fragile to be of any practical use, so the idea was dropped. However, some devices still use custom built, high impedance loudspeakers where it suits the circuit topology.

However, returning to the original question about loudspeaker impedance, there is a balance to be struck between low impedance (i.e. 4Ω) which requires high current at low voltage, and higher impedance (15Ω) which requires a lower current at high voltage. Clearly there was a need to standardise these values, so 15Ω and later 8Ω (actually 7.5Ω)* loudspeakers became the early standards. 8Ω loudspeakers became popular with the advent of transistor amplifiers for the reasons already set out, whilst 4Ω loudspeakers became commonplace because of the automotive industry, where the HT voltage is usually limited to 12 volts, (although trucks and boats often use 24 V.) However, you will note that the mathematical relationship of these values allows multiple low impedance speakers to be connected in series to create a higher impedance.

Given that all early amplifiers (including portable transistor radios) used output transformers the impedance of loudspeakers could be chosen to suit the design requirements at that time. In this regard I suspect it was easier to wind voice coils and output transformers with lots of thin gauge copper wire (rather than fewer windings of thicker wire), so high impedance (15Ω) loudspeakers became the preferred option to begin with. This also meant that loudspeakers could be positioned further away from amplifiers without the resistance of loudspeaker leads posing a problem. As above, 8Ω became popular with the advent of transistor amplifiers followed later by 4Ω for the automotive industry.

PA systems still use 70 volt and (more commonly nowadays) 100 volt line circuits. As Alan says, this is to minimise losses in long circuits around schools, factories and hospitals, etc., and also allows level to be set locally by changing transformer tappings. As an example, many schools were built with an integrated radio system, where a central receiver and amplifier provided sound for the entire building. (I think this was the BBC Home Service - now R4.) This could be switched on and off, and volume controlled in each classroom, although this didn't respond too well when switched on in every room - as happened in our school when JFK was assassinated!

Right, back to 2014 . I had better get on with some real work now!

Nigel.

* The 8Ω tapping on the original Quad II amplifiers was 7.5Ω, so it was exactly half of 15Ω. However, as Alan explained the impedance of loudspeakers varies with frequency, and is not critical, so this figure was probably rounded up to avoid confusion.

[Audio Al]

Nigel

Thanks for taking the time to reply

Great answers ( it's becoming clearer now )

Cheers