To make the electronics simpler, why not just design speakers to physically require a DC bias to achieve its neutral position. Wouldn't that potentially halve the number of transistors needed?

 

Look up Class-A amplifier design. The transistors are always conducting. Signal clarity is great, efficiency is horrendous. If the transistor is always conducting, you're going to loose a very significant amount of energy to heat. Not only will your batteries go dead very quickly, but you will also have to manage and get rid of all that heat somehow.

 

...why not just design speakers to physically require a DC bias to achieve its neutral position?

That's an oxymoron. How could you call it a neutral position if it requires energy to hold it there?

 

To make the electronics simpler, why not just design speakers to physically require a DC bias to achieve its neutral position. Wouldn't that potentially halve the number of transistors needed?

Can you make one?
How ever did you came with that idea?

 

The old tube radios and other equipment of the day used large electromagnets for the speakers. The DC from the power supply was thereby filtered as well by the large inductance.

 

To make the electronics simpler, why not just design speakers to physically require a DC bias to achieve its neutral position. Wouldn't that potentially halve the number of transistors needed?

Bad idea.
Class A amplifiers are very inefficient, and you'd have to design the speaker for the specific bias current of the amplifier design you are using.

And the cost of the additional transistors for a Class B stage is very small compared to the total amplifier cost.
You are trying to solve a non-existent problem.

 

But the "neutral" position in a mechanical system is one with no net force acting on it.
A "bias", by definition, is a net force acting to change a system to state that is NOT neutral.

The loudspeaker technology and amplifiers that drive them do not leave much room for any kind of improvement these days. Pretty much engineered to the proverbial Nth degree.

 

Agreed, dumb idea.

 

You would also limit the excursion of the speaker, which would limit it's power handling capability.

Bob

 

Back in the 1990's there was a company producing amplifiers that received position feedback from the subwoofers, and adjusted the output to make the motion of the subwoofer match the input signal. I think they were made by Alphasonik, I don't remember... can anyone remember the name, or the term? I can see them in my head, there was a coil suspended above the dust cap to read the cone position. I believe the goal was to make the enclosure volume less-significant, so you could still get big enclosure sound quality from smaller enclosures, at the expense of efficiency.

 

...and adjusted the output to make the motion of the subwoofer match the input signal.

Could you extrapolate on this? Why wouldn't it match? Does it go out of phase?

 

Motional Feedback is the term used for thee speakers Mr. S is talking about. Philips built speakers (mid 1960s) that sounded really good considering their size. They used sensors similar to tonearm cartridges as pick-ups. I've also seen dynamic mics and piezo elements in use.

Early radios (1930's) often used a field-coil type speaker. This allowed power supply electrolytics (expensive to make then) to be small. The field coil (easy to make and copper was cheap) replaced the choke in the power supply. The speaker had to have a bias to accommodate the current draw of the radio set.

You have to see these things in context, stepping stones that got use where we are today. Quite ingenious, considering what they had to work with! E

 

Could you extrapolate on this? Why wouldn't it match? Does it go out of phase?

The output electrical signal matching the input electrical signal is one thing, but the physical movement of the speaker matching the input signal is entirely different. For example, in a sealed enclosure (no vents) the movement of a subwoofer relative its input signal is extremely dependent on the airspace inside the enclosure. For example, given an arbitrary reference input, say a sine wave at 40Hz; if you place the speaker in a very small sealed enclosure, the speaker cone won't move much at all due to the great difference in air pressure relative the cone movement. If you place the same speaker with the same input signal in a very large enclosure the speaker cone will move a great deal more because the pressure variation won't be nearly as great. This very significantly affects the sound. So now you have the same input signal, but 2 very different sounds depending on the space you have available to build your speaker cabinet. The goal of the amplifier was to monitor the physical movement of the speaker and make realtime adjustments based on feedback, so that the speaker would ideally sound the same regardless of the enclosure space that you have to work with.

I never witnessed one in person, but I remember seeing the ads for them and always wanted to know if they performed as-advertised. They apparently didn't catch on, so I'm guessing they either didn't work well, or were too expensive.

 

If I'm going too far off-topic, slap my wrist. Were the speakers in 1950's televisions anything special?

 

Look up Class-A amplifier design. The transistors are always conducting. Signal clarity is great, efficiency is horrendous. If the transistor is always conducting, you're going to loose a very significant amount of energy to heat. Not only will your batteries go dead very quickly, but you will also have to manage and get rid of all that heat somehow.

That doesn't really address the question - most class-A amplifiers have an output transformer because its a really bad ide to have standing DC in a loudspeaker.

The Philips EE kits used a 150R speaker as the collector load in most audio amplifier projects. Depending on kit - the transistor was either an AC128 or a BC148, so the power was pretty low.

Not sure - but the telephone earpiece may also have standing DC.

 

To make the electronics simpler, why not just design speakers to physically require a DC bias to achieve its neutral position. Wouldn't that potentially halve the number of transistors needed?

I had a 1962 Pontiac Catalina with a Delco radio. It DID run DC through the speaker. Unfortunately, it was not a great radio. DC through a speaker does two things:
1) It creates distortion by off-setting the zero position of the cone. (Within Hooke's law, this may or may not be a problem)
2) DC saturates the voice coil reducing low frequency response.
3) Electrostatic speakers CAN work just fine with DC bias....they're just incredibly inefficient.
4) Blocking capacitors are cheap.

 

.....Back in the 1990's there was a company producing amplifiers that received position feedback from the subwoofers, and adjusted the output to make the motion of the subwoofer match the input signal. I think they were made by Alphasonik, I don't remember... can anyone remember the name, or the term?.....

Such speakers are still being made by Rythmik Audio, which uses a custom speaker having an auxiliary sense coil wound adjacent to the main voice coil to generate a signal proportional to the speaker motion.
This avoids any mechanical resonances or non-linearity that a motion sensing device mounted on the speaker cone might have.
The feedback from this winding to the custom subwoofer amp basically allows cancellation of much of the frequency response rolloff caused by the enclosure, and also reduces distortion.

I have one and it seems to work well.

 

Such speakers are still being made by Rythmik Audio, which uses a custom speaker having an auxiliary sense coil wound adjacent to the main voice coil to generate a signal proportional to the speaker motion.
This avoids any mechanical resonances or non-linearity that a motion sensing device mounted on the speaker cone might have.
The feedback from this winding to the custom subwoofer amp basically allows cancellation of much of the frequency response rolloff caused by the enclosure, and also reduces distortion.

I have one and it seems to work well.

But that's an entirely different thing. The whole idea is to include the mechanical components of the speaker in the over all feedback loop.

A sense coil only senses a DC magnetic field when the coil is moving - that would provide pretty limited information to whatever servo function existed.

 

@crutschow that's exactly what I'm talking about. The ones I remember seeing had a stationary coil suspended above the dust cap, but probably worked in the same way.

I believe you would get enough information to derive position vs. time, which would give you enough data to apply a correction to the output.

 

But that's an entirely different thing. The whole idea is to include the mechanical components of the speaker in the over all feedback loop.

A sense coil only senses a DC magnetic field when the coil is moving - that would provide pretty limited information to whatever servo function existed.

No, it is not significantly different for this application.
The coil generates a voltage when the voice coil moves in the magnetic field, which is the same as the cone movement, assuming the cone is well designed with stiffness sufficient that is does not move differently than the coil (which is reasonably easy to do for the low frequencies a subwoofer handles).
Thus the coil gives essentially the same information as a mechanical sensor mounted on the cone would.