There once was a "Loudspeaker" made by the Leak company that was "electrostatic" in its operation. They were used for Hi-Fi in the early 1960's - so I suppose they were good. Very slim-line, they were expensive and looked like a radiator! We don't see them anymore.
I think all modern speaker "transducers" operate on an electro-magnetic principle. We want to move the air at various frequencies to hear our voices and our music. Long slow strokes of the diaphragm for the bass, very short fast strokes for the treble. That's one reason for the many different sizes and, indeed for the development of the "CROSSOVER" circuit used in Hi-Fi to drive them. We send the higher frequencies to the small one and the lower freqs to the larger one. Sometimes there's a "Midrange" unit (speaker).
All these units are constructed on the magnet & coil principle. They are the opposite of a generator. Here we have a varying applied (ac) voltage that generates a changing current in a coil. The ensuing magnetic field alters in sympathy and reacts with it's core magnet thus causing the coil assembly to move. A thin, (was paper - now often plastic), cone is attached to move the air. Too simple? Go and have a look at one!
We need these units to be very "responsive" & that means following the applied current in an exact copy. It ain't gonna happen folks! There's a raft of problems. Not least of these is the inertia, the impetus and the IMPEDANCE. {Got there!}.
The fact that the impedance varies with frequency is a big problem. If we know the impedance at a certain spot fequency, it still only gives us a basic idea of how far the impedance will range.
CROSS-OVER DRIVE
We can use different size units, coupled with a frequency steering & impedance matching network, but we still get an awful problem of "LINEARITY" in the drive current between frequencies. A novel, much used way around that is by a device within the amplifier known as NEGATIVE FEEDBACK. Were it not for this I doubt that true High Fidelity would be possible. Even then - but wait, enough unto the day etc. MORE SOON.
NEGATIVE FEEDBACK
The effective gain (amplification) of the amplifier is altered by taking a sample of the actual output and reducing gain when it is too much and increasing it when too little. We are able to do this via frequency filters if we want to, it's just the principle we need to grasp. If we take a sample of what is delivered into the speaker (load) we can level things out. There!
Saturday 21 February 2009
Glossary of Terminology
Right here's an apology: - I'm sorry! I really should have forseen that someone would want to know what these words mean - in this context. I can/will add to this list as we may progress. For now just these: -
RESISTANCE - This is a measure of how well a DC (& sometimes an ac) current can flow in a circuit. We say that when 1 Volt causes a current of 1 Amp to flow the resistance in the circuit is 1 OHM. This then, is OHMS LAW. To get POWER in WATTS multiply the current (I) by the Volts (V).
CONDUCTANCE is the reverse of that with a unit called a MHO (mho) Get it? Don't worry about this, it rarely crops up!
IMPEDANCE - Still measured in ohms this is the complex ratio of sinusoidal (ac) voltage to current in a component or circuit consisting of two parts. The real & the imaginary. The real part is the RESISTANCE which will dissipate heat, the REACTANCE part is "imaginary" and does not dissipate.
REACTANCE - Measured in ohms this is the ac only part of any circuit or component. Characterised by the storage of energy rather than by Wattful dissipation.
NOTE:- Bear in mind that most/all practical electrical/electronic components are not "pure" in the sense that they have elements of all in them. In that regard, and for most practical circumstances, we can regard RESISTANCE as being for both DC & ac, but IMPEDANCE as being for ac circuits which are much less than pure. Usually containing INDUCTANCE and/or CAPACITANCE.
RMS - Root Mean Square. DC is what you get from a battery. It has a fixed polarity. Ohms law works! In ac power they use slip-rings on the generator to collect the current instead of a commutator - as in a motor or DC dynamo. There are advantages which outweigh the cons. Less arcing, less brush wear AND because the polarity keeps reversing (50Hz here - 60Hz in America), it is much safer for us! Muscles don't get held in relentless contraction. You can alter voltages up or down with a transformer. You can't do that with DC. The shape of the reversing wave is sinusoidal - that is relating to the cyclic operation of the generator. We need to know when this varying, & reversing voltage has the same heating effect as an equivalent DC voltage. Hence an average of what power is delivered. There's a formula which relates to the ROOT of 2. That's 1.414. The peak voltage will have to be 1.414 times higher than its DC equivalent to give the same heating effect. That means that a a 240Vac supply will have a peak of 1.414 X 240 = 339.6 Volts. In BOTH directions! That makes the Peak to Peak (p to p) Voltage = 679.4
The wave shape is traced in time like the valve on a wheel as it rolls. If the axle is at zero, the valve goes higher and lower than that describing a sine wave. Rotations / second is Hz. Think about it!
RESISTANCE - This is a measure of how well a DC (& sometimes an ac) current can flow in a circuit. We say that when 1 Volt causes a current of 1 Amp to flow the resistance in the circuit is 1 OHM. This then, is OHMS LAW. To get POWER in WATTS multiply the current (I) by the Volts (V).
CONDUCTANCE is the reverse of that with a unit called a MHO (mho) Get it? Don't worry about this, it rarely crops up!
IMPEDANCE - Still measured in ohms this is the complex ratio of sinusoidal (ac) voltage to current in a component or circuit consisting of two parts. The real & the imaginary. The real part is the RESISTANCE which will dissipate heat, the REACTANCE part is "imaginary" and does not dissipate.
REACTANCE - Measured in ohms this is the ac only part of any circuit or component. Characterised by the storage of energy rather than by Wattful dissipation.
NOTE:- Bear in mind that most/all practical electrical/electronic components are not "pure" in the sense that they have elements of all in them. In that regard, and for most practical circumstances, we can regard RESISTANCE as being for both DC & ac, but IMPEDANCE as being for ac circuits which are much less than pure. Usually containing INDUCTANCE and/or CAPACITANCE.
RMS - Root Mean Square. DC is what you get from a battery. It has a fixed polarity. Ohms law works! In ac power they use slip-rings on the generator to collect the current instead of a commutator - as in a motor or DC dynamo. There are advantages which outweigh the cons. Less arcing, less brush wear AND because the polarity keeps reversing (50Hz here - 60Hz in America), it is much safer for us! Muscles don't get held in relentless contraction. You can alter voltages up or down with a transformer. You can't do that with DC. The shape of the reversing wave is sinusoidal - that is relating to the cyclic operation of the generator. We need to know when this varying, & reversing voltage has the same heating effect as an equivalent DC voltage. Hence an average of what power is delivered. There's a formula which relates to the ROOT of 2. That's 1.414. The peak voltage will have to be 1.414 times higher than its DC equivalent to give the same heating effect. That means that a a 240Vac supply will have a peak of 1.414 X 240 = 339.6 Volts. In BOTH directions! That makes the Peak to Peak (p to p) Voltage = 679.4
The wave shape is traced in time like the valve on a wheel as it rolls. If the axle is at zero, the valve goes higher and lower than that describing a sine wave. Rotations / second is Hz. Think about it!
Friday 20 February 2009
Audio Electronics
I am often asked about various aspects of this somewhat confusing arena. I find I have to think back to all the short-cuts that I used in my occupation. Here are a few notes that I may choose to refine in the light of further interest and discussion.
HEADPHONE IMPEDANCE
Several articles quote typical 50 to 150ohms and up to 600 ohms. This is very different from my memory of so called Hi-Z types which were in Kohms. Further research needed. The impedance of my mp3 headphones is 18.8ohms each. That is at 1Khz - which is/was a common standard. Remember the XL increases with rises in frequency and you cannot get a reliable reading from a DC ohm multimeter!
So I've fixed my Inductance meter and the headphones for my mp3 player are 0.3mH and another similar set is 0.5mH. Is that sensible?
DECIBELS
This is a ratio unless you state an impedance. "Characteristic" impedance in the case of cable. We say this is, say, 22dBs up (or down). FROM WHAT? POWER & VOLTS are different when it comes to dBs. Read on. (1Watt is 1 joule / sec. This is energy as in calories).
There is confusion between various references. In particular the old Post Office Telephones standard for their nominally 600ohm network. Knowing that impedance one can calculate actual power for particular dBs (known as dBm). The standard is that 1mW into 600 ohms is set at 0dBm. The operation of the network is -13dBm. Very easy if you will settle for each 3db (of power - not volts) being double or half. So -3dB is half = 0.5mW. Do it again is 0.25mW then again is 0.125mW and one more time for -12dbm is 0.0625mW. That's the power for a telephone earpiece (well -1 more db actually). Why? Well to get 1mW we need an RMS voltage of 0.707 which is a peak ac wave of 1V. {0.707 is the root of 2} and 1/1.414 = 0.707. Now just almost forget it!
Forget it save for one more very useful thing. When people use dBs in audio it's useful to know that power (Watts) doubles for each addition of 3dB, as I just said. If one is measuring voltage however, it doubles for each rise of 6dB. So if you raise 3V by 6dB it becomes 6Volts. This Voltage calculation becomes useful because microphones for example are often quoted in dB outputs. Just as in music no-one ever says that E7 / C7 is actually a FLATTED seventh note in the scale, so it is that no-one ever says that the microphones are normalised to 1V (peak). {That's times 0.707 for RMS}.
So a -70db moving coil microphone will 70/6 = 12 X 6dB steps down approx. You divide 1Volt by 2 six times! (or by 2 to the power of 6)
1V is 1000mV so if you divide it by 128 you get 7.8mV peak as the output. Very slight error - but you can do it in your head! For RMS multiply by 0.707 = 5.5mV. Amplifiers are the same. The gain is quoted in dBs. It doesn't tell you much unless you assume the voltage going in. Now a passive guitar pick-up gives around 30mV. BUT NONE OF THIS SPEAKS OF ANY KIND OF IMPEDANCE. {An unloaded Voltage is reduced by input impedance}.
IMPEDANCE of CABLES
The capacitative reactance of screened cable has a greater effect when it shunted across a high impedance. High Frequency (HF) losses are therefore worse with long cables. {Typical value is 200pF/meter. One way around this is to use low impedance microphones with a step up matching transformer. Medium Z is 200-1000 ohms. Hi-Z is 50Kohm or more. Lo-Z is 30-50 ohms.
There is no absolutely pure impedance (Z). There's R+XC+XL in any practical component. They all contain: -
Pure RESISTANCE+Capacitive+Inductive = REACTANCE.
Z impedance formulas are for C in Farads (that's enormous and the largest unit we use is uF which is 1 millionth part or x 0.000001). For L this is in HENRYS which is also quite large. We use more mH and uH. They are 1 thousandth and 1 millionth part respectively. That's x 0.001 or x 0.000001.
SPEAKER IMPEDANCE
There's a standard. Well more than one! It started with - well no standard at all but then 16 ohms became common. With transistors came 8 & 4 ohms. The reason for the low impedance is to keep the coil assembly light & small. Then again thicker wire can be used for high Watt output powers.
For Public Address (PA - not Power Amplifier as in guitar), in buildings where there were speakers in many rooms & places, there came a "LINE" distribution and speakers were often around 100 ohms - or more - and all wired in series/parallel as needed to match to the nominal Impedance of the system. Transistorised outputs of any power are usually 8ohms, but tolerant to higher impedances without distortion.
Now then on to valve amplifiers and their speaker loads. In order to match the high output impedance of a valve (many Kohms), with a low impedance speaker, a transformer is needed. These are heavy & expensive but many musicians like the result they give. They are more commonly 16 ohms, or two 8ohm units in series. It is absolutely vital that you do not operate without the speaker load. There will be almost certain damage to the transformer from undamped back EMF. {TIP: - Good idea to use a shorting jack with a load resistor to replace the speaker if it is unplugged}. It's not so much the valve damage that you have to worry about. They are relatively rugged & cheap - even now. It's the damage to the transformer as it's interwinding insulation breaks down. They are expensive and much more difficult to fit.
MORE TO FOLLOW
HEADPHONE IMPEDANCE
Several articles quote typical 50 to 150ohms and up to 600 ohms. This is very different from my memory of so called Hi-Z types which were in Kohms. Further research needed. The impedance of my mp3 headphones is 18.8ohms each. That is at 1Khz - which is/was a common standard. Remember the XL increases with rises in frequency and you cannot get a reliable reading from a DC ohm multimeter!
So I've fixed my Inductance meter and the headphones for my mp3 player are 0.3mH and another similar set is 0.5mH. Is that sensible?
DECIBELS
This is a ratio unless you state an impedance. "Characteristic" impedance in the case of cable. We say this is, say, 22dBs up (or down). FROM WHAT? POWER & VOLTS are different when it comes to dBs. Read on. (1Watt is 1 joule / sec. This is energy as in calories).
There is confusion between various references. In particular the old Post Office Telephones standard for their nominally 600ohm network. Knowing that impedance one can calculate actual power for particular dBs (known as dBm). The standard is that 1mW into 600 ohms is set at 0dBm. The operation of the network is -13dBm. Very easy if you will settle for each 3db (of power - not volts) being double or half. So -3dB is half = 0.5mW. Do it again is 0.25mW then again is 0.125mW and one more time for -12dbm is 0.0625mW. That's the power for a telephone earpiece (well -1 more db actually). Why? Well to get 1mW we need an RMS voltage of 0.707 which is a peak ac wave of 1V. {0.707 is the root of 2} and 1/1.414 = 0.707. Now just almost forget it!
Forget it save for one more very useful thing. When people use dBs in audio it's useful to know that power (Watts) doubles for each addition of 3dB, as I just said. If one is measuring voltage however, it doubles for each rise of 6dB. So if you raise 3V by 6dB it becomes 6Volts. This Voltage calculation becomes useful because microphones for example are often quoted in dB outputs. Just as in music no-one ever says that E7 / C7 is actually a FLATTED seventh note in the scale, so it is that no-one ever says that the microphones are normalised to 1V (peak). {That's times 0.707 for RMS}.
So a -70db moving coil microphone will 70/6 = 12 X 6dB steps down approx. You divide 1Volt by 2 six times! (or by 2 to the power of 6)
1V is 1000mV so if you divide it by 128 you get 7.8mV peak as the output. Very slight error - but you can do it in your head! For RMS multiply by 0.707 = 5.5mV. Amplifiers are the same. The gain is quoted in dBs. It doesn't tell you much unless you assume the voltage going in. Now a passive guitar pick-up gives around 30mV. BUT NONE OF THIS SPEAKS OF ANY KIND OF IMPEDANCE. {An unloaded Voltage is reduced by input impedance}.
IMPEDANCE of CABLES
The capacitative reactance of screened cable has a greater effect when it shunted across a high impedance. High Frequency (HF) losses are therefore worse with long cables. {Typical value is 200pF/meter. One way around this is to use low impedance microphones with a step up matching transformer. Medium Z is 200-1000 ohms. Hi-Z is 50Kohm or more. Lo-Z is 30-50 ohms.
There is no absolutely pure impedance (Z). There's R+XC+XL in any practical component. They all contain: -
Pure RESISTANCE+Capacitive+Inductive = REACTANCE.
Z impedance formulas are for C in Farads (that's enormous and the largest unit we use is uF which is 1 millionth part or x 0.000001). For L this is in HENRYS which is also quite large. We use more mH and uH. They are 1 thousandth and 1 millionth part respectively. That's x 0.001 or x 0.000001.
SPEAKER IMPEDANCE
There's a standard. Well more than one! It started with - well no standard at all but then 16 ohms became common. With transistors came 8 & 4 ohms. The reason for the low impedance is to keep the coil assembly light & small. Then again thicker wire can be used for high Watt output powers.
For Public Address (PA - not Power Amplifier as in guitar), in buildings where there were speakers in many rooms & places, there came a "LINE" distribution and speakers were often around 100 ohms - or more - and all wired in series/parallel as needed to match to the nominal Impedance of the system. Transistorised outputs of any power are usually 8ohms, but tolerant to higher impedances without distortion.
Now then on to valve amplifiers and their speaker loads. In order to match the high output impedance of a valve (many Kohms), with a low impedance speaker, a transformer is needed. These are heavy & expensive but many musicians like the result they give. They are more commonly 16 ohms, or two 8ohm units in series. It is absolutely vital that you do not operate without the speaker load. There will be almost certain damage to the transformer from undamped back EMF. {TIP: - Good idea to use a shorting jack with a load resistor to replace the speaker if it is unplugged}. It's not so much the valve damage that you have to worry about. They are relatively rugged & cheap - even now. It's the damage to the transformer as it's interwinding insulation breaks down. They are expensive and much more difficult to fit.
MORE TO FOLLOW
Saturday 14 February 2009
Explanation
"LONDON APPRENTICE" is the general electronic engineering persona who deals with technical things in a nice peaceful fashion.
"MUSICAL BENT" can be found at Beresfordsmusicality.blogspot.com He deals with the live issues of musical performance
"SPECIAL BITTER" deals with thorny issues and is less that tolerant or pleasant. We would say he has "attitude."
"MUSICAL BENT" can be found at Beresfordsmusicality.blogspot.com He deals with the live issues of musical performance
"SPECIAL BITTER" deals with thorny issues and is less that tolerant or pleasant. We would say he has "attitude."
Subscribe to:
Posts (Atom)
